15

CFO TRENDS WITH ANALYTICS

A managerial movement that is now happening and gaining in popularity is the application of business analytics for organisations, which helps them gain insights so they can make good decisions and decide the best actions to take. This topic was once the domain of quantitative analysts and statistical geeks developing models in their cubicles. Today applying analytical methods is on the verge of becoming mainstream.

Which line management function may be in the best position to endorse and drive the adoption of business analytics? Is it marketing, operations or sales? Might it be the finance and accounting function? They already have a propensity for quantitative analysis. It is in their DNA.

One way to draw my conclusion about this emerging movement is to listen to the chatter and debate about the topic. Articles in IT magazines and on websites about ‘Big Data’ and the need for analytics of all flavours, such as segmentation analysis, are increasingly prominent. Debate is always healthy. Some IT analysts view applying analytics as a fad or being way overvalued. Others, such as leading proponents of analytics like authors Tom Davenport of Babson College in Massachusetts and Jeanne Harris of Accenture, claim that an organisation’s achievement of competencies with analytics will provide a competitive edge.

Predictive analytics is one type of analytical method that is getting much attention because senior executives appear to be shifting away from a command-and-control style of management—reacting after the fact to results—to a much more anticipatory style of managing. With predictive analytics executives, managers and employee teams can see the future coming at them, such as the volume and mix of demands to be placed on them. As a result they can adjust their resource capacity levels and types, such as the number of employees needed or spending amounts. They can also quickly address small problems before they become big ones. They can transform their mountains of raw data into information to test hypothesis, see trends and make better decisions.

ANALYTICS AS THE ONLY SUSTAINABLE COMPETITIVE ADVANTAGE

For the last few decades many executives and strategic consulting firms have followed the framework of the popular Harvard Business School professor Michael S. Porter. Porter’s writings include three types of generic strategies, discussed as follows. Notice that with today’s technology-driven markets and economies, each generic strategy is vulnerable:

1. Cost leadership strategy. This is accomplished by improving process efficiencies, unique access to low-cost inputs (eg, labour, materials), vertical integration, or by avoiding certain costs. However today other firms using lean management techniques and data analysis methods can quickly lower their costs.

2. Differentiation strategy. This is accomplished by developing products or services, or both, with unique traits valued by customers. However today there can be imitation or replication of products and services by competitors (eg, smart phones) or changes in customer tastes.

3. Focus strategy. This is accomplished by concentrating on a narrow customer segment with entrenched customer loyalty. However today broad market cost leaders or micro-segmenters can invade a supplier’s space and erode its customers’ loyalty.

So how will an organisation gain a competitive edge? In my opinion the best defence is agility, with quicker and smarter decision making. This is accomplished by achieving competency with business analytics that can provide a long-term, sustaining competitive advantage. It means creating an organisational culture for metrics and analytics.

RESISTANCE TO CHANGE AND PRESUMPTIONS OF EXISTING CAPABILITIES

Some organisations may believe because they hired or trained employees with analytical skills that they have fulfilled the need to be analytical. However there are misconceptions about what analytics really is. To demonstrate this, the following is a true experience of one of my work colleagues.

A large department store retailer accepted a brief meeting with my co-worker for possible clarification about how analytics can increase profit lift from individual customers. The company’s president, chief marketing officer and head of customer analytics attended. They were somewhat impatient because they were confident they already had an effective programme in place because many of their customers used a loyalty card at the checkout counter.

My colleague described that with access to each customer’s profile (eg, age, address, gender etc) and their purchase history, a real-time analytics system could substantially increase the probability that a customer will actually respond to an offer, deal or intervention—and when. The first answer comes from data mining and the latter from forecasting—two of the many components of business analytics.

After the brief presentation with only a few minutes of the scheduled meeting left, the head of customer analytics concluded that the company was already using appropriate techniques. My co-worker then took a risk. The day prior to the meeting he went to one of the retailer’s stores and purchased a travel-size shampoo and toothpaste using his loyalty card. He repeated the identical purchase a second time. In the meeting he placed both receipts on the table and turned them over. One receipt had a discount offer for a feminine hygiene product. The other receipt’s discount was for cat food. My colleague, a male, has no pet. The chief marketing officer asked the head of customer analytics for an explanation. The answer was that ‘those were among the hundred high-profit-margin products that are being promoted this month.’

In this example, there was no true connection to the individual customer, and the checkout register did not have sufficient technology to quickly access in real time customer-specific deals. The three executives had a kind of epiphany. They are now piloting a store entrance kiosk where customers can swipe their loyalty cards and receive personalised discounts and offers.

The kiosk substantially improves sales compared to the checkout register method. What I omitted from this example is how the kiosk knows what specific discount or deal to offer. That requires statistical analysis of different customer behaviours (eg, an online retailer’s feature showing ‘Others who bought what’s in your shopping cart also bought X’). Early results from the retailer’s pilot kiosk programme were substantial. For the checkout register receipt, a 1.8% response rate was the norm. The response rate rose to 30% for the real-time store entry kiosk.

The intended point here is that applying statistical analysis, data mining and forecasting with a goal of optimisation is now in reach—and some organisations that may think they are applying these methods are only just starting to develop them.

It may be that the ultimate sustainable business strategy is to foster analytical competency among an organisation’s work force. Today managers and employee teams do not need a doctorate degree in statistics to investigate data and gain insights. Commercial software tools are designed for the casual user. The accounting and finance function can lead this business analytics movement. However improvement in their skills and competencies will be needed.

EVIDENCE OF DEFICIENT USE OF BUSINESS ANALYTICS IN FINANCE AND ACCOUNTING

Research¹ by the publisher CFO.com reported deficiencies in current uses of analytics in finance. Roughly half of the 231 companies surveyed reported being less than ‘very effective’ at incorporating information for strategic and operational decision making purposes. Thirty-six percent of the respondents identified management intuition and experience as the primary decision criteria when making strategic and operating decisions. Only 17% said that statistical analysis and modelling are primary decision criteria, and the largest number of respondents (36%) reported that their companies make ‘little or no use’ of more sophisticated techniques.

Fifty-three percent of the respondents said that robust modelling and analytics should play a greater role in their organisation’s decision making. Finance executives appear willing to make the kind of investment needed to improve their ability to access and analyse business performance information. Seventy-six percent of the respondents anticipated that their companies will make at least a moderate investment in linking operational data to financial results.

Research by the IBM Institute for Business Value,² reports that the group of surveyed finance functions that demonstrate the highest effectiveness across the entire CFO agenda excels at two capabilities: finance efficiency brought about by process and data consistency, which helps unlock the power of analytics, and business insight to drive enterprise performance. Note the reference to the power of analytics. The study also states that this same group consistently applied five transformation enablers throughout their journey. These are addressing technology, enabling a sequential adoption of standard processes, using new operating models, applying better analytics and improving workforce efficiency. Again note the reference to applying better analytics.

The message from these two studies is clear. Analytics is playing an increasingly more important role with the CFO function. Analytics comes in many flavours, including, but not limited to, segmentation, correlation, regression and probabilistic analysis.

SOBERING INDICATION OF THE ADVANCES STILL NEEDED BY THE CFO FUNCTION

Another research study by The Data Warehouse Institute³ states that ‘finance can be a powerful agent of organizational change. It can leverage the information that it collects to assist executives and line of business managers to optimize processes, achieve goals, avert problems, and make decisions.’ The study goes on to say

forward-thinking finance departments have figured out how to transform themselves from back-office bookkeepers to strategic advisors. They have learned to partner with the IT department—more specifically, the business intelligence (BI) team—whose job is to manage information and deliver a single version of corporate truth. In so doing, they have liberated themselves from manual data collection and report production processes so they can engage in more value-added activities.

The research study further states that the finance function should be empowered to explore data on their own without IT assistance and that ‘armed with analytical insights, the finance department can collaborate with business managers to optimize pricing, reduce inventory, streamline procurement, or improve product profitability. They can help business managers evaluate options, such as whether to add more salespeople, change commission fees, partner with a new supplier, or change merchandising assortments.’

The study makes a sobering statement by saying

Unfortunately, the majority of finance departments have yet to adopt this new role to a significant degree. Our survey shows that although the finance department has made strides toward becoming a trusted partner with the business, it still has a long way to go. Less than half of financial professionals who responded to the survey believe their finance department, to a high degree, helps the organisation ‘achieve its objectives’ (41%), ‘refine strategies’ (35%), ‘drive sales’ (29%), or ‘optimise processes’ (29%). In fact, more than 20% of finance professionals gave their finance teams a low rating in these areas, with a larger percentage saying in effect that the finance department does little or nothing to help the business ‘optimise processes’ (43%) or ‘understand and help drive sales’ (50%).

MOVING FROM ASPIRATIONS TO PRACTICE WITH ANALYTICS

A problem with the research studies referenced is that they describe what the CFO function could be doing with analytics, with some blunt survey results describing the sizeable gap from the possibilities. However they do not provide tangible examples of the vision. Let’s consider a few.

Customer Profitability Analysis to Take Actions

A trend for customers is to increasingly view suppliers’ products and standard service lines as commodities. As a result, what customers now seek from suppliers are special services, ideas, innovation and thought leadership. Many suppliers have actively shifted their sales and marketing functions from being product-centric to customer-centric through the use of data mining and business intelligence⁴ tools to understand their customers’ behaviour—their preferences, purchasing habits and customer affinity groups. In some companies, the accounting function has supported this shift by reporting customer profitability information (including product gross profit margins) using the activity-based cost management principles described in Part 2. However, is this enough?

It is progressive for the accounting function to provide marketing and sales with reliable and accurate visibility of which customers are more and less profitable. Their company can also see why by observing the visibility and transparency of the internal process and activity costs that yield each customer’s contribution profit margin layers. Often, sales and marketing people are surprised to discover that due to special services, their largest customers in sales are not their most profitable ones and that a larger subset of customers than believed are only marginally profitable—or worse yet, unprofitable. However a ranking of profit—from highest to lowest—of each customer does not provide all the information about why. This ranking is illustrated in Figure 15-1. It is a start but without giving all the answers. This is where data mining and analytical techniques can answer the ‘why and so what’ questions.

The use of ABC data leads to activity-based management (ABM). There are some low-hanging fruit insights from ABC data. For example, one can see relative magnitudes of activity costs consumed among customers. There is also visibility into the quantity of activity drivers, such as the number of deliveries that cause activity costs to be high or low. However this does not provide sufficient insight to differentiate relatively high profitable customers from lower-profit or unprofitable customers.

Figure 15-1: Why Are Some Customers More Profitable Than Others?

Source: Created with SAS software. Copyright 2010, SAS Institute Inc., Cary, NC, USA. All Rights Reserved. Reproduced with permission of SAS Institute Inc., Cary, NC.

One can speculate what the differentiating characteristics or traits might be, such as a customer’s sales magnitude or location, but hypothesising (although an important analytics practice) can be time consuming. In attempting to identify the differentiating traits between more and less profitable customers, the major traits may not be intuitively obvious to an analyst. A more progressive technique is to use data mining and advanced statistical analytics techniques. This involves the use of segmentation analysis based on techniques involving decision trees and recursive partitioning. These techniques can give the sales and marketing functions insights into what actions, deals, services, offers, unbundled pricing and other decisions can elicit profit lift from customers.

The goal is to accelerate the identification of the differentiating drivers so that actions or interventions can be made to achieve that high-payback profit lift from varying types of customers. Analysts using ABC/M have benefited from applying online analytical processing multi-dimensional cubes to sort data. Even greater benefits and better decisions can come from applying data mining and advanced analytics.

Rationalising and Validating Key Performance Indicators in a Strategy Map and Balanced Scorecard

How do executives expect to realise their strategic objectives if all they look at is financial results like product profit margins; returns on equity; earnings and interest before interest, taxes, depreciation and amortisation (EBITDA); cash flow and other financial measurements? These are really not goals—they are results. They are consequences. As previously mentioned, measurements are not just about monitoring the summary dials of a balanced scorecard. They are about moving the dials of the operational dashboards that actually move the strategic balanced scorecard dials.

Worse yet, when measures are displayed in isolation of each other rather than with a chain of cause and effect linkages, then one cannot analyse how much influencing measures affect influenced measures. This is more than just leading indicators and lagging indicators. It is timing relationships. As described in Part 3, Chapter 7, ‘The Promise and Perils of the Balanced Scorecard,’ a balanced scorecard reports the causal linkages, and its key performance indicators (KPIs) should be derived from a strategy map. Any strategic measurement system that fails to start with a strategy map or reports measures in isolation, or both, is like a kite without a string. There is no steering or controlling.

In Chapter 7 we learned that the strategy map’s companion scorecard, on its surface, serves more as a feedback mechanism to allow everyone in the organisation, from front-line workers up to the executive team, to answer the question: ‘How are we doing on what is important?’ More importantly, the scorecard should facilitate analysis to also know why.

To go one step further, a truly complete scorecard system should have business analytics embedded in it. An obvious example would be correlation analysis to evaluate which influencing measures have what degree of explanatory contribution to influenced measures. Imagine a balanced scorecard’s strategy map in which the thickness of the KPI arrow reflects the degree of explanatory contribution. That is how analytics embedded in a methodology brings more value. With KPI and PI correlation analysis, scorecards and dashboards become like a laboratory to truly optimise size and complexity. Consider that the thicker arrows (ie, higher correlation) could mean to provide greater budget funding because those levers appear to drive higher results of other KPIs.

MOVING FROM POSSIBILITIES TO PROBABILITIES WITH ANALYTICS

What could possibly affect an organisation’s performance results? At the operational level, sales order volume could be up or down. Prices of purchased commodity materials like steel or coffee could be up or down. On a strategic macroeconomic level, consumer demand could be up or down. From a risk management perspective, weather fluctuations could adversely affect the best laid plans.

How could you know the impact, including the financial impact, since these ranges of possibilities occur at various levels? There are three broad ways: a single best guess; the worst, baseline and best likely outcomes; and a probabilistic scenario of the full range of outcomes. They all include predictions with analytics.

1. Single Best Guess. Most organisations plan for results based on their manager’s best assumptions of what they estimate. For example, in the annual budgeting exercise, managers forecast sales mix volume, labour rates and prices of purchases. Each is a single point estimate, and the accountants aggregate them to produce a single budget.

2. Worst, Baseline and Best Likely Outcomes. The more advanced organisations consider three ranges of outcomes: worst, baseline and best likely outcomes. Separate predictions are made for the key variables in the plan. Then the three overall possibilities are calculated. This provides a sense of the range of outcomes. These organisations might individually test the sensitivity of the key variables by increasing or decreasing each of them—one at a time—to se e the effect and outcome.

3. Multiple Probabilistic Scenarios. The most advanced organisations take this process to its ultimate limit, from three scenarios to the full range of possibilities. That is, they estimate the probability distribution of each variable, perhaps as percentage increments from the base (eg, -20%, -10, 0% base, +10%, +20%). By combining these, they move from the three single point outcomes to viewing a distribution curve of dozens and, conceivably, hundreds or thousands of outcomes. The benefit is they can have more certainty of the increasingly uncertain world they operate in. In addition, the variables become understood as drivers of the results in which the level of each one may be able to be proactively managed in advance of their occurrence. These three levels are illustrated in Figure 15-2.

Figure 15-2: Analytics: Probabilistic Planning Scenarios

Source: Copyright Gary Cokins. Used with permission.

The breadth and granularity of the distribution curve increases as the probability ranges for each variable is segmented, as more variables (not just the key ones) are added, and as each variable is sub-divided (eg, from a product family to its individual products). The three-scenario approach gives a limited view of risk in contrast to the multiple probabilistic distribution curve. With the latter, sensitivity analysis can become very refined, including automated increases and decreases of each variable to determine which variable drivers have more impact.

Now take this process to an even higher level by increasing the time interval frequency of re-forecasting one or more (or even all) of the variable drivers.

What influences the accuracy and quality of the distribution curve? A critical one is the forecasting of each variable. If the baseline is really far off, then incrementing it up or down is also going to include error.

Achieving this best practices approach requires a combination of advanced analytics, reliable forecasting techniques (eg, Monte Carlo methods) and a powerful computational software engine. If this is supplemented with robust reporting, visualisation and analytical power, then it is nirvana. The full range of probabilistic outcomes can be viewed and, at more frequent time intervals, approach near real time. The benefits are endless. Risk management becomes scientific. Rolling financial forecasts replace static and fixed-in-time annual budgets. Drivers can be proactively managed, such as supply chain logistics and inventory management.

Predictive analytics is becoming a hot term with enterprise performance management. With the opportunity to move from just discussing the possibilities to understanding the factors affecting an organisation and also taking actions based on the interdependent probabilities, is anyone surprised? The organisation shifts from possibility to probability—managed probability—of outcomes.

FILL IN THE BLANKS: WHICH X IS MOST LIKELY TO Y?

Business analytics allows organisations to make decisions and take actions they could not do (or do well) without the analytics capabilities. The finance function can assist its line managers and employee teams. Consider these examples:

Increased Employee Retention

Which of our employees will be the next most likely to resign and take a job with another company? By examining the traits and characteristics of employees who have voluntarily left (eg, age, time period between salary raises, % wage raise, years with the organisation, etc), business analytics can layer these patterns on the existing work force. The result is a rank order listing of employees most likely to leave and the reasons why. This allows managements’ selective intervention.

Increased Customer Profitability

Which customer will generate the most profit based on our least effort? As just described, by understanding various types of customers with segmentation analysis with recursive partitioning based on data about them (and others like them), business analytics can answer how much can optimally be spent retaining, growing, winning back and acquiring the attractive micro-segment types of customers that are desired.

Increased Product Shelf Opportunity

Which product in a retail store chain can generate the most profit without carrying excess inventory but also not have periods of stock outs? By integrating sales forecasts with actual near real time point-of-sale checkout register data, business analytics can optimise distribution cost economics with dynamic pricing to optimise product availability with accelerated sales throughput to maximise profit margins. Mark-down prices of inventories to be abandoned can be optimised.

These are three examples of the contribution that business analytics can provide. How can an organisation determine which X is most likely to Y? Hundreds of other examples exist in which the goal is to maximise or optimise something. With business analytics, the best and correct decisions can be made, and organisational performance can be tightly controlled and continuously improved. Without business analytics, an organisation operates on intuition, and optimisation cannot even be in that organisation’s vocabulary. The CFO function has the competencies involving quantitative analysis. It is in their nature.

THE CFO FUNCTION NEEDS TO PUSH THE ENVELOPE

Research by Ventana Research has confirmed that the gap between current and potential use of analytics remains wide. It reports that information technology should be a particular focus because most finance organisations are not using IT assets as intelligently as they could. In particular, the finance function often focuses only on efficiency and neglects opportunities to use IT to enhance their effectiveness. Finance functions have made considerable progress in addressing their basic information needs (referred to as 20th century reporting requirements), but most are a long way from providing the more complete, next level of information that can be used to improve performance (their 21st century requirements).

The benchmark research of this study shows there is important information that employees could receive—or already do receive—that would improve their organisation’s performance and help align its actions to strategy. Information deficits, combined with poorly designed processes, can severely limit how well all departments, including the finance function, do their jobs. The study’s recommendations are that CFOs and senior finance department executives focus on these three areas:

• Push the envelope when it comes to management reporting. To improve performance, companies must link more operational and financial data, make information available sooner and provide a richer set of data, including leading indicators for the business unit and relevant information about competitors, suppliers and factors that drive demand for the company’s products or services.

• Have a disciplined, sustained process in place to address information technology barriers, especially infrastructure complexity, and to enhance the use of enterprise resource planning (ERP) systems. Typically, these are the root causes of issues preventing finance organisations from improving process implementation and preventing the disconnections that obstruct better alignment of strategy and implementation.

• Assess where there are shortfalls in the people, process information and technology dimensions of key financial functions, then define a plan with specific objectives and timetables that addresses these shortfalls.

Pursuing the application of analytics is common sense. One could argue that this study omitted as a root cause barrier the natural resistance to change and preference for the status quo. Without analytics, insights and understanding for better decision making is limited.

Endnotes

1 Gearing Up for Growth: Financial Analytic Capabilities for Changing Times. CFO Research Services, May, 2011. Changing Times is published by CFO Publishing LLC.

2 Journey to a Value Integrator, IBM Institute for Business Value, www-935.ibm.com/services/us/gbs/thoughtleadership/ibv-journey-to-new-value-integrator.html.

3 Transforming Finance: How CFOs Use Business Intelligence to Turn Finance from Record Keepers to Strategic Advisors. The Data Warehouse Institute, First Quarter, 2010.

4 Data mining is the process of extracting patterns from large amounts of stored data by combining methods from statistics and database management systems. It is seen as an increasingly important tool to transform unprecedented quantities of digital data into meaningful information, nicknamed business intelligence, to give organisations an informational advantage. It is used in a wide range of profiling practices, such as marketing, surveillance, fraud detection and scientific discovery.

5 Financial Performance Management in the 21st Century—A CFO’s Agenda for Using IT to Align Strategy and Execution. Ventana Research, 2007.