J. of Acc. Ed.

xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

J. of Acc. Ed.
journal homepage: www.elsevier.com/locate/jaccedu

The role of data visualization and analytics in performance

management: Guiding entrepreneurial growth decisions
Julia Kokina a,⇑, Dessislava Pachamanova b, Andrew Corbett c
a
Accounting and Law Division, Babson College, 231 Forest St., Babson Park, MA 02457, United States
b
Analytics and Computational Finance, Zwerling Family Endowed Research Scholar, Math and Science Division, Babson College, 231 Forest St., Babson Park,
MA 02457, United States
c
John E. & Alice L. Butler Venture Accelerator, Entrepreneurship Division, Babson College, 231 Forest St., Babson Park, MA 02457, United States

a r t i c l e i n f o a b s t r a c t

Article history: This case introduces you to a small but growing e-commerce venture and asks you to
Received 30 June 2016 assume the role of an accountant tasked with designing a performance management sys-
Received in revised form 23 December 2016 tem that aligns with the venture’s growth objectives. Using a sample of actual customer,
Accepted 24 December 2016
order, and revenue data, you are guided to develop visualizations in Excel and Tableau
Available online xxxx
and communicate your findings. Finally, the case challenges you to map business problems
with analytical techniques such as regression, decision trees, and clustering in order to pri-
Keywords:
oritize activities and manage the growth the company has experienced to date. The case
Analytics
Big Data
addresses a growing need for accountants to develop competency in predictive analytics
Data visualization (PwC, 2015).
Predictive modeling Ó 2016 Published by Elsevier Ltd.
Performance management
Strategy map
Balanced scorecard
Entrepreneurial growth
Heatmap
Metrics
Key performance indicators
Dashboard

1. Introduction

The goal of most companies, whether a startup or a large multinational corporation, is to grow net income by increasing
revenues. It is logical to think that in order to grow, a company should pursue revenue opportunities from any source pos-
sible. However, most successful companies focus their growth efforts in areas where they have built their distinctive capa-
bilities; strengths that set them apart from everyone else (Leinwand & Mainardi, 2016). To make sound strategic growth
decisions, companies must establish a coherent performance management system that includes critical metrics that support
a company’s value proposition. To do so, companies must find ways to select, validate, and prioritize performance measures.
This project gives you hands-on experience with analytics and performance management in an e-commerce venture. This
project also challenges you to think of ways to incorporate financial and non-financial data to guide the venture at a critical
time in its evolution as it attempts to thrive and grow. This case is unique as it highlights the interdisciplinary nature of run-
ning a company where analytics help create a performance management system that informs strategic growth decisions.

⇑ Corresponding author.
E-mail addresses: [email protected] (J. Kokina), [email protected] (D. Pachamanova), [email protected] (A. Corbett).

http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
0748-5751/Ó 2016 Published by Elsevier Ltd.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
2 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx

1.1. Background: data and analytics

Data and analytics are critical components of successful performance measurement and management for growth. A com-
pany’s operations and interactions with customers are constant generators of data that are often rich with information. How-
ever, the abundance of data and plethora of data sources are also responsible for many challenges faced by companies trying
to drive decisions based on data. For example, some data are structured – stored in rows and columns in spreadsheets or in
relational databases – and are easy to retrieve and analyze. Other data – such as email data or website traffic data – are
referred to as semi-structured and require more effort to process and summarize in useful ways. Finally, there is often a
wealth of unstructured data associated with a company – customer comments, reviews, testimonials; social media discus-
sions and references; customer service call transcripts. Storing, retrieving, and managing all different types of data, as well
as prioritizing performance measures based on these data to help a company grow are complex problems faced by many
companies today.
Tracking performance via performance measures is an example of descriptive and diagnostic analytics; one can obtain
information about how the company is doing and diagnose potential issues. Many companies have grown to appreciate
the importance of employing various types of analytical techniques to move from the descriptive and diagnostic to predictive
and prescriptive (see, for example, Kiron, Prentice, & Ferguson, 2014; Parr-Rud, 2012). Predictive analytics involve employing
complex analysis based on statistical, data mining, and machine learning techniques to detect relationships and make pre-
dictions about the future or about categories of data that are not available (such as what type of customer would find a pro-
duct appealing). Prescriptive analytics go a step beyond and use modeling to make recommendations about the best course of
action given the predictions made; such as how to schedule sales representatives given predicted demand for goods and/or
services.
Accountants often play central roles in supplying senior management with timely and reliable information to inform cru-
cial strategic decision-making activities. Accountants, therefore, have the opportunity to be at the forefront of analytics
implementation in their companies. A necessary step toward accomplishing this goal is for accountants to achieve compe-
tency in the analytical techniques that take advantage of data from multiple sources and of multiple types – structured,
semi-structured, unstructured, big and small.

1.2. The business life cycle

Bombas, the startup in this case, is an early stage company that is in need of the advice from an ‘‘analytics-savvy” accoun-
tant. It is just a couple of years old and has been quite successful to date. Bombas has survived the difficult first stages of
development and now has a handful of products that it has sold to over one hundred thousand customers in just the
two-and-a-half years it has been in business. Fig. 1 contains the classic life cycle depiction which can be used to illustrate
either the life cycle of an individual product or the life cycle of a business. During the introduction stage, companies are
focused on trying to acquire the resources needed to get started and develop the necessary technical capabilities to get into
the marketplace. Next, as companies begin to grow, they have to overcome issues related to commercialization: the produc-
tion and scaling of their products and the business as a whole. As growth accelerates and the venture begins to move toward
maturity and stability, companies struggle with issues such as sales volumes, market share, profitability, internal controls,
and where their future growth will come from.
While the success Bombas has enjoyed is admirable, it now faces the challenge of stabilizing and growing the company.
The company is currently in the growth phase and is managing the myriad issues that come as a startup attempts to scale.
Top management relies on you as their accountant to assist in the analysis of data and to provide advice. Data and analytics
can be used to prioritize activities and manage the growth the company has experienced to date. This case provides you with

Fig. 1. Lifecycle of a product (or business) (adapted from Levitt (1965)).

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx 3

actual revenue data to develop and deploy analytical techniques to guide the continued success of a new venture in order to
avoid the natural progression toward decline.

2. Scenario

After graduating from the top entrepreneurship college in the US, brothers Andrew and David were eager to begin their
journeys as entrepreneurs. Their goal was to establish a company that would be both profitable and help those in need. After
learning that socks were the most requested item in homeless shelters across the US, their idea for an online athletic and
leisure sock company with a social mission was born. Along with their two other co-founders, they named their company
Bombas and spent two years on research and development to engineer a sock that was built for both athletic performance
and extreme comfort; a sock that looked and felt better than any other sock on the market. The Bombas business model is
similar to the business models of companies like Toms and Warby Parker – for each pair of socks sold, one pair is donated to
homeless shelters across the country.
After finalizing the design, the co-founders launched a crowdfunding campaign on indiegogo.com in August of 2013.
Bombas remains the #1 most funded apparel campaign on the platform to date, with $142,488 raised in purchase commit-
ments by 2751 people in one month.1
Bombas’ original and primary product has seven unique features: long staple, combed pima cotton, performance footbed,
stay-up technology, blister tab, y-stiched heel, honeycomb support system, and invisitoe. Since its initial product launch,
Bombas has continued to innovate, design and offer a number of different styles, colors and materials. Socks are Bombas’
only product, and they are sold exclusively online at bombas.com. Most pairs of socks currently retail for $12, with specialty
products reaching $18 per pair. To date, Bombas has donated more than one million pairs of socks to those in need.
Bombas has enjoyed great success with its products; it has seen a consistent increase in new customers and has an exten-
sive social media following. The founders are seeking the advice of management accountants. The Bombas leadership needs
help understanding revenue and order data, increasing repeat customer rates, optimizing website use, and spending on
marketing.

3. Software and data

You will use Excel and Tableau to develop analytics solutions for Bombas’ revenue data. Your instructor will provide you
with an Excel workbook titled, ‘‘BombasRevenueData.xlsx,” containing a random sample of 1100 customer orders. You will
use this dataset to complete the case requirements.
Visualization software Tableau’s website, http://tableausoftware.com, has a number of training resources. As part of the
preparation for the case, you can watch the first few videos at http://www.tableau.com/learn/training. (For example, the first
lesson in each of the groups of videos Getting Started, Connecting to Data, Visual Analytics, Dashboards and Stories, and
Mapping are a good place to start.) These videos demonstrate how to create a time series graph, a geographic visualization,
as well as a performance dashboard which will be useful as you complete this project.

4. Requirements

Assume you are a management accountant who is tasked by Bombas’ founders to use analytics and develop tools that
would help a cash-strapped e-commerce startup grow revenue. Perform the steps outlined below to complete this project:

1. Take a tour of Bombas’ website at bombas.com and watch the Indiegogo campaign at https://www.indiegogo.com/pro-
jects/bombas-better-socks-better-world-bee-better#/ to learn about the company. Familiarize yourself with Bombas’
social media presence by visiting Bombas on Facebook and Instagram. What are your initial impressions about the com-
pany? What are some of the challenges the company might be facing? Who are Bombas’ customers?
2. Read Appendix A and perform the step-by-step guidelines for analytics applications for Bombas’ revenue data using Excel
(pivot tables, time series graphs, heatmaps, and correlation analysis) and Tableau (geographic visualizations and perfor-
mance dashboards).
3. Interpret the Excel and Tableau output you generated. What recommendations would you make to Bombas’s founders
regarding its: customer location, behavior on the website, order patterns and revenues? When should Bombas schedule
the work hours of its customer service representatives? Where should Bombas target its future advertising campaigns?
4. Read Appendix B and think of ways for Bombas to measure its performance over time using the Balanced Scorecard
approach. Table 1 presents an incomplete Balanced Scorecard that Bombas is using. Your task is to recommend 2–3 more
measures and metrics to Bombas founders for each scorecard category: learning and growth, internal process, customer,
and financial.
5. What kind of data would you need for the metrics you recommend to Bombas? Are the data structured, semi-structured,
unstructured? (To answer this question, feel free to use external information from various online sources.)

1
The full indiegogo campaign is available at https://www.indiegogo.com/projects/bombas-better-socks-better-world-bee-better#/.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
4 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx

Table 1
Incomplete balanced scorecard.

Perspective Strategy Pursued Measure Metric

Financial Increase revenue Revenue growth Revenue increase/revenue last period
Customer Acquire and retain customers Customer Lifetime Valuea Average sale per customer  number of sales per customer
yearly  average retention time
Cost per Acquisition Acquisition spend/number of conversions
Internal Achieve timely delivery of product Shipping Cost Relative to Total Shipping cost/total revenue
Process at a reasonable cost Revenue
Learning and Learn how to create a customer- Maximize Loyalty Change in Net Promoter Scoreb
Growth focused culture
Create innovative products Gain customer insight on new Analysis of focus groups, social media, bloggers,
product creation influencers
a
Customer Lifetime Value (CLV) is the predicted revenue a customer will realize for the company over the lifetime of the customer’s relationship with the
company. The measure has been adapted from https://www.quora.com/ What-are-the-key-KPIs-to-track-for-a-consumer-internet-startup. For more ways
to calculate CLV, refer to https://blog.kissmetrics.com/how-to-calculate-lifetime-value/.
b
Net Promoter Score (NPS) is a measure of company’s performance through customer lens. Customers are asked the following question: ‘‘How likely is it
you would recommend us to a friend?” on a scale from 0 to 10. If a customer responds with 10 or 9, that customer is considered a ‘‘promoter”; if 8 or 7, that
customer is a ‘‘passive”; if between 0 and 6, a customer is a ‘‘detractor”. The NPS is computed by subtracting the percentage of detractors from the
percentage of promoters (Bain, 2016).

6. Read Appendix C. Keeping in mind the Balanced Scorecard you just designed, what are some of the advanced analytical
techniques you could use to help Bombas?
7. Present your findings in a brief PowerPoint presentation for the startup founders. Outline your recommendations and the
tools you have used to generate those recommendations. Your deliverable must include your output from Excel and
Tableau.

5. Teaching notes

5.1. Motivation and learning objectives

In a 2015 report based on results from a survey of over 1300 CEOs in more than 60 countries, PwC argued that, ‘‘Univer-
sities should infuse analytical exercises into existing curriculum to help students develop data analytics proficiency on top of
their core accounting skills” (PwC, 2015, 14). PwC also pointed out the growing need not only for, ‘‘skills for technical
problem-solving, but also on problem-framing, so students learn to ask the right questions and think strategically” (PwC,
2015, 1). This case helps develop analytics-related and problem-framing skills indicated by PwC and other major accounting
service providers as necessary for university graduates to succeed in the current business environment. It is also motivated
by the need for accounting programs in the US to integrate modern accounting and business technology to enable the devel-
opment of data analytics skills (AACSB, 2013).
The case introduces students to the process of developing logical accounting analytics solutions to address specific busi-
ness issues. Moreover, the case is based on real company data and presents issues relevant to many companies today. The
case focuses on the following analytics-related skills:

 Identifying key business issues at a particular growth stage for a company (Requirement 1).
 Developing data visualizations in Excel and Tableau to glean insights from revenue data (Requirement 2).
 Interpreting the results and recommending a course of action (Requirement 3).
 Determining appropriate metrics to track performance, diagnose problems, and look for solutions (Requirement 4 and 5).
 Matching the business issues with most suitable predictive analytics solutions (Requirement 6).
 Communicating the findings in a manner tailored to the receiving audience (Requirement 7).

5.2. Course use and student prerequisites

This case can be used with software packages like Excel and Tableau. Students are provided with specific examples of
analytics solutions in the Appendices. The case can be incorporated in managerial accounting courses at both the undergrad-
uate and the graduate levels to facilitate a richer discussion of data-driven performance measurement and management.
Specifically, it can be introduced when covering balanced scorecards and strategy maps. Another successful use of this case
could be in a course that introduces students to accounting analytics, such as a course in accounting information systems or
auditing. The case can be taught in conjunction with Janvrin, Raschke, and Dilla (2014) or Igou and Coe (2016). The only pre-
requisite for the case is basic competency in Excel. In our experience, students learn Tableau quickly through the instruc-
tional videos available on the software provider’s website.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx 5

5.3. Implementation guidance

To implement this case, students first need to become familiar with the case scenario and get to know Bombas as a com-
pany by watching its Indiegogo campaign video online. Subsequently, the students review and graph the Bombas transac-
tional data presented in the Excel spreadsheet. For example, students may use a heatmap from the initial analysis to
visualize the times of the day and the week when the most orders are placed. This can inform marketing campaign decisions
or operational decisions related to staffing of customer service representatives. Students may also use a map visualization to
summarize information about customer locations (Who is Bombas reaching?), the order activity in different geographical
areas (What areas should be targeted in the future?), and product preferences for different geographical areas (How much
of different products should be stocked in different locations?).
After viewing the data and acquainting themselves with the company, students should read the background information
on strategic performance measurement in Appendix B. Appendix B reviews the four important perspectives in building a
strategy map for a company: the financial perspective, the customer perspective, the internal process perspective, and
the learning and growth perspective. It also summarizes some important metrics used to measure performance from each
of these perspectives. The instructor could discuss these metrics with students and solicit input on additional metrics that
Bombas may want to consider.
While most accounting fundamentals discussions would stop at this point, a data analytics-centric discussion would
cover the following additional issues that are likely to be challenging in practice:

1. What kind of data would Bombas need to compute the metrics designed as part of the balanced scorecard? Where would
the data come from, and how would the data be stored and retrieved regularly for analysis purposes?
2. What kind of performance metric visual displays would be helpful for decision making?
3. Are these performance metrics sufficient for understanding the leading indicators of performance?

The performance metrics described in Appendix B should be tracked on a periodic basis, and the graphs in Appendix A
represent examples of visualizations that can be used. For example, the time series of a financial metric such as increase
in the number of orders over the previous period could be plotted to monitor order growth.
Item 3 in the list above leads to a discussion about applications of advanced analytical techniques: the observation of rela-
tionships and the discovery of deeper insights. One possible direction of discussion could be at the aggregate level; the study
of correlations and the use of predictive models such as regression to find out if a particular metric such as number of page
views correlates with an increase in the number of orders placed by a customer.
Another possible direction of discussion is at the individual customer level. Bombas could apply advanced analytical tech-
niques to understand its customers better. For example, Bombas could use regression techniques to estimate a particular
customer’s lifetime value (Bombas is already estimating this value through an external vendor). Cluster analysis can be used
to identify groups of customers who have similar behavior or demographics. If data were available on: the actual cost of each
pair of socks a customer purchases, the cost of shipping (if the order met the minimum for free shipping), and coupon dis-
counts, one could estimate the total cost of the customer. Subtracting the total cost from the sales price of the order would
determine profitability of each customer. One could request this additional data and then determine if customer profitability
were associated with particular types of customer behaviors.
Finally, one may request additional data to discuss product profitability and bundling. Association rules (also called mar-
ket basket analysis) could be used to determine if socks of certain colors or types were purchased together, which could
inform product bundling and new product decisions.
The advanced analytics discussion can be guided by Table 4 in Appendix C, which summarizes types of business questions
that can be answered with some common advanced analytical techniques. The instructor could start with the balanced
scorecard (Table 1) as a foundation for the discussion. Specifically, once students have identified metrics that are important
for company growth, the instructor could ask how these metrics could be improved. If Bombas’ goal is to increase average
Customer Lifetime Value (in the Customer category), for example, then decision trees or regression could be used to identify
characteristics that determine the Customer Lifetime Value, and the results could be used to design marketing campaigns
that target customers with the desired characteristics. If Bombas’ goal is to increase Email Click Rate (in the Customer cat-
egory), then text analytics can be used to identify the most effective wording for emails, and logistic regression and decision
trees can be used to identify customers most likely to click on the email, so that the targeting can be effective. Sentiment
analysis on social media data can be used to calculate the Customer Satisfaction Score identified in the Learning and Growth
section of the strategy map.

5.4. Scoring criteria and sample solution

We suggest to include the following items in developing a scoring criteria to evaluate student performance:

 The number and appropriateness of metrics suggested by the students that Bombas can use for performance evaluation.
 Whether the appropriate formats and visualizations are used to illustrate metrics and results from the application of
analytical techniques.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
6 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx

 Whether the presentation informs a decision maker enough so that the recommendations can be used to pursue a data-
driven growth strategy.
 Whether the student could suggest additional advanced analytical techniques that help uncover additional relationships
in the data and drive a deeper understanding of the company and its performance.

5.5. Evidence of efficacy

We collected efficacy data for two classes of an undergraduate managerial accounting course in a private college. The
instructor was one of the authors. Descriptive statistics for the class profile are presented in Panel A of Table 2. Panel B of
Table 2 summarizes the students’ competency in spreadsheet and visualization software prior to the case.
The case was presented at the end of the semester in two 75 min sessions. The first session began with a discussion on
performance management in the context of sustainability. In the last 25 min of the first session, the students watched the
Shark Tank video (Season 6, Episode 1, released on September 26, 2014) and were introduced to the case and the assignment.
They were then asked to complete Requirements 1–3 of the case before the following class.

Table 2
Descriptive statistics and efficacy data.

Pre-Case Post-Case
Panel A: Descriptive statistics
Gender
Male 42 (58.33%) 30 (50.85%)
Female 30 (41.67%) 29 (49.15%)
Average Age 19 19
Class standing
Freshman 0 (0%) 1 (1.69%)
Sophomore 65 (90.28%) 53 (89.83%)
Junior 6 (8.33%) 4 (6.79%)
Senior 1 (1.39%) 1 (1.69%)

Current estimated level of knowledge of MS Excel Tableau or other advanced Advanced data analytics and statistical modeling
analytics software visualization package software such as R, Minitab, SPSS, SAS
Panel B: Student familiarity with spreadsheet and visualization software packages before the case
None 4 (5.56%) 42 (58.33%) 10 (13.89%)
Beginner 12 (16.66%) 20 (27.78%) 40 (55.56%)
Intermediate 45 (62.5%) 9 (12.5%) 19 (26.39%)
Advanced 11 (15.28%) 1 (1.39%) 3 (4.16%)

Statement Pre n Post n Pre mean (SD) Post mean (SD) Significance
Panel C: Results of pre-case and post-case surveysa
1. I understand how data analytics can be used to answer 72 59 4.4028 (0.5972) 4.5932 (0.5607) p = 0.0643*
important business questions
2. I understand how to use data visualization tools to 72 59 4.0833 (0.6446) 4.4068 (0.5607) p = 0.0030**
explore data for business insights
3. I understand how advanced analytics techniques can be used 72 59 4.1667 (0.6713) 4.3390 (0.6594) p = 0.1431
to determine relationships in data and to predict behaviors
4. I understand how to evaluate both financial and non-financial 72 59 4.1944 (0.5474) 4.2712 (0.6652) p = 0.4701
aspects of business performance
5. I understand how to create performance metrics that are 72 59 3.8333 (0.5814) 4.3051 (0.6500) p < 0.0001***
aligned with the pursued strategy
Statement (n = 52) Mean (SD) Significance
(Difference from Neutral (3))
Panel D: Results of debriefing questionnaire
1. I found the case to be engaging and interesting 4.0680 (1.015) p < 0.0001***
2. I found the case to be an effective way to gain an understanding of 4.1017 (0.9772) p < 0.0001***
how data analytics can be used to answer important business questions
3. I found the case to be an effective way to gain an understanding of data 3.8140 (1.137) p < 0.0001***
analytics software
4. I found the case to be an effective way to develop skills and knowledge related 4.0339 (0.9462) p < 0.0001***
to data analytics from an accountant’s perspective
5. I think that this case should be used in future classes 4.1690 (1.085) p < 0.0001***

Scale: 1 = Strongly disagree, 2 = Disagree, 3 = Neither agree nor disagree, 4 = Agree, 5 = Strongly agree.
a
Due to time constraints, not all students completed the post-survey.
*
Significant at the 10% significance level.
**
Significant at the 5% significance level.
***
Significant at the 1% significance level.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx 7

The following class started with the Beyond the Tank video (Season 2, Episode 14, released on May 15, 2016) followed by
a discussion of the visualizations created by the students (a total of 20 min). The students were then asked to work in groups
for 15–20 min to develop strategies and metrics for Bombas performance management, keeping in mind the interpretation of
the visualizations. To wrap up the class, in the remaining 30 min each group was asked to share their findings with their
classmates.
To assess the efficacy of this case, we developed a questionnaire with five statements related to the case learning objec-
tives. Students were asked anonymously to indicate the level of agreement or disagreement with each statement before and
after the case. We used a five-point Likert scale where 1 = Strongly disagree, 2 = Disagree, 3 = Neither agree nor disagree,
4 = Agree, and 5 = Strongly agree. Pre-case and post-case results are presented in Panel C, Table 2.
The results of this questionnaire show a statistically significant improvement with regard to the students’ perceptions of
the primary objectives of the case: creating performance metrics that are aligned with the pursued strategy and understand-
ing of how to use visualization tools to explore data for business insights (Questions 2 and 5 in Panel C, Table 2). Students
also gained an understanding of how data visualization and more advanced analytics can be used to answer business ques-
tions (Questions 1 and 3). The results for Question 4 do not seem to indicate statistically significant improvement; however,
it should be noted that during the first session and before the pre-case survey was conducted, students were exposed to non-
financial metric evaluation as part of the sustainability discussion. Thus, the insignificant change between the first and the
second class session may be reflective of the fact that they already felt knowledgeable in this area. (The average pre-case
score on that question was 4.19 out of 5.) It should be noted also that by their sophomore year, undergraduate students
at the authors’ institution have taken at least two courses in analytics and modeling, and that may explain the comfort of
the class with applications of visualization and modeling techniques both in the pre-case and post-case survey. The results
from the pre- and post-case surveys indicate, however, that the case is effective at providing a context in which the students
in the class can appreciate the application of analytical techniques (Questions 1, 2 and 5).
We also assessed the students’ overall impressions of the case by conducting a short debriefing questionnaire adopted
from Igou and Coe (2016). The results of this questionnaire (see Panel D, Table 2) demonstrate that the students had a pos-
itive experience with the case and would like to see it used in future classes.
To assess student learning, we created a mapping exercise as shown in Panel A, Table 3. (Possible answers to the mapping
exercise are shown in the solutions file.) Panel B of Table 3 shows the results of a pre- and post-case survey of learning out-
comes. The students in the class were asked anonymously to answer questions about the most appropriate analytical tech-
nique to use to answer a particular business question. The p-values for the statistical tests conducted on the significance of
the difference in proportion of students who answered the question correctly before and after the case are provided. As the
results in Panel B, Table 3 illustrate, the case was effective in teaching students how to map analytical techniques to business
questions.
An abbreviated version of this case was also implemented in a graduate managerial accounting course, with one of the
authors as an instructor. Twenty-two students participated in this case and were asked to assess briefly their experience

Table 3
Mapping exercise and its results.

Scatterplot Pivot Time Map Numeric summaries (Means,

(contingency) series visualization correlations, etc.)
table plot
Panel A: Mapping exercise
Analyze the frequency of orders over a period of time
Summarize number of orders placed by time of day
Determine if some geographical areas have a higher
number of orders than others
Determine if the size of a customer’s order is related to
the amount of discount offered
Explore if repeat customers have characteristics that
single-time customers do not
Pre-Case (% Students Who Post-Case (% Students Who Significance
Answered Correctly, n = 72) Answered Correctly, n = 59)
Panel B: Pre-case and post-case results of the mapping exercise
Analyze the frequency of orders over a period of time 62.50% 79.66% p = 0.0270**
Summarize number of orders placed by time of day 65.28% 76.27% p = 0.1630
Determine if some geographical areas have a higher 81.95% 94.91% p = 0.0160⁄⁄
number of orders than others
Determine if the size of a customer’s order is related to 83.34% 81.36% p = 0.7860
the amount of discount offered
Explore if repeat customers have characteristics that 30.55% 45.76% p = 0.0720*
single-time customers do not
***
Significant at the 1% significance level.
*
Significant at the 10% significance level.
**
Significant at the 5% significance level.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
8 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx

Table 4
Map of business problems and analytical techniques.

Type of problem Analytical technique What the analytical technique does

Determine the impact of one or more factors Regression Calculates an equation that links the values of inputs to the value of the
on an output variable of interest output variable, so that one can predict the value of the output variable
based on the values of the inputs
Determine the factors affecting an output Decision Trees Calculates rules starting with the most important input variable that allow
variable and display them in order of one to predict the value of the output variable
importance
Group items by similarity Clustering Places items into clusters based on measures of similarity or distance
between the items
Determine if groups of items occur together Association Rules Based on how often groups of items occur together and how often subsets of
in customer purchase transactions (Market Basket these items occur when a particular group occurs, creates rules that, for
Analysis) example, allow one to determine whether customers have a high likelihood
of buying one or more additional items given what is in their shopping cart
so far, or what types of items are often purchased together
Extract information from unstructured (text) Text Analytics Analysis of text to identify key words and phrases that help predict an
data output variable of interest or summarize a sentiment (see also Sentiment
Analysis)
Extract information about customer opinion Sentiment Analysis Analysis of text with the goal of extracting information about the sentiment
from text (blogs, comments, etc.) expressed in the text. The result is usually a report on whether the sentiment
in a piece of text is positive, neutral, or negative, based on the occurrence of
expressions deemed useful for identifying the sentiment

with the case. Eighty-seven percent of the students expressed agreement that the case prompted them to learn more about
applications of data visualization and that the case increased their level of understanding of data-driven performance mea-
surement. Ninety-five percent of participants agreed that the case prompted them to think about meaningful ways to visu-
alize information.

5.6. Case extensions

The case of Bombas and its associated data provide instructors an opportunity to teach students about an important stage
in a company’s evolution. The case can be extended to a discussion of analytical performance management which examines
statistical relationships between financial and nonfinancial metrics (Davenport, 2008). A discussion could address how
regression modeling and the analytical hierarchy process (Campbell & Lu, 2010) could be implemented to prioritize perfor-
mance metrics.
The case can also be extended to provide a glimpse into how the advice and counsel of accountants assist top managers in
numerous business decisions across different functions. The data in the case can help students understand how they can use
sales data to inform the company about production-related issues.
Entrepreneurial ventures at this stage of development need to grow market share and find a future base for growth. Data
provided in the case can help the venture with growth decisions such as: where to spend ad dollars, which customers are
most profitable, which products are most profitable, are certain regions more or less likely to generate sales, and more.

6. Case solution and Excel data

The solution and the spreadsheet with the order data are available upon request from the corresponding author.

7. Conclusion

There is a growing need for accountants to incorporate various types of data and analytics to inform strategic decision-
making of management. This case introduces accounting students to data-driven performance management in the context of
a rapidly growing new venture. Students are asked to identify strategic objectives for the venture and design a balanced
scorecard with specific leading and lagging measures and metrics. Using a sample of the venture’s actual customer, order,
and revenue data, the case provides a hands-on experience with descriptive and diagnostic analytics using visualization.
It then provides a context for discussing more sophisticated statistical analytics techniques such as regression, decision trees,
and cluster analysis. The case can be completed using Excel and Tableau. Preliminary feedback suggests that this case
achieves its goal of helping students understand the role of accountants in using analytics to guide performance manage-
ment decisions.

Acknowledgements

We are very grateful to Andrew Heath and Alex Yu of Bombas for meeting with us and providing data for this case. We
thank PwC for providing financial support for writing the case. We also thank all the students who tested this case. Finally,

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx 9

we would like to thank the editor of the Journal of Accounting Education, Natalie T. Churyk, the special issue editors, Diane
Janvrin and Marcia Weidenmier Watson, as well as the two anonymous reviewers for their critiques, direction and
support.

Appendix A. Guidelines for performing analytics in Excel and Tableau

This appendix contains step-by-step guidelines for implementing visualizations and obtaining insight from Bombas’ rev-
enue data. While most students are familiar with Excel, they may be less familiar with Tableau. However, it is a relatively
simple and intuitive program and we encourage students to just begin using it. they will feel comfortable with it in a matter
of minutes.
To perform these analyses, use the data file titled BombasRevenueData.xlsx. The file contains the following columns: Cus-
tomer number, State, Zip code, Registration date, Last site visit, Number of site visits, Number of product page views, Number
of add to cart events, Number of checkout events, Number of orders placed, Number of items purchased, Order 1 revenue,
and Order 1 purchase date.
As you work with the data, please answer the following question and then map each business question to the most appro-
priate analytics technique (refer to Panel A, Table 3):
What variables in the case dataset do you think would be helpful to analyze to identify key business issues Bombas is
facing at this growth stage?
Visualize the frequency of orders placed during a time period.

 This could be accomplished by creating a time series graph in Excel.

 Transform ‘‘Order 1 purchase date” (this column includes the time of purchase along with month, date, and year) into
‘‘Order 1 MMDDYYYY” in order to create a count of orders by day. To do so, create a new column and use the following
formula:=DATE(YEAR(M2),MONTH(M2),DAY(M2)), where Column M is ‘‘Order 1 purchase date”.
 Create a PivotTable by clicking Insert, PivotTable. Make sure the option to create a New Worksheet is selected. Drag the
newly created field ‘‘Order 1 MMDDYYYY” to the area under ‘‘ROWS”. Then drag the same field ‘‘Order 1 MMDDYYYY”
P
to the area under ‘‘ VALUES”.
 To create a time series graph, click on Insert, Line Chart. Select the first line chart to the left under the 2-D Line.

Summarize the frequency of occurrence of a specific number of orders placed by individual customer for the duration of
the time period in the data set.

 This could be done using a pivot table.

 Create a PivotTable by clicking Insert, PivotTable. Make sure the option to create a New Worksheet is selected. Drag the field
P
‘‘Number of Orders Placed” to the area under ‘‘ROWS”. Then drag ‘‘Number of Orders Placed” to the area under ‘‘
P
VALUES”. Finally, click on the button ‘‘Sum of Number of Orders Placed” under ‘‘ VALUES” and select ‘‘Value Field
Setting,” then ‘‘Count.”

Visualize the frequency of orders placed by the day of the week and the time of the day.

 This could be done using a heatmap in Excel.

 Create a new column titled ‘‘WEEKDAY” by using the following formula:=WEEKDAY(N2,1), where ‘‘N2” is ‘‘Order 1 pur-
chase date” column and ‘‘1” is to indicate that the week begins with Sunday (1) and ends with Saturday (7).
 Create a new column titled ‘‘HOUR” by using the following formula:=HOUR(N2), where ‘‘N2” is ‘‘Order 1 purchase date”
column.
 Create a PivotTable by clicking Insert, PivotTable. Make sure the option to create a New Worksheet is selected. Drag the
newly created field ‘‘WEEKDAY” to the area under ‘‘ROWS”. Then drag the newly created field ‘‘HOUR” to the area under
P
‘‘COLUMNS”. Finally, drag ‘‘Order 1 MMDDYYYY” to the area under ‘‘ VALUES”.
 Under the ‘‘Home” tab, highlight the area with order frequencies (excluding the Grant Total column), click on ‘‘Conditional
Formatting”, then ‘‘Color Scales”, and select the first option on the left (Green – Yellow – Red Color Scale).

Determine the strength of the relationship between two variables.

 This could be done by running a correlation analysis in Excel.

 Under the ‘‘Data” tab, click on ‘‘Data Analysis.” (If you do not see ‘‘Data Analysis” as an option in the ‘‘Data” ribbon, click on
File, then ‘‘Options,” then ‘‘Add-Ins,” and select ‘‘Data Analysis Toolpak.”) From the list of analytical techniques, select
‘‘Correlations.” Select the two columns of numbers for which you would like to find the correlation. Click OK.
Alternatively, one can use the CORREL formula in Excel.
=CORREL(Array Reference in Column 1, Array Reference in Column 2)

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
10 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx

will return the correlation between the arrays in Column 1 and Column 2.
 It is always helpful to use a scatterplot to visualize the correlation. In the ‘‘Charts” section under the ‘‘Insert” tab, select
‘‘Scatter,” and then select the two columns to be plotted. (If the two columns with data are highlighted in advance, the
scatterplot will appear automatically. Press the Ctrl key to highlight two non-contiguous arrays of data.)

Visualizations with Tableau can help look into additional issues. Some of these visualizations can be done with Excel as
well; however, it takes more steps to do so.
Visualizations with Tableau:
Visualization software Tableau’s website, http://tableausoftware.com, has a number of training resources. As part of the
preparation for the case, you can watch the first few videos at http://www.tableau.com/learn/training. (For example, the first
lesson in each of the groups of videos Getting Started, Connecting to Data, Visual Analytics, Dashboards and Stories, and
Mapping are a good place to start.) These videos demonstrate how to create a time series graph, a geographic visualization,
as well as a performance dashboard which will be useful as you complete this project. In our experience, students get used to
Tableau’s interface quickly and are able to create their own visualizations within 5–10 min of seeing the software for the first
time.
Perform the following tasks in Tableau:

 Develop a time series plot of number of orders placed for every day in the data set.
 Visualize the aggregate number of orders placed for each day of the month.
 Show a map visualization with all the states in the US and values for the number of orders placed in each state and the
average revenue per order in that state.
 Graph the number of site visits per day for all dates in the dataset.
 Graph the number of page views for all dates in the dataset.
 Design a dashboard that simultaneously displays the map visualization and the graph with the number of page views for
all dates in the dataset.

Appendix B. Strategic performance measurement and management during growth

B.1. Strategy map and balanced scorecard

A strategy map is a helpful visualization tool built around the balanced scorecard concepts that illustrates cause-and-
effect relationships between strategic initiatives presented alongside performance measures or key performance indicators
(KPIs). A thoughtfully designed balanced scorecard is crucial for performance management and internal decision-making as
it enables companies to monitor the strategy execution process. The main attribute of a balanced scorecard is its focus on the
few key parameters that are critical to the success of the venture (Kaplan & Norton, 2004). The process of collecting data for
KPIs consumes organizational resources and could be costly. For example, Danaher, a DC-based company whose primary
business is to acquire and integrate new companies, reduced the number of KPIs from 150 to just 8 in 2006 because of metric
and data overload (Roth & Kleiner, 2016).
Typically, a strategy map presents four distinct areas for performance evaluation. They are (Kaplan & Norton, 2004):

 Financial perspective – shows ways to achieve sustainable growth to satisfy shareholders (lag indicators).
 Customer perspective – depicts success with customers and defines customer segments (a combination of lag and lead
indicators).
 Internal process perspective – demonstrates how value is delivered to customers (lead indicators).
 Learning and growth perspective – focuses on people, technology, and organizational climate (lead indicators).

B.2. Properties of metrics

To design KPIs, it is helpful to keep in mind that in order for a metric to be successful, it should be Simple to understand
and benchmark against; Map to key business activities, actions to results; Actionable – focus attention and guide right
behavior; Reliable and valid; and Timely (SMART). Metrics should also be dynamic where a change in a business process
should be reflected in the change in a metric. Also, metrics should not be collected simply because data are available. Instead,
metrics should ultimately link to company’s success through growth in revenue, profitability, or reduction of risk. Finally,
metrics should not be redundant or convey the same information (a correlation analysis could help identify redundant met-
rics). An example of an incomplete Balanced Scorecard for Bombas, an e-commerce retailer, is presented in Table 1. The focus
is on growing revenue from customers in their target markets while ensuring that they acquire new customers and retain
existing ones. Also, in order to satisfy their customers, Bombas must have systems in place to make sure that orders arrive
on time and that shipping costs are reasonable. Finally, Bombas must focus on fostering a customer-centric culture where
people can create an engaged customer base and sense the pulse for new products that can result in a pipeline of innovative
offerings. Please refer to Table 1 for an example of a Balanced Scorecard.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx 11

B.3. Dashboards

While many companies use strategy maps and balanced scorecards as foundational tools for performance management,
they often use electronic dashboards to view KPIs. A dashboard effectively depicts indicators using graphics which makes it
much easier to tell a story and communicate it throughout the company. It can also be equipped with warning signs or alerts
delivered when a metric is outside of preset parameters (Wolf, 2016).

Appendix C. Analytics background

‘‘Analytics” is a term most broadly used to describe using computational techniques from fields like statistics, computer
science, operations research, and mathematics to extract information from data. The process of extracting information from
data typically relies on building models but data visualization is a critical component of the process, as has been pointed out
also by Janvrin et al. (2014), Sun, Wu, and Liang (2013), and many others). A seminal framework by Keim, Kohlhammer, Ellis,
and Mansmann (2010) describes the process of visual analytics as involving multiple iterations of data visualizations and
model refinements. Specifically, data from multiple sources may be transformed and formatted for user-specific applications.
Ideally, there should be a way for the business user to interact with the data and answer questions of importance to the user
through the ability to ‘‘drill down” and query the data. This can be accomplished, for example, through interactive data visu-
alization software such as Tableau, IBM Cognos, Qlikview, SAP BusinessObjects Lumira, Oracle Data Visualization Cloud Ser-
vice, and others. Approaches such as data mining methods may be employed to extract additional information from the data
that can inform the business user. Patterns discovered through visualization and predictive models inform further analysis
and decision making. The data visualization and the model building therefore work hand in hand to enhance the knowledge
and insights extracted from data (see Fig. 2).
There is an infinite variety of business questions and an infinite variety of visualization and modeling techniques. How-
ever, many business questions can be addressed with a small subset of modeling techniques. In fact, the Rexer Analytics Data
Science Surveys (Rexer Analytics, 2015) have shown that three modeling techniques consistently top the list of algorithms
used by thousands of modelers around the world since the first survey was conducted in 2007: regression, decision trees, and

Fig. 2. Overview of the visual analytics process (adapted from Keim et al. (2010)).

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
12 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx

Fig. 3. The three most widely used types of data mining algorithms. Source: Rexer analytics data science survey (2015). Reprinted with permission.

cluster analysis (see Fig. 3). There are many good references on data mining algorithms for business applications, including
EMC Education Services (2015) and Shmueli, Bruce, and Patel (2016).
Regression in its many varieties (linear regression, logistic regression, etc.) has its roots in statistics and is used to create
models that predict an output variable of interest based on the values of a set of input variables. The model is an equation
that may be used, for example, to predict the lifetime value of a customer (an output variable) based on known customer
characteristics (input variables), or the likelihood that a customer will respond to a marketing offer (an output variable)
based on past customer behavior.
Decision trees hail from computer science and can be generally applied in many of the same situations as regression tech-
niques. Instead of an equation, they produce a set of rules that allow one to predict an output variable. For example, a rule
produced by a decision tree may be that a customer whose age is within a particular range and whose income is larger than a
particular threshold is more likely to buy a given product.
Cluster analysis is used to group items or people into groups (clusters) so that items in a cluster are ‘‘closer” to each other
than to items in other clusters based on a set of characteristics. Cluster analysis is widely used in the context of customer
segmentation as well as data preparation for input into various other algorithms.
Analytics techniques are very important in the context of accounting applications (PwC, 2015): exploratory multivariate
statistics, inferential statistics, visualization tools, optimization methods, machine learning, and predictive analysis tools are
used in auditing applications; visualization in tax applications; data discovery and visualization techniques in the context of
risk management. The ability to identify and frame key business decisions and their related metrics, as well as to commu-
nicate recommendations based on complex analytical techniques are also skills highly prized in the market today.
Table 4 summarizes the correspondence of some important business questions and analytical techniques, and explains
what the analytical techniques do.

Appendix D. Supplementary material

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jaccedu.
2016.12.005.

References

AACSB. (2013). Accounting accreditation standards: Standard A7. Available at: <http://www.aacsb.edu/~/media/AACSB/Publications/white-papers/
accounting-accreditation-standard-7.ashx>.
Bain & Company. (2016). Measuring your net promoter score. Available at: <http://www.netpromotersystem.com/about/measuring-your-net-promoter-
score.aspx>.
Campbell, D. & Lu, R. (2010). LG display. Harvard Business School Case 9-111-072.
Davenport, T. H. (2008). The rise of analytical performance management. SAS White Paper. Available at: <http://www.gleanster.com/system/resources/
W1siZiIsIjIwMTIvMTAvMzAvMTYvNTIvNTYvMzAwL3dwXzU1OTYucGRmIl1d/wp_5596.pdf>.
EMC Education Services (2015). Data science and big data analytics: Discovering, analyzing, visualizing and presenting data. J. Wiley & Sons.
Igou, A., & Coe, M. (2016). Vistabeans coffee shop data analytics teaching case. Journal of Accounting Education, 36, 75–86.
Janvrin, D. J., Raschke, R. L., & Dilla, W. N. (2014). Making sense of complex data using interactive data visualization. Journal of Accounting Education, 32,
31–48.
Kaplan, R. S., & Norton, D. P. (2004). Strategy maps. Converting intangible assets into tangible outcomes. Harvard Business School Press.
Keim, D. A., Kohlhammer, J., Ellis, G., & Mansmann, F. (2010). Mastering the information age: Solving problems with visual analytics. Florian Mansmann.
Kiron, D., Prentice, P. K., & Ferguson, R. B. (2014). The analytics mandate. MIT Sloan Management Review Research Report. Available at: <http://www.
sas.com/content/dam/SAS/en_us/doc/research2/mit-analytics-mandate-107098.pdf>.
Leinwand, P., & Mainardi, C. (2016). Strategy that works. How winning companies close the strategy-to-execution gap. Harvard Business Review Press.
Levitt, T. (1965). Exploit the product life cycle. Harvard Business Review, 43(November–December), 81–94.
Parr-Rud, O. (2012). Driving your business with predictive analytics. SAS Report.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005
 J. Kokina et al. / J. of Acc. Ed. xxx (2017) xxx–xxx 13

PricewaterhouseCoopers (PwC). (2015). Data driven: What students need to succeed in a rapidly changingbusiness world. Available at: <https://www.pwc.com/
us/en/faculty-resource/assets/PwC-Data-driven-paper-Feb2015.pdf>.
Rexer analytics data science survey (2015). Available from <http://www.rexeranalytics.com>.
Roth, G. L. & Kleiner, A. (2016). Danaher’s instruments of change. Strategy + Business. Issue 82, Spring. Available at: <http://www.strategy-business.com/
article/Danahers-Instruments-of-Change>.
Shmueli, G., Bruce, P., & Patel, N. (2016). Data mining for business analytics: Concepts, techniques and applications (3rd Ed). J. Wiley & Sons.
Sun, G. D., Wu, Y. C., & Liang, R. H. (2013). A survey of visual analytics techniques and applications: State-of-the-art research and future challenges. Journal of
Computer Science and Technology, 28(5), 852–867.
Wolf, R. (2016). Dashboard design. Strategic Finance. June 1.

Please cite this article in press as: Kokina, J., et al. The role of data visualization and analytics in performance management: Guiding entre-
preneurial growth decisions. Journal of Accounting Education (2017), http://dx.doi.org/10.1016/j.jaccedu.2016.12.005