Click here to have a similar A+ quality paper

Order Now

4.2 DV and Ease of Analysis Question 2: In your opinion, does usage of DV applications significantly ease the analysis and appreciation of all forms of data, including granular or nano-data? There was an overwhelming affirmation from all the participants surveyed so far as large data sets was concerned. The following response by a business manager participant from the FMCG industry illustrates this line of thinking: “Absolutely. There is no doubt about it at all. By and large, I do not get to use much of this information, but I ought to prepare it for senior management perusal. And, personally, I’ve seen it work wonders with large sets of data. There are commercial programs that we use and we need to feed data specifically into it and it pops out pretty visual representations of trends. That obviously eases analyses.” The following response from a top management participant in the financial service industry provides further insight: “For us top management individuals who have a lot of things to take care of, we often ought to deal with large sets of information. For instance, on a quarterly basis, I need to understand all the frauds that have taken place in a bank. Now there are numerous kinds of frauds like ATM frauds, fake accounts, personal identity thefts, internet-based fraud, credit card frauds etc. Within these kinds of frauds, across all our operations, we end up with significant number of issues. Now that is only one kind of issue that I ought to deal with. In my position, the biggest challenge is dealing with individuals, ensuring that they are on track, they are motivated to achieve the outlined goals. So, operational things take a back burner. In such an instance, data visualization comes as a big boon. We had several individuals spending a large amount of time working on data just to make sure that it is easy to understand and that the analysis is in appropriate. With DV tools, analysis becomes tremendously easy. Ofcourse, we still need to spend time on the raw data itself, which for its own reasons is collected in diverse ways, and thus requires to be cleaned thoroughly. So, in a nutshell, for large sets of data, DV tools are definitely greatly helpful.” The following response by a business manager from the retail industry indicates that the scenario is pretty similar as that in FMCG industry: “As a marketing manager, my primary focus is on ensuring that the brand is performing upto the budgets. Although, I tend to work more on marketing material and brand management team, I ought to look at sales data on a regular basis. I ought to understand which distributors are picking up our brands, which areas are performing properly, which retailers are taking our brand and so on. And then at the retail level, there is tremendous amount of consumer information, especially with regards to SKUs. So, we use DV tools mostly on sales information and consumer-level SKU purchase information. Only now are we beginning to use more recent forms of DVs like key text graphing, wherein we are mapping our market research information to provide us ready insights. Dealing with such large sets of data, I can tell you that DV has been greatly helpful. Additionally, DVs also help us present a coherent investment scenario to top management. Often top management does not have time to dwell upon the data and it might become very difficult to walk them through all the insights. However, DV aids us in mapping such information very easy and also in aesthetic formats. In essence, I can present entire sales data from our global sales on one single slide!” The following is a response by business manager from the financial services industry which connects well with the response of the top management participant from the same industry: “As a credit card departmental manager, I ought to look at sales figures, credit defaults, collections, new customer acquisition, customer attrition, customer retention, extension of the credit card to more and more vendors, customer loyalty and points and usage and so on. As you can see, on a national scale, I have to deal with tremendous amounts of information. DV tools aids in a great manner. I can look at customer acquisition, customer retention, customer attrition, and customer loyalty all in one-go with the help of DV tools and techniques. Most importantly, when presenting this information to the top management in periodic meetings, the discussion becomes more fruitful as the insight is readily visible and we needn’t run the top management through our entire analysis. Prior to using DV tools, we used to sweat out a lot to make coherent graphs and slides, and quite literally we were at a loss always putting in unholy hours. Thus DV tools have not only improved the ease of analysis, but have also reduced our workload dramatically.” The easing of burden was used as a prompt in interviews that followed the above interviews. Interestingly, this prompt provided further insights into the ease of analysis enabled by DV tools and techniques. To put things in context, the following is the response of a top management participant from the IT services industry: “Undoubtedly, DV tools are greatly beneficial when we look at large data sets. For example, on an yearly basis we look at the total number of projects completed, projects in pipeline, work-in-progress projects and so on. Since we are in a project management environment, we are pretty adamant on data collection as that helps us understand whether projects will get finished on time and in stipulated budget. And here’s where the magic of DV lies. I can take one project and put all information pertinent to it on one page and that will give me all the information that is required. So, it is in this context that I want to comment on your easing of burden point. You see, analysis is eased only when we are familiar with a certain kind of representation. For instance, a simple pie chart. That’s something we have seen for ages and we know pretty much what a pie chart involves and indicates. So, our focus is no more on the representation, but on the actual information. However, if you pop up some very fancy new thing, then things go for a toss. I remember, recently this new project manager brought an analysis with him in the format that he used in his previous company. Believe me, the first thirty minutes of the meeting, we spent on understanding what the whole thing was and only then did we begin looking at the information. Now ofcourse, if he presents information in a similar format, probably, we’ll take no more than ten minutes to orient our minds. But if he goes back to the company format, we’ll get uncomfortable again. You see, this trade-off needs to be kept in mind always.” The following response by a business manager from the financial services also shows a similar viewpoint: “I’ve been using DV tools for some time now. I also tend to do a little bit of R&D. It is much more beneficial to be aware and knowledge of what needs to be done before your subordinates pick it up. Very recently, I saw this heat map which caught my fancy. Although I didn’t have all the information to create a pitch-perfect heat map, I went ahead and produced one to show the total number of mobile phone users in the country. Now heat maps are not something that are difficult to understand. We see them all the time in newspapers. At first, when I popped it up in the boardroom, there was appreciation. The analysis and insights were readily visible. However, our top management got into the nitty-gritties of it. I believe this is called xenophobia or some related term. Standing there in the boardroom, I thought to myself-damn, why didn’t I expect such a situation. It is much easier to present information in ways that one is used to. Nevertheless, only by presenting new formats and getting people acquainted to them, is the way forward.” However, some participants observed that since there is no common base for most recent DVs, investment of time required to understand the visualized data negates the benefits of the visualization. In terms of old forms of DVs (e.g. time-series, spider-charts etc.), all participants praised DVs as being the easiest method of analysis. With regards to granular or nano-data, the following is the response of a top-management participant from the financial services industry: “I would say that DV is definitely very useful even for granular data. For instance, if I wanted to understand the ATM transactions of a particular customer, a visual representation would be very helpful to see the trends and patterns. But with respect to nano-data, I don’t DV would be significantly helpful. For instance, in the case of a particular customer, if I wanted to check their cheque book requests across a time span, visualization of that data is as helpful as the raw data itself, simply because the quantum of information is very little. In fact, I would say that DV for nano-data would be more of a nuisance that actual use. Although the same can be said about granular data, I believe that by and large, for granular data DV definitely are very helpful.” Top management participants and almost all of the line managers presented similar perspectives of applicability of DV to granular and nano-data. The following response from a line manager from the retail industry exemplifies this: “Sometimes, you just to look at raw numbers. See, DVs are deemed to very helpful when you trying to identify trends or patterns, i.e. when there is need for analysis. Most often with nano-data, e.g. sales of brand X on Sunday across each hour, there is actually no need of an analysis. Here, the raw numbers are required. I would say that as nano data generally doesn’t require analysis, the usage of DV in such a case is not recommended.” The following response by a top management participant from FMCG industry also expresses similar viewpoint: “By granular or nano-data, I believe that you are referring to information pertaining to small sets like say sale of SKU x of brand y in territory z for the first week of January. In such a scenario, I don’t think that DVs are really require. I mean, when given such a kind of information, I’d rather worry about the absolute number than proportional representation which is what DVs do. However, if you are suggesting that this information is presented in comparison to sale of SKU x of brand y in territory z across the year, then maybe, there could be a hidden trend. But then again, we are talking about DVs dealing with large sets of data. So, I’m sorry, but I don’t see the advantage of using DVs for granular or nanoo-data.” The information from the IT services industry also was on the same line although contextually phrased. The following is a response of a business manager participant from the IT services industry: “When you say granular or nano-data, I’m interpreting it as information on a very minute scale. Say, the total number of developers used on a particular day on a particular project. In these cases, there isn’t really a trend that we want to identify. And that is the strength of DVs, isn’t it? Identifying hidden trends. So, when there are not really any hidden trends, then why would one put the energy to use DVs at all? In this case, I’d rather go with the raw information vis-a-vis a pretty visual presentation.” To generalize, the objections to usage of DV for nano-data were the quantum of information, no real requirement for analysis, higher utility of raw information vis-a-vis DVs and stronger merits of data as compared to visualization. 4.3 DV and Gut-Driven Decision Making Question 3: DVs move managements away from decisions that are gut-driven to those that are data or evidence-driven. Your opinion on that statement is… Most of the line managers agreed with the statement in a limited manner. The following response by a line manager from the FMCG company indicates the discord with the statement: “Undoubtedly on the face of it DVs definitely aid in having data or evidence-driven decision making. However, if you consider the fact that most decisions that are required to be taken pertain to people, then your statement goes for toss. Because when it is related to people, data or evidence is not always the best way of taking decisions. For instance, if I had to understand the productivity of a particular department in the store, sure DV can make a fantastic infographic, but the corrective measures would be based not on data, but on gut as you mentioned.” Some of the other line managers also indicated similar responses. The following response by line manager from the retail industry provides another angle to the discussion: “DV tools can definitely present tremendous insights into the data which has been gathered. However, when you say business decisions, they normally pertain to more than one than one area. For example, If I wanted to remove brand Z from the top-shelf space of a premier location in a retail store, then I ought to know which other brands can take that spot. I need to understand what are the costs and benefits of such an action. In a management simulation or a school, the problem can be neatly outlined as evaluate the costs and benefits of replacing brand Z with other brands in the store at a super premium location in the retail space. However, most situations in life cannot be framed so neatly. So, while in the former case DV seems to be an elegant and useful way out, in real life it fails at several levels. Now, there might possibly be an insight about other brands that can be replaced into that spot. But the essence of the problem is that it is not articulated at a certain point in time. So only under certain conditions, the first insight is not highly relevant until the second insight. DV can give me both the insights, but it cannot connect the dots between the two insights. I would say that DV and context put together, are perhaps better suited to evidence-driven decision making as compared to only DV.” Evidently, the above manager indicated that DV can definitely provide insights, but without the right context, such insights might not be actionable. Thus for DVs to provide relevant insights, the context ought to be there for the insight to become meaningful. An additional problem with DVs is that the problem cannot be stated neatly and coherently in real life situations. In other words, often business managers have no idea of what they are solving. Thus, even one insight might not be very useful unless other insights are present. Again, the contextual nature and timeliness of insight presentation is also relevant to decision making. Another area where line managers felt that DVs and data-driven decision making is not necessary is that of routine decisions or areas where processes and procedures have already been in place for a large amount of time. As an example consider the following response by a business manager from the financial services industry: “DVs do give us evidence-derived insights. However, can such insights always be actionable? For instance, sometimes we already know what the issue might be and we might be looking at the data merely to confirm our hypothesis. Say for instance, sales incentives. We have for long working with various kinds of sales incentives and we know what kind of incentives are working and what kinds are not. If an incentive is working only to a certain extent, then we know how to modulate it for desired effects. In such scenarios, we might use DVs to present us with insights. Our understanding might be altered in a minor way or simply we might get affirmation of our hunches or gut-instincts. Then, we can go ahead and take the pre-devised decision. In these cases, yes, DVs definitely move us towards evidence-driven decision making as compared to gut-driven decision making. However, not always do we have the liberty of time or the luxury of pre-conceived problems. There are several decisions which need to be taken on the spur of the moment. In most of these cases, it is not reasonable to expect a business manager to go in for a thorough data analysis and then take a decision. Here, gut instincts are all that counts. To simplify, for routine operational decisions where we have the benefit of experience, DVs are not necessary good or put it another way, gut-driven decision making is more important than evidence-driven decision making, thus removing the necessity to use DVs.” To simplify the above response, DVs are not required in several decision making processes which are entrenched in the industry with well-established escalation and decision-making procedures. In such a circumstance, DVs definitely do not add value through evidence-derived insights that lead to decisions. From the other responses, it seemed that DVs have limited scope so far as driving evidence-driven decision making is concerned. The luxury of time that allows for neatly framing a given situation or issue is not always available, thereby reducing the utility of DVs. The top management on the other hand had a completely different perspective. The following response by a top management participant from the FMCG industry showcases this line of thinking: “So you are suggesting that machines should overtake us now? I’m not completely against the notion as I’ve seen automation increase and replace manpower in this industry. Definitely, DV exemplifies analysis of a given situation. But then, people like us would always be required to overlay our crystallized experience of years when it comes to decision-making. So, we do use DVs which are based on data. But when it comes to taking decisions, especially strategic decisions, it is always a thing of chance. Maybe someday informatics would become so pervasive that the probability of success of taking a particular route is known before hand. Even in that case, I do envision manual over-ride every now and then. And we are nowhere close to such a futuristic situation. So, I’m going to disagree with your statement.” The following response by a top management participant from the retail industry provides further insight into this line of thinking: “Business decisions often affect several stakeholders. So DV can bring to me information about a particular set of stakeholders. That is definitely evidence-driven. Does that make the decision-making process simpler? Or does it entirely command the decision-making process? In my experience, seldom do such situations arise where evidence-driven insight is sufficient to take a decision. For every decision taken, we need to weigh in the perspectives for a wide range of stakeholders who are bound to be affected by a single decision. For instance, say that I want to increase the purchase a particular brand because DV shows that the sales of this brand are very strong. Further analysis also reveals that sales are happening from select stores. So, in essence my decision boils down to increasing purchase of a particular brand and increasing the inventory of that brand at certain stores. Now, my internal competing brand will suffer because of it. How do I estimate the effects of such a decision on the internal brand? I need to change the distribution a little bit more so as to carry that much extra amount of that particular brand. How does the decision effect my supply chain? Also, do I go with more shelf-space or just more stock? Can I circumvent loss of sale due to lower inventory levels through other means? Here, the DV has given me the insight that I ought to buy more of a particular brand. But by no means has it simplified the decision making process. In several cases, we take decisions based on gut-instincts and no matter how good DVs become, they’ll not be able to replace gut-driven decision making.” To simplify the above response, data-driven or evidence-driven decision making is not entirely possible as decision making needs to be done with the interests of several stakeholders, several conflicts or trade-offs in mind. Thus, no matter how effect DVs become in ensuring evidence-driven insights, business decision making cannot rely on DVs. The following response by a top management participant from the IT services industry showcases another opposition to the given statement: “Business decision is not like law and order, and the justice department of states! You simply look at evidence and then you draw your conclusions. It is not as simple as that. DVs today, by far, are being used to analyze historical information. In other words, they are not powered to give us projections or scenarios of the future. When they are powered to do so, they often take into account the gut-instincts of individuals making the projections. What use is business decision making based on historical information? Often, business decision making entails taking decision that will affect the future direction of the department or the company. So, in that case, we ought to make several assumptions as no one has ever seen the future. In other words, we are dealing with a tremendous degree of ambiguity and uncertainty. In such cases, there is no more evidence that is driving the thinking process but gut-instincts or hunches. Most projections everywhere are based on assumptions which are in essence gut-instincts. So, I disagree with your statement that DVs are moving us towards evidence-driven decision making.” The above respondent clearly indicated the major shortcoming of DVs in regards to business decision making: lack of evidence-driven assumptions. Although, historical data can be used as a proxy for future trends, some experiential overlay is necessary to make them sound realistic, which implies that business decision making is bound to be gut-driven to a large extent. The following response by a top management participant from the financial services industry also indicates discord, however in very different words and thoughts: “Business intelligence has definitely come a long way since its inception. Today, your enterprise tools are combining information from different departments and giving us very valuable insights into our business. However, this is not happening because of automation. There is natural intelligence that says okay let’s combine this and that, and check what is the insight. So, the decision that is borne from such a process cannot be completely be credited as that derived from evidence. There is gut-instinct somewhere in the process. Similarly, when it comes to final decision-making, one cannot do away with gut-instinct entirely despite very strong evidence-driven insights at one’s disposal.” The above clearly indicates that though DV has come a long way, it still has a long way to go given the shortcoming in artificial intelligence that powers such tools and processes. The primary objection to DVs leading to data-driven decision making by top management participants was the necessity to overlay experiential wisdom on the data analyzed and then take decisions based on hunches due to lack of tools that provide probability of success. The other objections raised included dealing with ambiguity and uncertainty while taking decisions which necessarily requires gut-instincts, inherent shortcoming in the artificial intelligence that powers the DV tools and techniques, and applicability of DVs only to historical information that ends up providing merely a proxy for decision making but not all the evidence required for the decision. 4.4 DV and Problem Identification Question 4: Do you think that DVs ease problem identification and formulation, and data preparation and exploration? Again, there was an strong agreement with the given statement from all respondents without exception. The following response by a line manager participant in the FMCG industry exemplifies this: “Problem identification is by far the greatest advantage that DV brings to the table. The essence of DV is identification of concealed insights in the data gathered. Once data sets are put in visual format, one can readily see how the things are moving and panning out in practice, which directly leads to problem identification. In many cases, a set of DVs also leads to problem formulation. I strongly agree with your statement.” The following response by a business manager from retail services industry further substantiates this line of thinking: “Absolutely. DVs make the problem pop out from heaps of data. There is a popular adage: a picture is worth a thousand words and DV personify that adage. You pool in consumer purchases of a particular brand across a particular day across all our stores in comparison to another brand across the same day and across all the same stores, and we might see that one brand did not sell as much as another brand. That is just one example, but there is inherently a problem in there. So, DV shows us that there is a problem. Problem formulation can also be achieved through DVs. For instance, in the above DV, if we had placed discount information on both the brands, we might find that discount was a problem area. Similarly, if more than one criteria were to be placed onto the DV, although it might be a bit difficult to read and interpret, problem identification and formulation becomes very easy.” Despite a general agreement with the first part of the stated assertion, there was hesitation from some of the participants. The following response by a top management participant from the IT industry shows this line of thinking: “Problem identification is very easy only if the interpretation of the DV is straight-forward. Now, back in the olden days, we could look at three or four bar charts and say that there is a problem area. However, current jazzy DVs negate this effect simply because they are too jazzy. One needs to invest time in first getting friendly with the given DV so as to be able to leverage it to understand DVs. That being said, I’ll agree with your statement because, at my position, I get the problem statement as the title or the footer of the page and then I see the DV. Obviously to me, the problem becomes very obvious simply by looking at the DV then.” The above response clearly indicates that ease of understanding the DV determines the ease of problem identification based on the DV. The respondent clearly indicates that difficulty to understand DVs negate one of the important benefits that is problem identification. However, this is limited to newer forms of DVs which are not entrenched in psychology of business managers unlike the old forms of simple DVs like pie charts, bar charts etc. The following response by a top management participant from the financial services industry also indicates dissonance with the first part of the given assertion: “DVs definitely help in problem identification, I will not contest that. But, I definitely do not agree that they are helpful in problem formulation.” Upon further probing, the participant continued: “A series of line graphs can tell me the trend of my company’s ROI across the years. In it, if I notice that there is a dip in a particular year, then I know that there is a problem there. Now, if I were to look at the trend of my returns or my investment over the same period of time, I’ll know where the problem lies. So, DVs are definitely helpful in problem identification. But I ought to look at returns or investments in juxtaposition to my ROI trend. This is not done by a DV. This is done by human intelligence. No matter how good DVs and business intelligence gets, there needs to be human intelligence in the background that is formulating problems and connecting the right set of data with each other. Only then do DVs deliver in problem identification.” The above response hints at the fact the DVs cannot be automated to connect all sorts of data that are captured in an organization. Thereby, they cannot aid in problem identification. The participant also insisted that problem formulation is not the realm of DVs but that of humans who are engaging (using) the DV. With regards to data preparation and exploration, there were some inhibitions from the participants. The following response by a line manager participant from the financial services industry showcases this: “It is not always possible to gather data in the way that these tools require them or how we require them based on the analysis desired. Just that the raw data forms are near infinite, I don’t see how DV tools and techniques can encompass all forms of raw information. And also, data cleansing is a big activity. This is an area where there is tremendous scope for improvement for DV tools. Frankly, I don’t know how and when this barrier shall be broken down, but it definitely is required. This single thing, data preparation is so time consuming and so frustrating that it is limiting the widespread application of DV tools and techniques.” The above response indicates that data preparation and cleansing is a difficult task as the raw data is captured in numerous forms. This implies that all that data must first be brought to the same basis before putting it through analysis of any kind, and more so in the case of DV where rigor is much higher. The following is the response of a top management participant from the retail industry: “Data preparation is always a big headache. Different kinds of data are gathered by different departments and this is dynamic, as each manager has a preference for a particular style. Thus, ensuring that the data captured is on the same units and basis as required for a DV analysis is not always so straightforward. To an extent, this process can be automated. However, we’ll need a subjective overlay each and everytime to ensure that the cleansing has happened properly. And I’ve already pointed out the problem formulation bit. I believe data exploration leads to that.” One of the reasons why raw data is captured in numerous forms is managerial preferences and dynamism in capturing data due to managerial attrition. Each incoming manager prefers to see data in particular formats, and thus the raw data that is captured and the subsequent analysis tends to change with time. The following is the response of a top management participant from the FMCG industry: “We have machines to capture information in our supply. We have sales force and human entry to capture data at the retail level. We have order systems to capture data at the distributor level. We have automated systems to capture information at the manufacturing level. We have manual systems to capture information within the organization. Evidently, there are numerous means of capturing data. In other words, the rigor with which data is captured and the error in data collection varies across the organization. So, it is easier to conduct DV analysis for data within one system, say manufacturing. However, that is a mediocre use of DV. If one were to analyse data at an holistic organization level, then obvious one needs to invest in thorough data preparation. Like I said, since the rigor changes with data collection systems and processes, the manual input required in data preparation is also a variant. At each system level, it does not make business sense to invest human capital simply to prepare data, for one might never use the data that is available. It makes more sense to prepare only that data that leads to problem formulation, which means that the persons accountable for identifying business problems and developing solutions need only deal with data preparation and exploration.” The participant from the FMCG industry indicated the large investment in data collection systems and processes. At the same time, it was confessed that despite such a high investment in data collection, a similar investment is not made in data preparation and exploration as it does not make business sense to make such a human capital investment. A business manager from the IT services industry gave the following response to the data preparation and exploration question: “Our operations span a variety of industries and consequently deal with a diverse set of industries. Our operations are not only culture-dependent, but also time-bound as work happens in different times. When you deal on a such a global scale, data collected simply cannot be on the same terms. Someone in Spain might be willing to provide some kind of data while a business manager in India can only capture data in a certain format. Thus, the scale and diversity of operations leads to data complexity. Given such a complexity, the amount of time that needs to be invested in data preparation is humungous. Simply mind-boggling. Although, we are from the computing industry, I can assure you that it is still a humungous requirement. Forget setting up the appropriate systems, the continuous maintenance of such systems simply makes investment in data preparation and exploration completely a loss making proposition. Also, when you say data exploration, it is not an easy job. The one who needs to drive analysis should be the one who needs to say that they want to look at particular kinds of data. A machine, no matter how smart, still cannot do this job in a satisfactory manner. Don’t get me wrong. Like I said, we are from the computing industry and we understand the capabilities and limitations of artificial intelligence. And I can assure you that data exploration is still not the domain of machines.” The above response clearly indicates another reason for the humungous diversity of data–scale and diversity of operations. There is an indication that data preparation and exploration systems and teams can be set-up. However, continuous running of the such systems and teams would negate the benefits accrued from such a set-up, as data analysis is only a small part of business operations. This clearly shows that investments in data preparation and exploration cannot be developed as a coherent business case. To summarize DV and problem identification: most participants felt that DV tools have scope for improvement in terms of data preparation and exploration. Additionally, dedicated resource of data crunching and exploration is not possible in all industries, despite their capital bases. The evidence from the research indicates that this is one of the primary reasons for lack of widespread adoption of DV tools and techniques. 4.5 DV and Profitability Question 5: In your view, do DV applications simplify the evaluation and monitoring of returns? Question 6: In your opinion, how is usage of DVs associated with the profitability of the company? With regards to the DV and returns, almost all participants requested for further clarity. It was made known to them that the question sought returns with regards to improvement in business processes after applying DVs to such businesses processes. The response by a line manager from the retail industry is reproduced in verbatim: “That is an interesting question. One definitely needs to measure returns to understand the benefits of something. I have never thought about it earlier. But let me think aloud. We had began using DV to understand consumer spending patterns on discounts of national brands and our in-store brands, spending patterns across various times of the day, and spending patterns across the stores. Now that I’ve spelled it out, I don’t think we have applied DVs to specific processes. We are merely measuring end results and then going back to study the value chain or process flow. So, I really cannot answer this question.” The above response presents a very interesting facet of DV. Most visualization are applied to end results followed by a tracking of the entire value flow to isolate the underlying problem. Most of the subsequent analysis might not be engaging visualization. This clearly indicates that visualization is not being applied to entire data processes. The following is the response of a business manager from financial services industry that indicates another line of thought: “In my opinion, trying to understand returns of is simplified only when the process itself is streamlined and one understand the exact investment and the output. Then such investments and outputs need to be measured objectively. What happens practically is that not all processes are streamlined. And for the processes that are streamlined, in actual operations, there are always deviations. In such a dynamic scenario, measurement of inputs and outputs becomes very difficult. Though, we can use DV today, tomorrow the same DV might not be applicable as additional data inputs have been understood in the process. This is almost always the case as human input is a critical part of the process and measuring this objectively is always a challenge. So, in the initial stages we assume it to be not significant. However, as time progresses, someone says that measuring human input in this way makes sense, so let use it in the evaluation of returns. So then, the old DV needs to be updated. You see, there are too many elements that are changing in this picture. So, simplified is not a word that I would want to use.” The above response clearly indicates that processes themselves need to be streamlined for a DV to work. It is reported that though most processes are streamlined, in actual practice there are bound to be deviations. Such deviations cause errors in measurements of inputs and outputs, and sometimes also lead to addition or deletion of data sets. Thus, a DV that has been used once might not be applicable at a later stage. The following is a response by a top management participant from the IT services industry: “Simplification happens when things are well understood. Definitely DV can aid in evaluation and monitoring of returns of business processes. However, I would contest the word simplify here. You see, new kinds of DVs are not standardized. Unless we were using pie charts and simple bar charts all the time, we are not simplifying anything at all. I’ve already given you the example of a project manager who used something we were not aware of at all. In that case, it was really difficult to understand the kind of returns that we were getting on the investments in our business processes.” The above argument states that since the field of DV is itself in a state of flux with new visual representations being introduced constantly, there is no simplification in the usage. On the contrary, the introduction of newer kinds of visualizations coupled with the fact that the users themselves are not thoroughly trained in handling such visualizations, is complicating the matter further. The above responses clearly point out to the fact that since DV is not standardized, evaluation and monitoring of returns is not being simplified currently. The following response by a top management participant from the FMCG industry indicates that line of thinking: “If processes are streamlined and the method of analysis frozen, then one can definitely say that DV aids the evaluation and monitoring if returns.” Upon further probing for elaboration, the participant continued: “Say, I wanted to evaluate frauds that are occurring in our credit card division. Through a series of analysis, let’s say that we determined that a particular physical component on the cards is making the card vulnerable to thefts. So we decide to invest in a superior material and improved tracking. All through we have been using the same DVs to understand the investments, total fraud cases and overall returns. In time, we notice that there were infact positive returns in terms of reduced number of frauds and improved customer satisfaction. Now, in this case the processes are streamlined, as in we know what is going in, how it is flowing in the value stream and what is coming out. Additionally, we are using the same set of DV tools and analysis all throughout. In that case, the DVs generated are easy to interpret.” Most participants were of the opinion that DV does aid in evaluation and monitoring of returns if the processes are streamlined. The key point of contention was that since DV is itself an evolving field with new tools and techniques coming into place, the analysis and presentation of information is in a state of flux. Given such a dynamic scenario, consolidation of specific methods of analysis becomes difficult. This consequently leads to evaluation and monitoring of returns using different techniques all the time. Thus, the participants were cautious to opine that DVs simplify evaluation and monitoring of returns. However, when probed further, participants accepted that once a particular kind of analysis is deemed appropriate and frozen, DV greatly aids in understanding their returns. In terms of association of DV with profitability, most participants were of the opinion that DV definitely contributed to time and associated cost savings. Some participants cited that improved decision making due to DV does aid in improving top-line as well. The following response by a line manager in IT services industry presents this point of view: “I have seen DV contribute to improvement in top-line. For instance, last year, we analyzed our sales patterns with our top quartile clients. This data was analyzed used different kinds of DVs. From that analysis we learnt that certain kinds of customers were ordering only certain kinds of services, time and again. When we dug deeper into that analysis again with DVs, we realized that our track-record in those services has been brilliant. Although, there was a tacit understanding that the service line in discussion was a core competency for us, there was really no data-driven proof for it. However, post that analysis, we developed off-shoots for that service line. Again from the DV, we understood which clients should be approached first. And the results were very encouraging.” The above response indicates that DV has aided in top-line contributions. It is this improvement that can be considered as a contribution to profitability as improvements in revenues is bound to improve profits. The following response by a top management participant from the financial services industry indicates a different point: “We have repeatedly used DVs to understand the cost structures of our various strategic business units (SBUs). Firstly, the DVs aid us in understanding what is happening in the costs of each SBU and secondly, they provide a means of comparison across the SBUs. Such an analysis obviously leads to thoughts like “If SBU A is using similar processes as that of SBU B, then why does SBU A have a higher cost structure”, or “If SBU A has used this much in the past, why is it using so much now”. Such an evaluation obviously leads to further analysis from which we can isolate the actual problems. Once we understand the inherent problems in the processes, we obviously correct it some way or another. This leads directly to cost savings. And yes, that leads to improved profitability. On a one-on-one basis, I would say that the contribution is too meagre. On an aggregate basis both on time and across SBUs, though the contribution is low, it is still considerable. In such times of economic crunch and financial prudence, even such contributions are credit-worthy. Also, I would envision that DVs would have a stronger role in profitability the more they proliferate across business processes.” The above response clearly indicates that DVs have aided in cost savings that directly lead to improvement in profitability. Similar responses were also obtained from some other participants and most of them reported that the contribution is still very low in this regard. However, given the requirement of thorough cost control, even such contributions are greatly appreciated. The following response by business manager from the retail industry indicates a different point of view: “I would say that the biggest advantage of DVs is that of time savings. Despite the initial investment that needs to be made in understanding a particular kind of DV and data preparation, I still think that it is a good investment. Once something is standardized, it becomes a matter of plug and play. For instance, I began using a heat map to understand the sales of a particular brand across all our retail outlets in the nation. The first two or three times, we did it, it did take a lot of time. Now, it is simply a plug and play thing–clean the data and feed it, and voila, we have the analysis out. We’ve been using that analysis for the last one year. Once we have that time saving, we can invest it in other activities. I’m not saying that increased time in other activities will impact the profitability directly. However, there has got to be some kind of butterfly effect in place, right?” Time saving was cited as one of the biggest advantages of using DVs. Although, some participants were adamantly against the increased time investment required in understanding and data preparation required for the DV, almost all participants agreed that once the DV is standardized for a particular kind of analysis, it definitely results in time savings in subsequent usages. Most participants were of the opinion that the time saved due to DVs could be invested in other productive activities that improved the overall productivity, and subsequently profitability. However, not all participants could point out the link between productivity improvements and profitability improvements. The scenario in the FMCG industry was slightly different. The following is a response of a top management participant from the FMCG industry: “Undoubtedly, once a specific kind of analysis is frozen, it definitely aids in improved cost and time savings. However, the kind of data sets that we see in the FMCG industry, it is not always possible to create simply a plug-and-play solution. We need thorough quality control and quality assurance before we can even use the data in the analysis. It is this specific aspect of the FMCG industry that I think makes DVs less effective.” The researcher presented the viewpoint that in the retail industry as well the data sets are very large, but in that industry, DVs are seen to beneficial to profitability through time savings. To this, the above participant continued: “Undoubtedly, the data sets even in the case of the retail industry are humungous. But you also need to understand that a majority of that industry is automated. Everywhere the data is being captured through computers. In the case of FMCG, we are dealing with too many people. Yes, the people are using computers too at almost every point. But the very fact that there is so much reliance on human entry necessitates that we perform quality control and assurance each and every single time. This would be the primary difference in my opinion.” In the FMCG industry, humungous data sets and contrarily lack of data were cited as major deterrents. 4.6 DV and Overall Decision Making Question 7: Decisions made on the basis of DV are significantly more precise, informed and inclusive. Your thoughts. All participants agreed with the given statement with some inhibitions. Participants felt that DVs generally combine varying sets of data to generate one visual representation, they become inclusive by default. At the same time, since underlying data is being used to create the representations, they ought to be informed. In terms of precision of the output, some participants felt that it was dependent on the creator of such representations and could not be relied upon always. The following response by a top management participant from the financial services industry indicates this line of thinking: “I agree that DV generally ensures inclusiveness and leads to informed decisions. However, in terms of precision, I have my doubts. Several times, I’ve had to spend time in understanding a particular visualization. And once I see something new, I’m keen on understanding it entirely, especially when it is concerned with business. Often, I’ve seen that the precision of such visualizations is not very good. Sometimes, I feel that more data should have been included. Sometimes, I feel that less data should have been included. Sometimes, I feel that the person handling the generation of that visualization doesn’t entirely understand its merits and thereby leads to suboptimal usage. You see, I can go on and one about the precision of DVs. Thus, I would not agree that decisions made on the basis of DVs are always precise since DVs themselves are not always precise.” The primary concern with regards to the precision of decision made using DV is the precision of DVs itself. Additionally, some participants stated that the user of the DVs–the one who uses the analysis–is not always in sync with the visual representation as they are a relatively new phenomenon. Given the varied personalities that might be using the same DV to take varied decision as specific to their context, there cannot be one consensus on the right interpretation of the given visualization.

4.2 DV and Ease of Analysis

Question 2: In your opinion, does usage of DV applications significantly ease the analysis and appreciation of all forms of data, including granular or nano-data?

There was an overwhelming affirmation from all the participants surveyed so far as large data sets was concerned. The following response by a business manager participant from the FMCG industry illustrates this line of thinking:

Absolutely. There is no doubt about it at all. By and large, I do not get to use much of this information, but I ought to prepare it for senior management perusal. And, personally, I’ve seen it work wonders with large sets of data. There are commercial programs that we use and we need to feed data specifically into it and it pops out pretty visual representations of trends. That obviously eases analyses.”

The following response from a top management participant in the financial service industry provides further insight:

For us top management individuals who have a lot of things to take care of, we often ought to deal with large sets of information. For instance, on a quarterly basis, I need to understand all the frauds that have taken place in a bank. Now there are numerous kinds of frauds like ATM frauds, fake accounts, personal identity thefts, internet-based fraud, credit card frauds etc. Within these kinds of frauds, across all our operations, we end up with significant number of issues. Now that is only one kind of issue that I ought to deal with. In my position, the biggest challenge is dealing with individuals, ensuring that they are on track, they are motivated to achieve the outlined goals. So, operational things take a back burner. In such an instance, data visualization comes as a big boon. We had several individuals spending a large amount of time working on data just to make sure that it is easy to understand and that the analysis is in appropriate. With DV tools, analysis becomes tremendously easy. Ofcourse, we still need to spend time on the raw data itself, which for its own reasons is collected in diverse ways, and thus requires to be cleaned thoroughly. So, in a nutshell, for large sets of data, DV tools are definitely greatly helpful.”

The following response by a business manager from the retail industry indicates that the scenario is pretty similar as that in FMCG industry:

As a marketing manager, my primary focus is on ensuring that the brand is performing upto the budgets. Although, I tend to work more on marketing material and brand management team, I ought to look at sales data on a regular basis. I ought to understand which distributors are picking up our brands, which areas are performing properly, which retailers are taking our brand and so on. And then at the retail level, there is tremendous amount of consumer information, especially with regards to SKUs. So, we use DV tools mostly on sales information and consumer-level SKU purchase information. Only now are we beginning to use more recent forms of DVs like key text graphing, wherein we are mapping our market research information to provide us ready insights. Dealing with such large sets of data, I can tell you that DV has been greatly helpful. Additionally, DVs also help us present a coherent investment scenario to top management. Often top management does not have time to dwell upon the data and it might become very difficult to walk them through all the insights. However, DV aids us in mapping such information very easy and also in aesthetic formats. In essence, I can present entire sales data from our global sales on one single slide!”

The following is a response by business manager from the financial services industry which connects well with the response of the top management participant from the same industry:

As a credit card departmental manager, I ought to look at sales figures, credit defaults, collections, new customer acquisition, customer attrition, customer retention, extension of the credit card to more and more vendors, customer loyalty and points and usage and so on. As you can see, on a national scale, I have to deal with tremendous amounts of information. DV tools aids in a great manner. I can look at customer acquisition, customer retention, customer attrition, and customer loyalty all in one-go with the help of DV tools and techniques. Most importantly, when presenting this information to the top management in periodic meetings, the discussion becomes more fruitful as the insight is readily visible and we needn’t run the top management through our entire analysis. Prior to using DV tools, we used to sweat out a lot to make coherent graphs and slides, and quite literally we were at a loss always putting in unholy hours. Thus DV tools have not only improved the ease of analysis, but have also reduced our workload dramatically.”

The easing of burden was used as a prompt in interviews that followed the above interviews. Interestingly, this prompt provided further insights into the ease of analysis enabled by DV tools and techniques. To put things in context, the following is the response of a top management participant from the IT services industry:

Undoubtedly, DV tools are greatly beneficial when we look at large data sets. For example, on an yearly basis we look at the total number of projects completed, projects in pipeline, work-in-progress projects and so on. Since we are in a project management environment, we are pretty adamant on data collection as that helps us understand whether projects will get finished on time and in stipulated budget. And here’s where the magic of DV lies. I can take one project and put all information pertinent to it on one page and that will give me all the information that is required. So, it is in this context that I want to comment on your easing of burden point. You see, analysis is eased only when we are familiar with a certain kind of representation. For instance, a simple pie chart. That’s something we have seen for ages and we know pretty much what a pie chart involves and indicates. So, our focus is no more on the representation, but on the actual information. However, if you pop up some very fancy new thing, then things go for a toss. I remember, recently this new project manager brought an analysis with him in the format that he used in his previous company. Believe me, the first thirty minutes of the meeting, we spent on understanding what the whole thing was and only then did we begin looking at the information. Now ofcourse, if he presents information in a similar format, probably, we’ll take no more than ten minutes to orient our minds. But if he goes back to the company format, we’ll get uncomfortable again. You see, this trade-off needs to be kept in mind always.”

The following response by a business manager from the financial services also shows a similar viewpoint:

I’ve been using DV tools for some time now. I also tend to do a little bit of R&D. It is much more beneficial to be aware and knowledge of what needs to be done before your subordinates pick it up. Very recently, I saw this heat map which caught my fancy. Although I didn’t have all the information to create a pitch-perfect heat map, I went ahead and produced one to show the total number of mobile phone users in the country. Now heat maps are not something that are difficult to understand. We see them all the time in newspapers. At first, when I popped it up in the boardroom, there was appreciation. The analysis and insights were readily visible. However, our top management got into the nitty-gritties of it. I believe this is called xenophobia or some related term. Standing there in the boardroom, I thought to myself-damn, why didn’t I expect such a situation. It is much easier to present information in ways that one is used to. Nevertheless, only by presenting new formats and getting people acquainted to them, is the way forward.”

However, some participants observed that since there is no common base for most recent DVs, investment of time required to understand the visualized data negates the benefits of the visualization. In terms of old forms of DVs (e.g. time-series, spider-charts etc.), all participants praised DVs as being the easiest method of analysis.

With regards to granular or nano-data, the following is the response of a top-management participant from the financial services industry:

I would say that DV is definitely very useful even for granular data. For instance, if I wanted to understand the ATM transactions of a particular customer, a visual representation would be very helpful to see the trends and patterns. But with respect to nano-data, I don’t DV would be significantly helpful. For instance, in the case of a particular customer, if I wanted to check their cheque book requests across a time span, visualization of that data is as helpful as the raw data itself, simply because the quantum of information is very little. In fact, I would say that DV for nano-data would be more of a nuisance that actual use. Although the same can be said about granular data, I believe that by and large, for granular data DV definitely are very helpful.”

Top management participants and almost all of the line managers presented similar perspectives of applicability of DV to granular and nano-data. The following response from a line manager from the retail industry exemplifies this:

Sometimes, you just to look at raw numbers. See, DVs are deemed to very helpful when you trying to identify trends or patterns, i.e. when there is need for analysis. Most often with nano-data, e.g. sales of brand X on Sunday across each hour, there is actually no need of an analysis. Here, the raw numbers are required. I would say that as nano data generally doesn’t require analysis, the usage of DV in such a case is not recommended.”

The following response by a top management participant from FMCG industry also expresses similar viewpoint:

By granular or nano-data, I believe that you are referring to information pertaining to small sets like say sale of SKU x of brand y in territory z for the first week of January. In such a scenario, I don’t think that DVs are really require. I mean, when given such a kind of information, I’d rather worry about the absolute number than proportional representation which is what DVs do. However, if you are suggesting that this information is presented in comparison to sale of SKU x of brand y in territory z across the year, then maybe, there could be a hidden trend. But then again, we are talking about DVs dealing with large sets of data. So, I’m sorry, but I don’t see the advantage of using DVs for granular or nanoo-data.”

The information from the IT services industry also was on the same line although contextually phrased. The following is a response of a business manager participant from the IT services industry:

When you say granular or nano-data, I’m interpreting it as information on a very minute scale. Say, the total number of developers used on a particular day on a particular project. In these cases, there isn’t really a trend that we want to identify. And that is the strength of DVs, isn’t it? Identifying hidden trends. So, when there are not really any hidden trends, then why would one put the energy to use DVs at all? In this case, I’d rather go with the raw information vis-a-vis a pretty visual presentation.”

To generalize, the objections to usage of DV for nano-data were the quantum of information, no real requirement for analysis, higher utility of raw information vis-a-vis DVs and stronger merits of data as compared to visualization.

4.3 DV and Gut-Driven Decision Making

Question 3: DVs move managements away from decisions that are gut-driven to those that are data or evidence-driven. Your opinion on that statement is…

Most of the line managers agreed with the statement in a limited manner. The following response by a line manager from the FMCG company indicates the discord with the statement:

Undoubtedly on the face of it DVs definitely aid in having data or evidence-driven decision making. However, if you consider the fact that most decisions that are required to be taken pertain to people, then your statement goes for toss. Because when it is related to people, data or evidence is not always the best way of taking decisions. For instance, if I had to understand the productivity of a particular department in the store, sure DV can make a fantastic infographic, but the corrective measures would be based not on data, but on gut as you mentioned.”

Some of the other line managers also indicated similar responses. The following response by line manager from the retail industry provides another angle to the discussion:

DV tools can definitely present tremendous insights into the data which has been gathered. However, when you say business decisions, they normally pertain to more than one than one area. For example, If I wanted to remove brand  Z from the top-shelf space of a premier location in a retail store, then I ought to know which other brands can take that spot. I need to understand what are the costs and benefits of such an action. In a management simulation or a school, the problem can be neatly outlined as evaluate the costs and benefits of replacing brand Z with other brands in the store at a super premium location in the retail space. However, most situations in life cannot be framed so neatly. So, while in the former case DV seems to be an elegant and useful way out, in real life it fails at several levels. Now, there might possibly be an insight about other brands that can be replaced into that spot. But the essence of the problem is that it is not articulated at a certain point in time. So only under certain conditions, the first insight is not highly relevant until the second insight. DV can give me both the insights, but it cannot connect the dots between the two insights. I would say that DV and context put together, are perhaps better suited to evidence-driven decision making as compared to only DV.”

Evidently, the above manager indicated that DV can definitely provide insights, but without the right context, such insights might not be actionable. Thus for DVs to provide relevant insights, the context ought to be there for the insight to become meaningful. An additional problem with DVs is that the problem cannot be stated neatly and coherently in real life situations. In other words, often business managers have no idea of what they are solving. Thus, even one insight might not be very useful unless other insights are present. Again, the contextual nature and timeliness of insight presentation is also relevant to decision making.

Another area where line managers felt that DVs and data-driven decision making is not necessary is that of routine decisions or areas where processes and procedures have already been in place for a large amount of time. As an example consider the following response by a business manager from the financial services industry:

DVs do give us evidence-derived insights. However, can such insights always be actionable? For instance, sometimes we already know what the issue might be and we might be looking at the data merely to confirm our hypothesis. Say for instance, sales incentives. We have for long working with various kinds of sales incentives and we know what kind of incentives are working and what kinds are not. If an incentive is working only to a certain extent, then we know how to modulate it for desired effects. In such scenarios, we might use DVs to present us with insights. Our understanding might be altered in a minor way or simply we might get affirmation of our hunches or gut-instincts. Then, we can go ahead and take the pre-devised decision. In these cases, yes, DVs definitely move us towards evidence-driven decision making as compared to gut-driven decision making. However, not always do we have the liberty of time or the luxury of pre-conceived problems. There are several decisions which need to be taken on the spur of the moment. In most of these cases, it is not reasonable to expect a business manager to go in for a thorough data analysis and then take a decision. Here, gut instincts are all that counts. To simplify, for routine operational decisions where we have the benefit of experience, DVs are not necessary good or put it another way, gut-driven decision making is more important than evidence-driven decision making, thus removing the necessity to use DVs.”

To simplify the above response, DVs are not required in several decision making processes which are entrenched in the industry with well-established escalation and decision-making procedures. In such a circumstance, DVs definitely do not add value through evidence-derived insights that lead to decisions. From the other responses, it seemed that DVs have limited scope so far as driving evidence-driven decision making is concerned. The luxury of time that allows for neatly framing a given situation or issue is not always available, thereby reducing the utility of DVs.

The top management on the other hand had a completely different perspective. The following response by a top management participant from the FMCG industry showcases this line of thinking:

So you are suggesting that machines should overtake us now? I’m not completely against the notion as I’ve seen automation increase and replace manpower in this industry. Definitely, DV exemplifies analysis of a given situation. But then, people like us would always be required to overlay our crystallized experience of years when it comes to decision-making. So, we do use DVs which are based on data. But when it comes to taking decisions, especially strategic decisions, it is always a thing of chance. Maybe someday informatics would become so pervasive that the probability of success of taking a particular route is known before hand. Even in that case, I do envision manual over-ride every now and then. And we are nowhere close to such a futuristic situation. So, I’m going to disagree with your statement.”

The following response by a top management participant from the retail industry provides further insight into this line of thinking:

Business decisions often affect several stakeholders. So DV can bring to me information about a particular set of stakeholders. That is definitely evidence-driven. Does that make the decision-making process simpler? Or does it entirely command the decision-making process? In my experience, seldom do such situations arise where evidence-driven insight is sufficient to take a decision. For every decision taken, we need to weigh in the perspectives for a wide range of stakeholders who are bound to be affected by a single decision. For instance, say that I want to increase the purchase a particular brand because DV shows that the sales of this brand are very strong. Further analysis also reveals that sales are happening from select stores. So, in essence my decision boils down to increasing purchase of a particular brand and increasing the inventory of that brand at certain stores. Now, my internal competing brand will suffer because of it. How do I estimate the effects of such a decision on the internal brand? I need to change the distribution a little bit more so as to carry that much extra amount of that particular brand. How does the decision effect my supply chain? Also, do I go with more shelf-space or just more stock? Can I circumvent loss of sale due to lower inventory levels through other means? Here, the DV has given me the insight that I ought to buy more of a particular brand. But by no means has it simplified the decision making process. In several cases, we take decisions based on gut-instincts and no matter how good DVs become, they’ll not be able to replace gut-driven decision making.”

To simplify the above response, data-driven or evidence-driven decision making is not entirely possible as decision making needs to be done with the interests of several stakeholders, several conflicts or trade-offs in mind. Thus, no matter how effect DVs become in ensuring evidence-driven insights, business decision making cannot rely on DVs.

The following response by a top management participant from the IT services industry showcases another opposition to the given statement:

Business decision is not like law and order, and the justice department of states! You simply look at evidence and then you draw your conclusions. It is not as simple as that. DVs today, by far, are being used to analyze historical information. In other words, they are not powered to give us projections or scenarios of the future. When they are powered to do so, they often take into account the gut-instincts of individuals making the projections. What use is business decision making based on historical information? Often, business decision making entails taking decision that will affect the future direction of the department or the company. So, in that case, we ought to make several assumptions as no one has ever seen the future. In other words, we are dealing with a tremendous degree of ambiguity and uncertainty. In such cases, there is no more evidence that is driving the thinking process but gut-instincts or hunches. Most projections everywhere are based on assumptions which are in essence gut-instincts. So, I disagree with your statement that DVs are moving us towards evidence-driven decision making.”

The above respondent clearly indicated the major shortcoming of DVs in regards to business decision making: lack of evidence-driven assumptions. Although, historical data can be used as a proxy for future trends, some experiential overlay is necessary to make them sound realistic, which implies that business decision making is bound to be gut-driven to a large extent.

The following response by a top management participant from the financial services industry also indicates discord, however in very different words and thoughts:

Business intelligence has definitely come a long way since its inception. Today, your enterprise tools are combining information from different departments and giving us very valuable insights into our business. However, this is not happening because of automation. There is natural intelligence that says okay let’s combine this and that, and check what is the insight. So, the decision that is borne from such a process cannot be completely be credited as that derived from evidence. There is gut-instinct somewhere in the process. Similarly, when it comes to final decision-making, one cannot do away with gut-instinct entirely despite very strong evidence-driven insights at one’s disposal.”

The above clearly indicates that though DV has come a long way, it still has a long way to go given the shortcoming in artificial intelligence that powers such tools and processes.

The primary objection to DVs leading to data-driven decision making by top management participants was the necessity to overlay experiential wisdom on the data analyzed and then take decisions based on hunches due to lack of tools that provide probability of success. The other objections raised included dealing with ambiguity and uncertainty while taking decisions which necessarily requires gut-instincts, inherent shortcoming in the artificial intelligence that powers the DV tools and techniques, and applicability of DVs only to historical information that ends up providing merely a proxy for decision making but not all the evidence required for the decision.

4.4 DV and Problem Identification

Question 4: Do you think that DVs ease problem identification and formulation, and data preparation and exploration?

Again, there was an strong agreement with the given statement from all respondents without exception. The following response by a line manager participant in the FMCG industry exemplifies this:

Problem identification is by far the greatest advantage that DV brings to the table. The essence of DV is identification of concealed insights in the data gathered. Once data sets are put in visual format, one can readily see how the things are moving and panning out in practice, which directly leads to problem identification. In many cases, a set of DVs also leads to problem formulation. I strongly agree with your statement.”

The following response by a business manager from retail services industry further substantiates this line of thinking:

Absolutely. DVs make the problem pop out from heaps of data. There is a popular adage: a picture is worth a thousand words and DV personify that adage. You pool in consumer purchases of a particular brand across a particular day across all our stores in comparison to another brand across the same day and across all the same stores, and we might see that one brand did not sell as much as another brand. That is just one example, but there is inherently a problem in there. So, DV shows us that there is a problem. Problem formulation can also be achieved through DVs. For instance, in the above DV, if we had placed discount information on both the brands, we might find that discount was a problem area. Similarly, if more than one criteria were to be placed onto the DV, although it might be a bit difficult to read and interpret, problem identification and formulation becomes very easy.”

Despite a general agreement with the first part of the stated assertion, there was hesitation from some of the participants. The following response by a top management participant from the IT industry shows this line of thinking:

Problem identification is very easy only if the interpretation of the DV is straight-forward. Now, back in the olden days, we could look at three or four bar charts and say that there is a problem area. However, current jazzy DVs negate this effect simply because they are too jazzy. One needs to invest time in first getting friendly with the given DV so as to be able to leverage it to understand DVs. That being said, I’ll agree with your statement because, at my position, I get the problem statement as the title or the footer of the page and then I see the DV. Obviously to me, the problem becomes very obvious simply by looking at the DV then.”

The above response clearly indicates that ease of understanding the DV determines the ease of problem identification based on the DV. The respondent clearly indicates that difficulty to understand DVs negate one of the important benefits that is problem identification. However, this is limited to newer forms of DVs which are not entrenched in psychology of business managers unlike the old forms of simple DVs like pie charts, bar charts etc.

The following response by a top management participant from the financial services industry also indicates dissonance with the first part of the given assertion:

DVs definitely help in problem identification, I will not contest that. But, I definitely do not agree that they are helpful in problem formulation.”

Upon further probing, the participant continued:

A series of line graphs can tell me the trend of my company’s ROI across the years. In it, if I notice that there is a dip in a particular year, then I know that there is a problem there. Now, if I were to look at the trend of my returns or my investment over the same period of time, I’ll know where the problem lies. So, DVs are definitely helpful in problem identification. But I ought to look at returns or investments in juxtaposition to my ROI trend. This is not done by a DV. This is done by human intelligence. No matter how good DVs and business intelligence gets, there needs to be human intelligence in the background that is formulating problems and connecting the right set of data with each other. Only then do DVs deliver in problem identification.”

The above response hints at the fact the DVs cannot be automated to connect all sorts of data that are captured in an organization. Thereby, they cannot aid in problem identification. The participant also insisted that problem formulation is not the realm of DVs but that of humans who are engaging (using) the DV.

With regards to data preparation and exploration, there were some inhibitions from the participants. The following response by a line manager participant from the financial services industry showcases this:

It is not always possible to gather data in the way that these tools require them or how we require them based on the analysis desired. Just that the raw data forms are near infinite, I don’t see how DV tools and techniques can encompass all forms of raw information. And also, data cleansing is a big activity. This is an area where there is tremendous scope for improvement for DV tools. Frankly, I don’t know how and when this barrier shall be broken down, but it definitely is required. This single thing, data preparation is so time consuming and so frustrating that it is limiting the widespread application of DV tools and techniques.”

The above response indicates that data preparation and cleansing is a difficult task as the raw data is captured in numerous forms. This implies that all that data must first be brought to the same basis before putting it through analysis of any kind, and more so in the case of DV where rigor is much higher.

The following is the response of a top management participant from the retail industry:

Data preparation is always a big headache. Different kinds of data are gathered by different departments and this is dynamic, as each manager has a preference for a particular style. Thus, ensuring that the data captured is on the same units and basis as required for a DV analysis is not always so straightforward. To an extent, this process can be automated. However, we’ll need a subjective overlay each and everytime to ensure that the cleansing has happened properly. And I’ve already pointed out the problem formulation bit. I believe data exploration leads to that.”

One of the reasons why raw data is captured in numerous forms is managerial preferences and dynamism in capturing data due to managerial attrition. Each incoming manager prefers to see data in particular formats, and thus the raw data that is captured and the subsequent analysis tends to change with time.

The following is the response of a top management participant from the FMCG industry:

We have machines to capture information in our supply. We have sales force and human entry to capture data at the retail level. We have order systems to capture data at the distributor level. We have automated systems to capture information at the manufacturing level. We have manual systems to capture information within the organization. Evidently, there are numerous means of capturing data. In other words, the rigor with which data is captured and the error in data collection varies across the organization. So, it is easier to conduct DV analysis for data within one system, say manufacturing. However, that is a mediocre use of DV. If one were to analyse data at an holistic organization level, then obvious one needs to invest in thorough data preparation. Like I said, since the rigor changes with data collection systems and processes, the manual input required in data preparation is also a variant. At each system level, it does not make business sense to invest human capital simply to prepare data, for one might never use the data that is available. It makes more sense to prepare only that data that leads to problem formulation, which means that the persons accountable for identifying business problems and developing solutions need only deal with data preparation and exploration.”

The participant from the FMCG industry indicated the large investment in data collection systems and processes. At the same time, it was confessed that despite such a high investment in data collection, a similar investment is not made in data preparation and exploration as it does not make business sense to make such a human capital investment.

A business manager from the IT services industry gave the following response to the data preparation and exploration question:

Our operations span a variety of industries and consequently deal with a diverse set of industries. Our operations are not only culture-dependent, but also time-bound as work happens in different times. When you deal on a such a global scale, data collected simply cannot be on the same terms. Someone in Spain might be willing to provide some kind of data while a business manager in India can only capture data in a certain format. Thus, the scale and diversity of operations leads to data complexity. Given such a complexity, the amount of time that needs to be invested in data preparation is humungous. Simply mind-boggling. Although, we are from the computing industry, I can assure you that it is still a humungous requirement. Forget setting up the appropriate systems, the continuous maintenance of such systems simply makes investment in data preparation and exploration completely a loss making proposition. Also, when you say data exploration, it is not an easy job. The one who needs to drive analysis should be the one who needs to say that they want to look at particular kinds of data. A machine, no matter how smart, still cannot do this job in a satisfactory manner. Don’t get me wrong. Like I said, we are from the computing industry and we understand the capabilities and limitations of artificial intelligence. And I can assure you that data exploration is still not the domain of machines.”

The above response clearly indicates another reason for the humungous diversity of data–scale and diversity of operations. There is an indication that data preparation and exploration systems and teams can be set-up. However, continuous running of the such systems and teams would negate the benefits accrued from such a set-up, as data analysis is only a small part of business operations. This clearly shows that investments in data preparation and exploration cannot be developed as a coherent business case.

To summarize DV and problem identification: most participants felt that DV tools have scope for improvement in terms of data preparation and exploration. Additionally, dedicated resource of data crunching and exploration is not possible in all industries, despite their capital bases. The evidence from the research indicates that this is one of the primary reasons for lack of widespread adoption of DV tools and techniques.

4.5 DV and Profitability

Question 5: In your view, do DV applications simplify the evaluation and monitoring of returns? Question 6: In your opinion, how is usage of DVs associated with the profitability of the company?

With regards to the DV and returns, almost all participants requested for further clarity. It was made known to them that the question sought returns with regards to improvement in business processes after applying DVs to such businesses processes.

The response by a line manager from the retail industry is reproduced in verbatim:

That is an interesting question. One definitely needs to measure returns to understand the benefits of something. I have never thought about it earlier. But let me think aloud. We had began using DV to understand consumer spending patterns on discounts of national brands and our in-store brands, spending patterns across various times of the day, and spending patterns across the stores. Now that I’ve spelled it out, I don’t think we have applied DVs to specific processes. We are merely measuring end results and then going back to study the value chain or process flow. So, I really cannot answer this question.”

The above response presents a very interesting facet of DV. Most visualization are applied to end results followed by a tracking of the entire value flow to isolate the underlying problem. Most of the subsequent analysis might not be engaging visualization. This clearly indicates that visualization is not being applied to entire data processes.

The following is the response of a business manager from financial services industry that indicates another line of thought:

In my opinion, trying to understand returns of is simplified only when the process itself is streamlined and one understand the exact investment and the output. Then such investments and outputs need to be measured objectively. What happens practically is that not all processes are streamlined. And for the processes that are streamlined, in actual operations, there are always deviations. In such a dynamic scenario, measurement of inputs and outputs becomes very difficult. Though, we can use DV today, tomorrow the same DV might not be applicable as additional data inputs have been understood in the process. This is almost always the case as human input is a critical part of the process and measuring this objectively is always a challenge. So, in the initial stages we assume it to be not significant. However, as time progresses, someone says that measuring human input in this way makes sense, so let use it in the evaluation of returns. So then, the old DV needs to be updated. You see, there are too many elements that are changing in this picture. So, simplified is not a word that I would want to use.”

The above response clearly indicates that processes themselves need to be streamlined for a DV to work. It is reported that though most processes are streamlined, in actual practice there are bound to be deviations. Such deviations cause errors in measurements of inputs and outputs, and sometimes also lead to addition or deletion of data sets. Thus, a DV that has been used once might not be applicable at a later stage.

The following is a response by a top management participant from the IT services industry:

Simplification happens when things are well understood. Definitely DV can aid in evaluation and monitoring of returns of business processes. However, I would contest the word simplify here. You see, new kinds of DVs are not standardized. Unless we were using pie charts and simple bar charts all the time, we are not simplifying anything at all. I’ve already given you the example of a project manager who used something we were not aware of at all. In that case, it was really difficult to understand the kind of returns that we were getting on the investments in our business processes.”

The above argument states that since the field of DV is itself in a state of flux with new visual representations being introduced constantly, there is no simplification in the usage. On the contrary, the introduction of newer kinds of visualizations coupled with the fact that the users themselves are not thoroughly trained in handling such visualizations, is complicating the matter further.

The above responses clearly point out to the fact that since DV is not standardized, evaluation and monitoring of returns is not being simplified currently.

The following response by a top management participant from the FMCG industry indicates that line of thinking:

If processes are streamlined and the method of analysis frozen, then one can definitely say that DV aids the evaluation and monitoring if returns.”

Upon further probing for elaboration, the participant continued:

Say, I wanted to evaluate frauds that are occurring in our credit card division. Through a series of analysis, let’s say that we determined that a particular physical component on the cards is making the card vulnerable to thefts. So we decide to invest in a superior material and improved tracking. All through we have been using the same DVs to understand the investments, total fraud cases and overall returns. In time, we notice that there were infact positive returns in terms of reduced number of frauds and improved customer satisfaction. Now, in this case the processes are streamlined, as in we know what is going in, how it is flowing in the value stream and what is coming out. Additionally, we are using the same set of DV tools and analysis all throughout. In that case, the DVs generated are easy to interpret.”

Most participants were of the opinion that DV does aid in evaluation and monitoring of returns if the processes are streamlined. The key point of contention was that since DV is itself an evolving field with new tools and techniques coming into place, the analysis and presentation of information is in a state of flux. Given such a dynamic scenario, consolidation of specific methods of analysis becomes difficult. This consequently leads to evaluation and monitoring of returns using different techniques all the time. Thus, the participants were cautious to opine that DVs simplify evaluation and monitoring of returns. However, when probed further, participants accepted that once a particular kind of analysis is deemed appropriate and frozen, DV greatly aids in understanding their returns.

In terms of association of DV with profitability, most participants were of the opinion that DV definitely contributed to time and associated cost savings. Some participants cited that improved decision making due to DV does aid in improving top-line as well. The following response by a line manager in IT services industry presents this point of view:

I have seen DV contribute to improvement in top-line. For instance, last year, we analyzed our sales patterns with our top quartile clients. This data was analyzed used different kinds of DVs. From that analysis we learnt that certain kinds of customers were ordering only certain kinds of services, time and again. When we dug deeper into that analysis again with DVs, we realized that our track-record in those services has been brilliant. Although, there was a tacit understanding that the service line in discussion was a core competency for us, there was really no data-driven proof for it. However, post that analysis, we developed off-shoots for that service line. Again from the DV, we understood which clients should be approached first. And the results were very encouraging.”

The above response indicates that DV has aided in top-line contributions. It is this improvement that can be considered as a contribution to profitability as improvements in revenues is bound to improve profits.

The following response by a top management participant from the financial services industry indicates a different point:

We have repeatedly used DVs to understand the cost structures of our various strategic business units (SBUs). Firstly, the DVs aid us in understanding what is happening in the costs of each SBU and secondly, they provide a means of comparison across the SBUs. Such an analysis obviously leads to thoughts like “If SBU A is using similar processes as that of SBU B, then why does SBU A have a higher cost structure”, or “If SBU A has used this much in the past, why is it using so much now”. Such an evaluation obviously leads to further analysis from which we can isolate the actual problems. Once we understand the inherent problems in the processes, we obviously correct it some way or another. This leads directly to cost savings. And yes, that leads to improved profitability. On a one-on-one basis, I would say that the contribution is too meagre. On an aggregate basis both on time and across SBUs, though the contribution is low, it is still considerable. In such times of economic crunch and financial prudence, even such contributions are credit-worthy. Also, I would envision that DVs would have a stronger role in profitability the more they proliferate across business processes.”

The above response clearly indicates that DVs have aided in cost savings that directly lead to improvement in profitability. Similar responses were also obtained from some other participants and most of them reported that the contribution is still very low in this regard. However, given the requirement of thorough cost control, even such contributions are greatly appreciated.

The following response by business manager from the retail industry indicates a different point of view:

I would say that the biggest advantage of DVs is that of time savings. Despite the initial investment that needs to be made in understanding a particular kind of DV and data preparation, I still think that it is a good investment. Once something is standardized, it becomes a matter of plug and play. For instance, I began using a heat map to understand the sales of a particular brand across all our retail outlets in the nation. The first two or three times, we did it, it did take a lot of time. Now, it is simply a plug and play thing–clean the data and feed it, and voila, we have the analysis out. We’ve been using that analysis for the last one year. Once we have that time saving, we can invest it in other activities. I’m not saying that increased time in other activities will impact the profitability directly. However, there has got to be some kind of butterfly effect in place, right?”

Time saving was cited as one of the biggest advantages of using DVs. Although, some participants were adamantly against the increased time investment required in understanding and data preparation required for the DV, almost all participants agreed that once the DV is standardized for a particular kind of analysis, it definitely results in time savings in subsequent usages. Most participants were of the opinion that the time saved due to DVs could be invested in other productive activities that improved the overall productivity, and subsequently profitability. However, not all participants could point out the link between productivity improvements and profitability improvements.

The scenario in the FMCG industry was slightly different. The following is a response of a top management participant from the FMCG industry:

Undoubtedly, once a specific kind of analysis is frozen, it definitely aids in improved cost and time savings. However, the kind of data sets that we see in the FMCG industry, it is not always possible to create simply a plug-and-play solution. We need thorough quality control and quality assurance before we can even use the data in the analysis. It is this specific aspect of the FMCG industry that I think makes DVs less effective.”

The researcher presented the viewpoint that in the retail industry as well the data sets are very large, but in that industry, DVs are seen to beneficial to profitability through time savings. To this, the above participant continued:

Undoubtedly, the data sets even in the case of the retail industry are humungous. But you also need to understand that a majority of that industry is automated. Everywhere the data is being captured through computers. In the case of FMCG, we are dealing with too many people. Yes, the people are using computers too at almost every point. But the very fact that there is so much reliance on human entry necessitates that we perform quality control and assurance each and every single time. This would be the primary difference in my opinion.”

In the FMCG industry, humungous data sets and contrarily lack of data were cited as major deterrents.

4.6 DV and Overall Decision Making

Question 7: Decisions made on the basis of DV are significantly more precise, informed and inclusive. Your thoughts.

All participants agreed with the given statement with some inhibitions. Participants felt that DVs generally combine varying sets of data to generate one visual representation, they become inclusive by default. At the same time, since underlying data is being used to create

4.2 DV and Ease of Analysis

Question 2: In your opinion, does usage of DV applications significantly ease the analysis and appreciation of all forms of data, including granular or nano-data?

There was an overwhelming affirmation from all the participants surveyed so far as large data sets was concerned. The following response by a business manager participant from the FMCG industry illustrates this line of thinking:

Absolutely. There is no doubt about it at all. By and large, I do not get to use much of this information, but I ought to prepare it for senior management perusal. And, personally, I’ve seen it work wonders with large sets of data. There are commercial programs that we use and we need to feed data specifically into it and it pops out pretty visual representations of trends. That obviously eases analyses.”

The following response from a top management participant in the financial service industry provides further insight:

For us top management individuals who have a lot of things to take care of, we often ought to deal with large sets of information. For instance, on a quarterly basis, I need to understand all the frauds that have taken place in a bank. Now there are numerous kinds of frauds like ATM frauds, fake accounts, personal identity thefts, internet-based fraud, credit card frauds etc. Within these kinds of frauds, across all our operations, we end up with significant number of issues. Now that is only one kind of issue that I ought to deal with. In my position, the biggest challenge is dealing with individuals, ensuring that they are on track, they are motivated to achieve the outlined goals. So, operational things take a back burner. In such an instance, data visualization comes as a big boon. We had several individuals spending a large amount of time working on data just to make sure that it is easy to understand and that the analysis is in appropriate. With DV tools, analysis becomes tremendously easy. Ofcourse, we still need to spend time on the raw data itself, which for its own reasons is collected in diverse ways, and thus requires to be cleaned thoroughly. So, in a nutshell, for large sets of data, DV tools are definitely greatly helpful.”

The following response by a business manager from the retail industry indicates that the scenario is pretty similar as that in FMCG industry:

As a marketing manager, my primary focus is on ensuring that the brand is performing upto the budgets. Although, I tend to work more on marketing material and brand management team, I ought to look at sales data on a regular basis. I ought to understand which distributors are picking up our brands, which areas are performing properly, which retailers are taking our brand and so on. And then at the retail level, there is tremendous amount of consumer information, especially with regards to SKUs. So, we use DV tools mostly on sales information and consumer-level SKU purchase information. Only now are we beginning to use more recent forms of DVs like key text graphing, wherein we are mapping our market research information to provide us ready insights. Dealing with such large sets of data, I can tell you that DV has been greatly helpful. Additionally, DVs also help us present a coherent investment scenario to top management. Often top management does not have time to dwell upon the data and it might become very difficult to walk them through all the insights. However, DV aids us in mapping such information very easy and also in aesthetic formats. In essence, I can present entire sales data from our global sales on one single slide!”

The following is a response by business manager from the financial services industry which connects well with the response of the top management participant from the same industry:

As a credit card departmental manager, I ought to look at sales figures, credit defaults, collections, new customer acquisition, customer attrition, customer retention, extension of the credit card to more and more vendors, customer loyalty and points and usage and so on. As you can see, on a national scale, I have to deal with tremendous amounts of information. DV tools aids in a great manner. I can look at customer acquisition, customer retention, customer attrition, and customer loyalty all in one-go with the help of DV tools and techniques. Most importantly, when presenting this information to the top management in periodic meetings, the discussion becomes more fruitful as the insight is readily visible and we needn’t run the top management through our entire analysis. Prior to using DV tools, we used to sweat out a lot to make coherent graphs and slides, and quite literally we were at a loss always putting in unholy hours. Thus DV tools have not only improved the ease of analysis, but have also reduced our workload dramatically.”

The easing of burden was used as a prompt in interviews that followed the above interviews. Interestingly, this prompt provided further insights into the ease of analysis enabled by DV tools and techniques. To put things in context, the following is the response of a top management participant from the IT services industry:

Undoubtedly, DV tools are greatly beneficial when we look at large data sets. For example, on an yearly basis we look at the total number of projects completed, projects in pipeline, work-in-progress projects and so on. Since we are in a project management environment, we are pretty adamant on data collection as that helps us understand whether projects will get finished on time and in stipulated budget. And here’s where the magic of DV lies. I can take one project and put all information pertinent to it on one page and that will give me all the information that is required. So, it is in this context that I want to comment on your easing of burden point. You see, analysis is eased only when we are familiar with a certain kind of representation. For instance, a simple pie chart. That’s something we have seen for ages and we know pretty much what a pie chart involves and indicates. So, our focus is no more on the representation, but on the actual information. However, if you pop up some very fancy new thing, then things go for a toss. I remember, recently this new project manager brought an analysis with him in the format that he used in his previous company. Believe me, the first thirty minutes of the meeting, we spent on understanding what the whole thing was and only then did we begin looking at the information. Now ofcourse, if he presents information in a similar format, probably, we’ll take no more than ten minutes to orient our minds. But if he goes back to the company format, we’ll get uncomfortable again. You see, this trade-off needs to be kept in mind always.”

The following response by a business manager from the financial services also shows a similar viewpoint:

I’ve been using DV tools for some time now. I also tend to do a little bit of R&D. It is much more beneficial to be aware and knowledge of what needs to be done before your subordinates pick it up. Very recently, I saw this heat map which caught my fancy. Although I didn’t have all the information to create a pitch-perfect heat map, I went ahead and produced one to show the total number of mobile phone users in the country. Now heat maps are not something that are difficult to understand. We see them all the time in newspapers. At first, when I popped it up in the boardroom, there was appreciation. The analysis and insights were readily visible. However, our top management got into the nitty-gritties of it. I believe this is called xenophobia or some related term. Standing there in the boardroom, I thought to myself-damn, why didn’t I expect such a situation. It is much easier to present information in ways that one is used to. Nevertheless, only by presenting new formats and getting people acquainted to them, is the way forward.”

However, some participants observed that since there is no common base for most recent DVs, investment of time required to understand the visualized data negates the benefits of the visualization. In terms of old forms of DVs (e.g. time-series, spider-charts etc.), all participants praised DVs as being the easiest method of analysis.

With regards to granular or nano-data, the following is the response of a top-management participant from the financial services industry:

I would say that DV is definitely very useful even for granular data. For instance, if I wanted to understand the ATM transactions of a particular customer, a visual representation would be very helpful to see the trends and patterns. But with respect to nano-data, I don’t DV would be significantly helpful. For instance, in the case of a particular customer, if I wanted to check their cheque book requests across a time span, visualization of that data is as helpful as the raw data itself, simply because the quantum of information is very little. In fact, I would say that DV for nano-data would be more of a nuisance that actual use. Although the same can be said about granular data, I believe that by and large, for granular data DV definitely are very helpful.”

Top management participants and almost all of the line managers presented similar perspectives of applicability of DV to granular and nano-data. The following response from a line manager from the retail industry exemplifies this:

Sometimes, you just to look at raw numbers. See, DVs are deemed to very helpful when you trying to identify trends or patterns, i.e. when there is need for analysis. Most often with nano-data, e.g. sales of brand X on Sunday across each hour, there is actually no need of an analysis. Here, the raw numbers are required. I would say that as nano data generally doesn’t require analysis, the usage of DV in such a case is not recommended.”

The following response by a top management participant from FMCG industry also expresses similar viewpoint:

By granular or nano-data, I believe that you are referring to information pertaining to small sets like say sale of SKU x of brand y in territory z for the first week of January. In such a scenario, I don’t think that DVs are really require. I mean, when given such a kind of information, I’d rather worry about the absolute number than proportional representation which is what DVs do. However, if you are suggesting that this information is presented in comparison to sale of SKU x of brand y in territory z across the year, then maybe, there could be a hidden trend. But then again, we are talking about DVs dealing with large sets of data. So, I’m sorry, but I don’t see the advantage of using DVs for granular or nanoo-data.”

The information from the IT services industry also was on the same line although contextually phrased. The following is a response of a business manager participant from the IT services industry:

When you say granular or nano-data, I’m interpreting it as information on a very minute scale. Say, the total number of developers used on a particular day on a particular project. In these cases, there isn’t really a trend that we want to identify. And that is the strength of DVs, isn’t it? Identifying hidden trends. So, when there are not really any hidden trends, then why would one put the energy to use DVs at all? In this case, I’d rather go with the raw information vis-a-vis a pretty visual presentation.”

To generalize, the objections to usage of DV for nano-data were the quantum of information, no real requirement for analysis, higher utility of raw information vis-a-vis DVs and stronger merits of data as compared to visualization.

4.3 DV and Gut-Driven Decision Making

Question 3: DVs move managements away from decisions that are gut-driven to those that are data or evidence-driven. Your opinion on that statement is…

Most of the line managers agreed with the statement in a limited manner. The following response by a line manager from the FMCG company indicates the discord with the statement:

Undoubtedly on the face of it DVs definitely aid in having data or evidence-driven decision making. However, if you consider the fact that most decisions that are required to be taken pertain to people, then your statement goes for toss. Because when it is related to people, data or evidence is not always the best way of taking decisions. For instance, if I had to understand the productivity of a particular department in the store, sure DV can make a fantastic infographic, but the corrective measures would be based not on data, but on gut as you mentioned.”

Some of the other line managers also indicated similar responses. The following response by line manager from the retail industry provides another angle to the discussion:

DV tools can definitely present tremendous insights into the data which has been gathered. However, when you say business decisions, they normally pertain to more than one than one area. For example, If I wanted to remove brand  Z from the top-shelf space of a premier location in a retail store, then I ought to know which other brands can take that spot. I need to understand what are the costs and benefits of such an action. In a management simulation or a school, the problem can be neatly outlined as evaluate the costs and benefits of replacing brand Z with other brands in the store at a super premium location in the retail space. However, most situations in life cannot be framed so neatly. So, while in the former case DV seems to be an elegant and useful way out, in real life it fails at several levels. Now, there might possibly be an insight about other brands that can be replaced into that spot. But the essence of the problem is that it is not articulated at a certain point in time. So only under certain conditions, the first insight is not highly relevant until the second insight. DV can give me both the insights, but it cannot connect the dots between the two insights. I would say that DV and context put together, are perhaps better suited to evidence-driven decision making as compared to only DV.”

Evidently, the above manager indicated that DV can definitely provide insights, but without the right context, such insights might not be actionable. Thus for DVs to provide relevant insights, the context ought to be there for the insight to become meaningful. An additional problem with DVs is that the problem cannot be stated neatly and coherently in real life situations. In other words, often business managers have no idea of what they are solving. Thus, even one insight might not be very useful unless other insights are present. Again, the contextual nature and timeliness of insight presentation is also relevant to decision making.

Another area where line managers felt that DVs and data-driven decision making is not necessary is that of routine decisions or areas where processes and procedures have already been in place for a large amount of time. As an example consider the following response by a business manager from the financial services industry:

DVs do give us evidence-derived insights. However, can such insights always be actionable? For instance, sometimes we already know what the issue might be and we might be looking at the data merely to confirm our hypothesis. Say for instance, sales incentives. We have for long working with various kinds of sales incentives and we know what kind of incentives are working and what kinds are not. If an incentive is working only to a certain extent, then we know how to modulate it for desired effects. In such scenarios, we might use DVs to present us with insights. Our understanding might be altered in a minor way or simply we might get affirmation of our hunches or gut-instincts. Then, we can go ahead and take the pre-devised decision. In these cases, yes, DVs definitely move us towards evidence-driven decision making as compared to gut-driven decision making. However, not always do we have the liberty of time or the luxury of pre-conceived problems. There are several decisions which need to be taken on the spur of the moment. In most of these cases, it is not reasonable to expect a business manager to go in for a thorough data analysis and then take a decision. Here, gut instincts are all that counts. To simplify, for routine operational decisions where we have the benefit of experience, DVs are not necessary good or put it another way, gut-driven decision making is more important than evidence-driven decision making, thus removing the necessity to use DVs.”

To simplify the above response, DVs are not required in several decision making processes which are entrenched in the industry with well-established escalation and decision-making procedures. In such a circumstance, DVs definitely do not add value through evidence-derived insights that lead to decisions. From the other responses, it seemed that DVs have limited scope so far as driving evidence-driven decision making is concerned. The luxury of time that allows for neatly framing a given situation or issue is not always available, thereby reducing the utility of DVs.

The top management on the other hand had a completely different perspective. The following response by a top management participant from the FMCG industry showcases this line of thinking:

So you are suggesting that machines should overtake us now? I’m not completely against the notion as I’ve seen automation increase and replace manpower in this industry. Definitely, DV exemplifies analysis of a given situation. But then, people like us would always be required to overlay our crystallized experience of years when it comes to decision-making. So, we do use DVs which are based on data. But when it comes to taking decisions, especially strategic decisions, it is always a thing of chance. Maybe someday informatics would become so pervasive that the probability of success of taking a particular route is known before hand. Even in that case, I do envision manual over-ride every now and then. And we are nowhere close to such a futuristic situation. So, I’m going to disagree with your statement.”

The following response by a top management participant from the retail industry provides further insight into this line of thinking:

Business decisions often affect several stakeholders. So DV can bring to me information about a particular set of stakeholders. That is definitely evidence-driven. Does that make the decision-making process simpler? Or does it entirely command the decision-making process? In my experience, seldom do such situations arise where evidence-driven insight is sufficient to take a decision. For every decision taken, we need to weigh in the perspectives for a wide range of stakeholders who are bound to be affected by a single decision. For instance, say that I want to increase the purchase a particular brand because DV shows that the sales of this brand are very strong. Further analysis also reveals that sales are happening from select stores. So, in essence my decision boils down to increasing purchase of a particular brand and increasing the inventory of that brand at certain stores. Now, my internal competing brand will suffer because of it. How do I estimate the effects of such a decision on the internal brand? I need to change the distribution a little bit more so as to carry that much extra amount of that particular brand. How does the decision effect my supply chain? Also, do I go with more shelf-space or just more stock? Can I circumvent loss of sale due to lower inventory levels through other means? Here, the DV has given me the insight that I ought to buy more of a particular brand. But by no means has it simplified the decision making process. In several cases, we take decisions based on gut-instincts and no matter how good DVs become, they’ll not be able to replace gut-driven decision making.”

To simplify the above response, data-driven or evidence-driven decision making is not entirely possible as decision making needs to be done with the interests of several stakeholders, several conflicts or trade-offs in mind. Thus, no matter how effect DVs become in ensuring evidence-driven insights, business decision making cannot rely on DVs.

The following response by a top management participant from the IT services industry showcases another opposition to the given statement:

Business decision is not like law and order, and the justice department of states! You simply look at evidence and then you draw your conclusions. It is not as simple as that. DVs today, by far, are being used to analyze historical information. In other words, they are not powered to give us projections or scenarios of the future. When they are powered to do so, they often take into account the gut-instincts of individuals making the projections. What use is business decision making based on historical information? Often, business decision making entails taking decision that will affect the future direction of the department or the company. So, in that case, we ought to make several assumptions as no one has ever seen the future. In other words, we are dealing with a tremendous degree of ambiguity and uncertainty. In such cases, there is no more evidence that is driving the thinking process but gut-instincts or hunches. Most projections everywhere are based on assumptions which are in essence gut-instincts. So, I disagree with your statement that DVs are moving us towards evidence-driven decision making.”

The above respondent clearly indicated the major shortcoming of DVs in regards to business decision making: lack of evidence-driven assumptions. Although, historical data can be used as a proxy for future trends, some experiential overlay is necessary to make them sound realistic, which implies that business decision making is bound to be gut-driven to a large extent.

The following response by a top management participant from the financial services industry also indicates discord, however in very different words and thoughts:

Business intelligence has definitely come a long way since its inception. Today, your enterprise tools are combining information from different departments and giving us very valuable insights into our business. However, this is not happening because of automation. There is natural intelligence that says okay let’s combine this and that, and check what is the insight. So, the decision that is borne from such a process cannot be completely be credited as that derived from evidence. There is gut-instinct somewhere in the process. Similarly, when it comes to final decision-making, one cannot do away with gut-instinct entirely despite very strong evidence-driven insights at one’s disposal.”

The above clearly indicates that though DV has come a long way, it still has a long way to go given the shortcoming in artificial intelligence that powers such tools and processes.

The primary objection to DVs leading to data-driven decision making by top management participants was the necessity to overlay experiential wisdom on the data analyzed and then take decisions based on hunches due to lack of tools that provide probability of success. The other objections raised included dealing with ambiguity and uncertainty while taking decisions which necessarily requires gut-instincts, inherent shortcoming in the artificial intelligence that powers the DV tools and techniques, and applicability of DVs only to historical information that ends up providing merely a proxy for decision making but not all the evidence required for the decision.

4.4 DV and Problem Identification

Question 4: Do you think that DVs ease problem identification and formulation, and data preparation and exploration?

Again, there was an strong agreement with the given statement from all respondents without exception. The following response by a line manager participant in the FMCG industry exemplifies this:

Problem identification is by far the greatest advantage that DV brings to the table. The essence of DV is identification of concealed insights in the data gathered. Once data sets are put in visual format, one can readily see how the things are moving and panning out in practice, which directly leads to problem identification. In many cases, a set of DVs also leads to problem formulation. I strongly agree with your statement.”

The following response by a business manager from retail services industry further substantiates this line of thinking:

Absolutely. DVs make the problem pop out from heaps of data. There is a popular adage: a picture is worth a thousand words and DV personify that adage. You pool in consumer purchases of a particular brand across a particular day across all our stores in comparison to another brand across the same day and across all the same stores, and we might see that one brand did not sell as much as another brand. That is just one example, but there is inherently a problem in there. So, DV shows us that there is a problem. Problem formulation can also be achieved through DVs. For instance, in the above DV, if we had placed discount information on both the brands, we might find that discount was a problem area. Similarly, if more than one criteria were to be placed onto the DV, although it might be a bit difficult to read and interpret, problem identification and formulation becomes very easy.”

Despite a general agreement with the first part of the stated assertion, there was hesitation from some of the participants. The following response by a top management participant from the IT industry shows this line of thinking:

Problem identification is very easy only if the interpretation of the DV is straight-forward. Now, back in the olden days, we could look at three or four bar charts and say that there is a problem area. However, current jazzy DVs negate this effect simply because they are too jazzy. One needs to invest time in first getting friendly with the given DV so as to be able to leverage it to understand DVs. That being said, I’ll agree with your statement because, at my position, I get the problem statement as the title or the footer of the page and then I see the DV. Obviously to me, the problem becomes very obvious simply by looking at the DV then.”

The above response clearly indicates that ease of understanding the DV determines the ease of problem identification based on the DV. The respondent clearly indicates that difficulty to understand DVs negate one of the important benefits that is problem identification. However, this is limited to newer forms of DVs which are not entrenched in psychology of business managers unlike the old forms of simple DVs like pie charts, bar charts etc.

The following response by a top management participant from the financial services industry also indicates dissonance with the first part of the given assertion:

DVs definitely help in problem identification, I will not contest that. But, I definitely do not agree that they are helpful in problem formulation.”

Upon further probing, the participant continued:

A series of line graphs can tell me the trend of my company’s ROI across the years. In it, if I notice that there is a dip in a particular year, then I know that there is a problem there. Now, if I were to look at the trend of my returns or my investment over the same period of time, I’ll know where the problem lies. So, DVs are definitely helpful in problem identification. But I ought to look at returns or investments in juxtaposition to my ROI trend. This is not done by a DV. This is done by human intelligence. No matter how good DVs and business intelligence gets, there needs to be human intelligence in the background that is formulating problems and connecting the right set of data with each other. Only then do DVs deliver in problem identification.”

The above response hints at the fact the DVs cannot be automated to connect all sorts of data that are captured in an organization. Thereby, they cannot aid in problem identification. The participant also insisted that problem formulation is not the realm of DVs but that of humans who are engaging (using) the DV.

With regards to data preparation and exploration, there were some inhibitions from the participants. The following response by a line manager participant from the financial services industry showcases this:

It is not always possible to gather data in the way that these tools require them or how we require them based on the analysis desired. Just that the raw data forms are near infinite, I don’t see how DV tools and techniques can encompass all forms of raw information. And also, data cleansing is a big activity. This is an area where there is tremendous scope for improvement for DV tools. Frankly, I don’t know how and when this barrier shall be broken down, but it definitely is required. This single thing, data preparation is so time consuming and so frustrating that it is limiting the widespread application of DV tools and techniques.”

The above response indicates that data preparation and cleansing is a difficult task as the raw data is captured in numerous forms. This implies that all that data must first be brought to the same basis before putting it through analysis of any kind, and more so in the case of DV where rigor is much higher.

The following is the response of a top management participant from the retail industry:

Data preparation is always a big headache. Different kinds of data are gathered by different departments and this is dynamic, as each manager has a preference for a particular style. Thus, ensuring that the data captured is on the same units and basis as required for a DV analysis is not always so straightforward. To an extent, this process can be automated. However, we’ll need a subjective overlay each and everytime to ensure that the cleansing has happened properly. And I’ve already pointed out the problem formulation bit. I believe data exploration leads to that.”

One of the reasons why raw data is captured in numerous forms is managerial preferences and dynamism in capturing data due to managerial attrition. Each incoming manager prefers to see data in particular formats, and thus the raw data that is captured and the subsequent analysis tends to change with time.

The following is the response of a top management participant from the FMCG industry:

We have machines to capture information in our supply. We have sales force and human entry to capture data at the retail level. We have order systems to capture data at the distributor level. We have automated systems to capture information at the manufacturing level. We have manual systems to capture information within the organization. Evidently, there are numerous means of capturing data. In other words, the rigor with which data is captured and the error in data collection varies across the organization. So, it is easier to conduct DV analysis for data within one system, say manufacturing. However, that is a mediocre use of DV. If one were to analyse data at an holistic organization level, then obvious one needs to invest in thorough data preparation. Like I said, since the rigor changes with data collection systems and processes, the manual input required in data preparation is also a variant. At each system level, it does not make business sense to invest human capital simply to prepare data, for one might never use the data that is available. It makes more sense to prepare only that data that leads to problem formulation, which means that the persons accountable for identifying business problems and developing solutions need only deal with data preparation and exploration.”

The participant from the FMCG industry indicated the large investment in data collection systems and processes. At the same time, it was confessed that despite such a high investment in data collection, a similar investment is not made in data preparation and exploration as it does not make business sense to make such a human capital investment.

A business manager from the IT services industry gave the following response to the data preparation and exploration question:

Our operations span a variety of industries and consequently deal with a diverse set of industries. Our operations are not only culture-dependent, but also time-bound as work happens in different times. When you deal on a such a global scale, data collected simply cannot be on the same terms. Someone in Spain might be willing to provide some kind of data while a business manager in India can only capture data in a certain format. Thus, the scale and diversity of operations leads to data complexity. Given such a complexity, the amount of time that needs to be invested in data preparation is humungous. Simply mind-boggling. Although, we are from the computing industry, I can assure you that it is still a humungous requirement. Forget setting up the appropriate systems, the continuous maintenance of such systems simply makes investment in data preparation and exploration completely a loss making proposition. Also, when you say data exploration, it is not an easy job. The one who needs to drive analysis should be the one who needs to say that they want to look at particular kinds of data. A machine, no matter how smart, still cannot do this job in a satisfactory manner. Don’t get me wrong. Like I said, we are from the computing industry and we understand the capabilities and limitations of artificial intelligence. And I can assure you that data exploration is still not the domain of machines.”

The above response clearly indicates another reason for the humungous diversity of data–scale and diversity of operations. There is an indication that data preparation and exploration systems and teams can be set-up. However, continuous running of the such systems and teams would negate the benefits accrued from such a set-up, as data analysis is only a small part of business operations. This clearly shows that investments in data preparation and exploration cannot be developed as a coherent business case.

To summarize DV and problem identification: most participants felt that DV tools have scope for improvement in terms of data preparation and exploration. Additionally, dedicated resource of data crunching and exploration is not possible in all industries, despite their capital bases. The evidence from the research indicates that this is one of the primary reasons for lack of widespread adoption of DV tools and techniques.

4.5 DV and Profitability

Question 5: In your view, do DV applications simplify the evaluation and monitoring of returns? Question 6: In your opinion, how is usage of DVs associated with the profitability of the company?

With regards to the DV and returns, almost all participants requested for further clarity. It was made known to them that the question sought returns with regards to improvement in business processes after applying DVs to such businesses processes.

The response by a line manager from the retail industry is reproduced in verbatim:

That is an interesting question. One definitely needs to measure returns to understand the benefits of something. I have never thought about it earlier. But let me think aloud. We had began using DV to understand consumer spending patterns on discounts of national brands and our in-store brands, spending patterns across various times of the day, and spending patterns across the stores. Now that I’ve spelled it out, I don’t think we have applied DVs to specific processes. We are merely measuring end results and then going back to study the value chain or process flow. So, I really cannot answer this question.”

The above response presents a very interesting facet of DV. Most visualization are applied to end results followed by a tracking of the entire value flow to isolate the underlying problem. Most of the subsequent analysis might not be engaging visualization. This clearly indicates that visualization is not being applied to entire data processes.

The following is the response of a business manager from financial services industry that indicates another line of thought:

In my opinion, trying to understand returns of is simplified only when the process itself is streamlined and one understand the exact investment and the output. Then such investments and outputs need to be measured objectively. What happens practically is that not all processes are streamlined. And for the processes that are streamlined, in actual operations, there are always deviations. In such a dynamic scenario, measurement of inputs and outputs becomes very difficult. Though, we can use DV today, tomorrow the same DV might not be applicable as additional data inputs have been understood in the process. This is almost always the case as human input is a critical part of the process and measuring this objectively is always a challenge. So, in the initial stages we assume it to be not significant. However, as time progresses, someone says that measuring human input in this way makes sense, so let use it in the evaluation of returns. So then, the old DV needs to be updated. You see, there are too many elements that are changing in this picture. So, simplified is not a word that I would want to use.”

The above response clearly indicates that processes themselves need to be streamlined for a DV to work. It is reported that though most processes are streamlined, in actual practice there are bound to be deviations. Such deviations cause errors in measurements of inputs and outputs, and sometimes also lead to addition or deletion of data sets. Thus, a DV that has been used once might not be applicable at a later stage.

The following is a response by a top management participant from the IT services industry:

Simplification happens when things are well understood. Definitely DV can aid in evaluation and monitoring of returns of business processes. However, I would contest the word simplify here. You see, new kinds of DVs are not standardized. Unless we were using pie charts and simple bar charts all the time, we are not simplifying anything at all. I’ve already given you the example of a project manager who used something we were not aware of at all. In that case, it was really difficult to understand the kind of returns that we were getting on the investments in our business processes.”

The above argument states that since the field of DV is itself in a state of flux with new visual representations being introduced constantly, there is no simplification in the usage. On the contrary, the introduction of newer kinds of visualizations coupled with the fact that the users themselves are not thoroughly trained in handling such visualizations, is complicating the matter further.

The above responses clearly point out to the fact that since DV is not standardized, evaluation and monitoring of returns is not being simplified currently.

The following response by a top management participant from the FMCG industry indicates that line of thinking:

If processes are streamlined and the method of analysis frozen, then one can definitely say that DV aids the evaluation and monitoring if returns.”

Upon further probing for elaboration, the participant continued:

Say, I wanted to evaluate frauds that are occurring in our credit card division. Through a series of analysis, let’s say that we determined that a particular physical component on the cards is making the card vulnerable to thefts. So we decide to invest in a superior material and improved tracking. All through we have been using the same DVs to understand the investments, total fraud cases and overall returns. In time, we notice that there were infact positive returns in terms of reduced number of frauds and improved customer satisfaction. Now, in this case the processes are streamlined, as in we know what is going in, how it is flowing in the value stream and what is coming out. Additionally, we are using the same set of DV tools and analysis all throughout. In that case, the DVs generated are easy to interpret.”

Most participants were of the opinion that DV does aid in evaluation and monitoring of returns if the processes are streamlined. The key point of contention was that since DV is itself an evolving field with new tools and techniques coming into place, the analysis and presentation of information is in a state of flux. Given such a dynamic scenario, consolidation of specific methods of analysis becomes difficult. This consequently leads to evaluation and monitoring of returns using different techniques all the time. Thus, the participants were cautious to opine that DVs simplify evaluation and monitoring of returns. However, when probed further, participants accepted that once a particular kind of analysis is deemed appropriate and frozen, DV greatly aids in understanding their returns.

In terms of association of DV with profitability, most participants were of the opinion that DV definitely contributed to time and associated cost savings. Some participants cited that improved decision making due to DV does aid in improving top-line as well. The following response by a line manager in IT services industry presents this point of view:

I have seen DV contribute to improvement in top-line. For instance, last year, we analyzed our sales patterns with our top quartile clients. This data was analyzed used different kinds of DVs. From that analysis we learnt that certain kinds of customers were ordering only certain kinds of services, time and again. When we dug deeper into that analysis again with DVs, we realized that our track-record in those services has been brilliant. Although, there was a tacit understanding that the service line in discussion was a core competency for us, there was really no data-driven proof for it. However, post that analysis, we developed off-shoots for that service line. Again from the DV, we understood which clients should be approached first. And the results were very encouraging.”

The above response indicates that DV has aided in top-line contributions. It is this improvement that can be considered as a contribution to profitability as improvements in revenues is bound to improve profits.

The following response by a top management participant from the financial services industry indicates a different point:

We have repeatedly used DVs to understand the cost structures of our various strategic business units (SBUs). Firstly, the DVs aid us in understanding what is happening in the costs of each SBU and secondly, they provide a means of comparison across the SBUs. Such an analysis obviously leads to thoughts like “If SBU A is using similar processes as that of SBU B, then why does SBU A have a higher cost structure”, or “If SBU A has used this much in the past, why is it using so much now”. Such an evaluation obviously leads to further analysis from which we can isolate the actual problems. Once we understand the inherent problems in the processes, we obviously correct it some way or another. This leads directly to cost savings. And yes, that leads to improved profitability. On a one-on-one basis, I would say that the contribution is too meagre. On an aggregate basis both on time and across SBUs, though the contribution is low, it is still considerable. In such times of economic crunch and financial prudence, even such contributions are credit-worthy. Also, I would envision that DVs would have a stronger role in profitability the more they proliferate across business processes.”

The above response clearly indicates that DVs have aided in cost savings that directly lead to improvement in profitability. Similar responses were also obtained from some other participants and most of them reported that the contribution is still very low in this regard. However, given the requirement of thorough cost control, even such contributions are greatly appreciated.

The following response by business manager from the retail industry indicates a different point of view:

I would say that the biggest advantage of DVs is that of time savings. Despite the initial investment that needs to be made in understanding a particular kind of DV and data preparation, I still think that it is a good investment. Once something is standardized, it becomes a matter of plug and play. For instance, I began using a heat map to understand the sales of a particular brand across all our retail outlets in the nation. The first two or three times, we did it, it did take a lot of time. Now, it is simply a plug and play thing–clean the data and feed it, and voila, we have the analysis out. We’ve been using that analysis for the last one year. Once we have that time saving, we can invest it in other activities. I’m not saying that increased time in other activities will impact the profitability directly. However, there has got to be some kind of butterfly effect in place, right?”

Time saving was cited as one of the biggest advantages of using DVs. Although, some participants were adamantly against the increased time investment required in understanding and data preparation required for the DV, almost all participants agreed that once the DV is standardized for a particular kind of analysis, it definitely results in time savings in subsequent usages. Most participants were of the opinion that the time saved due to DVs could be invested in other productive activities that improved the overall productivity, and subsequently profitability. However, not all participants could point out the link between productivity improvements and profitability improvements.

The scenario in the FMCG industry was slightly different. The following is a response of a top management participant from the FMCG industry:

Undoubtedly, once a specific kind of analysis is frozen, it definitely aids in improved cost and time savings. However, the kind of data sets that we see in the FMCG industry, it is not always possible to create simply a plug-and-play solution. We need thorough quality control and quality assurance before we can even use the data in the analysis. It is this specific aspect of the FMCG industry that I think makes DVs less effective.”

The researcher presented the viewpoint that in the retail industry as well the data sets are very large, but in that industry, DVs are seen to beneficial to profitability through time savings. To this, the above participant continued:

Undoubtedly, the data sets even in the case of the retail industry are humungous. But you also need to understand that a majority of that industry is automated. Everywhere the data is being captured through computers. In the case of FMCG, we are dealing with too many people. Yes, the people are using computers too at almost every point. But the very fact that there is so much reliance on human entry necessitates that we perform quality control and assurance each and every single time. This would be the primary difference in my opinion.”

In the FMCG industry, humungous data sets and contrarily lack of data were cited as major deterrents.

4.6 DV and Overall Decision Making

Question 7: Decisions made on the basis of DV are significantly more precise, informed and inclusive. Your thoughts.

All participants agreed with the given statement with some inhibitions. Participants felt that DVs generally combine varying sets of data to generate one visual representation, they become inclusive by default. At the same time, since underlying data is being used to create the representations, they ought to be informed. In terms of precision of the output, some participants felt that it was dependent on the creator of such representations and could not be relied upon always. The following response by a top management participant from the financial services industry indicates this line of thinking:

I agree that DV generally ensures inclusiveness and leads to informed decisions. However, in terms of precision, I have my doubts. Several times, I’ve had to spend time in understanding a particular visualization. And once I see something new, I’m keen on understanding it entirely, especially when it is concerned with business. Often, I’ve seen that the precision of such visualizations is not very good. Sometimes, I feel that more data should have been included. Sometimes, I feel that less data should have been included. Sometimes, I feel that the person handling the generation of that visualization doesn’t entirely understand its merits and thereby leads to suboptimal usage. You see, I can go on and one about the precision of DVs. Thus, I would not agree that decisions made on the basis of DVs are always precise since DVs themselves are not always precise.”

The primary concern with regards to the precision of decision made using DV is the precision of DVs itself. Additionally, some participants stated that the user of the DVs–the one who uses the analysis–is not always in sync with the visual representation as they are a relatively new phenomenon. Given the varied personalities that might be using the same DV to take varied decision as specific to their context, there cannot be one consensus on the right interpretation of the given visualization.

representations, they ought to be informed. In terms of precision of the output, some participants felt that it was dependent on the creator of such representations and could not be relied upon always. The following response by a top management participant from the financial services industry indicates this line of thinking:

I agree that DV generally ensures inclusiveness and leads to informed decisions. However, in terms of precision, I have my doubts. Several times, I’ve had to spend time in understanding a particular visualization. And once I see something new, I’m keen on understanding it entirely, especially when it is concerned with business. Often, I’ve seen that the precision of such visualizations is not very good. Sometimes, I feel that more data should have been included. Sometimes, I feel that less data should have been included. Sometimes, I feel that the person handling the generation of that visualization doesn’t entirely understand its merits and thereby leads to suboptimal usage. You see, I can go on and one about the precision of DVs. Thus, I would not agree that decisions made on the basis of DVs are always precise since DVs themselves are not always precise.”

The primary concern with regards to the precision of decision made using DV is the precision of DVs itself. Additionally, some participants stated that the user of the DVs–the one who uses the analysis–is not always in sync with the visual representation as they are a relatively new phenomenon. Given the varied personalities that might be using the same DV to take varied decision as specific to their context, there cannot be one consensus on the right interpretation of the given visualization.

Click here to have a similar A+ quality paper

Order Now