• November 1, 2007
  • By Colin Beasty, (former) Associate Editor, CRM Magazine

Predicting Profitability

Article Featured Image
Profitability is the ultimate determinant of a company's success or failure. Yet as crucial as this measure is, it remains a mystery to many management teams, especially at the level of customer profitability. Those assessments are only as good as the metrics or methodologies you're using to measure them with, and those have historically differed from department to department, from company to company, and from industry to industry. Enter the new age of profitability analytics. As customer-centric strategies and their accompanying software solutions have gone enterprisewide, so have profitability analytics -- though now retooled and redeployed to meet the needs of all employees. One major advance is the ability to go beyond old data and to look ahead, allowing enterprises to take a more predictive look at which customers will be their most profitable going forward. "Profitability analytics have come a long way, and companies have hit plenty of roadblocks," says Gareth Herschel, a research director at Gartner. "But now the boundary is being crossed, and companies are learning how to truly leverage profitability analytics." Finance...Meet Marketing Early customer profitability systems bubbled out of the finance department, but as businesses began to embrace customer-centric strategies, marketing embraced profitability analytics as the vehicle by which to measure success. That's the friction point inherent in any enterprisewide profitability strategy. Finance has traditionally focused on cost-and-revenue analysis, with the end goal of assigning profit and loss (P&L) across multiple lines of business and allocating revenue across different product lines or services, typically relying on data from general ledgers and ERP systems. On the flip side, marketing has taken a more customer-centric focus, looking to segment an organization's most profitable customers via CRM data to develop campaigns and identify cross-sell and upsell opportunities. But as finance has begun to aggregate and analyze profitability data on a more granular level, and as the marketing and sales departments look to make good on their growth and churn predictions for the coming year, the goals, methodologies, and agendas of disparate departments collide. As a result, the company is caught in an analysis paralysis. "It's the convergence of two worlds," says Jeff Lovett, director of finance and performance management for Teradata. "Today, organizations want the big picture, the corporate performance management picture, so to speak. The problem becomes, 'Whose methodology do I use?' " The resulting fallout can land with a resounding thud in corporate board meetings, and leads to a top-to-bottom approach in which corporate is attempting to appease all the hands in the profitability cookie jar. "It happens all the time," Lovett says. "Finance is snickering and writing notes because they don't believe in the methodology, and vice versa. The result is key metrics are sidetracked and business opportunities are lost." The solution to this mayhem, and the way to keep all parties satisfied, is to take a bottom-to-top approach, Lovett says. A perfect example of this comes courtesy of the financial services industry, which typically struggles to allocate costs across millions of customers and thousands of varying transactions. Historically, banks have allocated the revenue generated from prepaid interest or mortgage rates to a specific date or time. While technically correct, the result means that customer profitability models can be thrown off and predictive analytics put into disarray. "If you want to understand the customer, the right way to allocate that $3,000, $4,000, or $10,000 is over the life of the loan," Lovett says. "That way, if marketing wants to query the amount of revenue that customer has generated over the previous six months, they can garner an accurate profitability snapshot. If finance needs to look at product profitability, they can. By driving your methodology from the lowest possible level and storing the data that way, you're providing a common denominator that meets the needs of all departments, and allowing that profitability data to be rolled up in multiple ways." The result is a bulletproof metric or series of metrics that the entire organization can buy into, says Jeff Gilleland, global strategist for SAS Institute's customer intelligence products. Instead of the corporation managing thousands of different pricing levers on the proverbial switchboard that is profitability analytics, corporate is managing just the big ones, allowing each department to drill down into the specific metrics inherent with its line of business. "Your independent departments might be calculating profitability at a very granular level, and your business might end up with 10,000 different cost metrics, but as the CFO or CMO, you don't need all of that. You're just going to focus on those five or six that allow you to do your financial planning and forecasting, and that give your company a competitive differentiator," he says. By tying any results back to the transaction level, the bottom-to-top approach also helps to ensure credibility of not just the metrics but also of the methodology used to calculate them. In essence, the organization reaches the apex of credibility, says Tony Adkins, product manager of SAS profitability products. "The credibility and accuracy comes when you can tie your activity-based costing models back to the transaction level, at the point of revenue," he says. "If you can do that, it's a very credible way to show other departments within your organization -- and your customers -- the reason for a business change, whether it's a price increase or an adjustment to next year's sales forecast." Analytics Today for Profitability Tomorrow
The science of actually measuring customer profitability isn't difficult, and usually boils down to analyzing the transactional and behavioral information so generously stored in a CRM system. Solutions from vendors such as SAS, SPSS, Teradata, Hyperion (now part of Oracle), and others automate many of the mind-numbing tasks associated with these calculations, and allow marketing, finance, and sales to group customers into a range of profitability segments, including product and channel preferences, P&L, account history, and more. But while targeting customer segments that have been historically profitable is all well and good in the short term, it's only a matter of time before the other dollar drops, says Colin Shearer, vice president of customer analytics at SPSS. As a result, leading organizations are using the data garnered from profitability solutions to help solve for "P" in the predictive analytics equation. "Your customer base might include a lot of struggling young medical students who are poor, but boy, down the road are they going to make money for you," Shearer says. "It's about being able to apply models that identify which customers are worth investing in because you believe if you can keep them happy for another five years, they'll become incredibly profitable." Today, leading organizations are using profitability modeling at the transaction level, determining campaign and transaction triggers and real-time decision-making to assess a customer's lifetime value. To that end, the customer-value metric is emerging as a centerpiece to measuring customer profitability at the transaction phase because it can account for profit cannibalization that sometimes occurs from one product or service to the next. Gilleland offers financial services as an example yet again: Typically, three out of every four cross sells in that industry actually destroy profitability. "A service agent might simply be trying to resolve a high balance on a credit card, or getting solicitation for an equity line of credit, and in the process, the bank migrates the customer's same balance from a high-margin to a low-margin product," he says. "When you factor in the cost of the call, administrative costs, etc., the net impact is you've destroyed that customer's profitability." In response, Shearer says, companies are letting profitability and predictive modeling trickle down into the operational segments of their business, namely the contact center and marketing. In the contact center, some enterprises are starting to run these models at the moment the interaction is taking place, providing recommendations to agents in real time, and letting those models determine the strategy to drive customers to specific channels, such as onto the Web or into the store. On the marketing side, the newer profitability solutions are permitting marketing departments to better understand (and utilize) customer profitability and propensity before running the campaigns, with the newer, simplified user interfaces acting as "a layer of business insulation between them [the marketers] and the hardcore analytics at the heart of it," Shearer says. "It's bringing profitability analytics to the masses." Smart companies are taking the bottom-to-top approach to drive strategic decision-making. By analyzing the aforementioned "bulletproof" metrics, businesses are redefining their client value measures and making strategic investments in certain customer segments to ensure future profitability. "You're streamlining the decision-making and reporting for marketing and sales initiatives with an accurate and concise assessment of client portfolios and their values," says Kevin Purkiss, manager of customer analytics at RBC Royal, a $294 billion Canadian financial services firm, where leveraging customer data to drive profitability is nothing new. The company was one of the first to implement data warehouses, and by the early 1990s was letting Teradata power its data quality initiatives. Following a study of its customers, the company realized it needed a powerful analytic application to leverage that data, and to score and manage its 11 million personal and commercial clients by value. To further refine its service and product offerings, cost management, pricing initiatives, and marketing programs, RBC Royal upgraded its existing profitability models so it could gain a more granular measure of client profitability. After a lengthy evaluation of third-party products and vendors, RBC Royal decided to stick with the vendor that had laid the foundation of its analytic initiative and chose Teradata Value Analyzer. By adding Teradata Value Analyzer, the bank could move from product portfolio reporting to client segment reporting. Initial design to full operation was completed in just over six months, which included defining business requirements and an extensive data-verification process. Using Value Analyzer, RBC Royal's 10 million retail customers are grouped into discrete segments based on such factors as current and potential profitability and channel preferences. Strategies are then developed for each segment, as well as for subsegments. Individual treatment strategies are tested on small cells of clients to establish what works and what doesn't. "We recalculated customer profitability using the account-level metric produced by Value Analyzer, which showed that 75 percent of our customers moved two or more deciles," Purkiss says. Taking things one step further, RBC Royal integrated the application of client-value metrics into the day-to-day activities of the front-line staff, enabling the bank's client sales and service system to proactively pursue opportunities with clients. And through the use of current and historical client-value measures, Royal Bank assesses client segmentation at a more granular level, by taking into account factors such as life-stage changes. "From current value, we have progressed to potential value," says Ted Brewer, vice president of CRM and information management at RBC Royal. "In this way, we can manage a relationship, actualizing CRM by ensuring the right product is available at the right time in the customer's life." And that's showing in the bank's direct marketing campaigns, where response rates are up -- reaching as high as 40 percent at times -- compared to an industry average between 2 percent and 4 percent. For its Retirement Savings Plans (RSP), RBC Royal used its CRM solution to execute a targeted marketing program that led to an 11 percent increase in RSP deposits. The bank also used Value Analyzer to determine that 90 percent of its Royal Certified Service accounts were unprofitable by drilling into the transaction data and seeing that a large portion of customers' account transactions were higher-cost bill payments at ATMs. Rather than raising the package prices, the bank introduced new service channels, including interactive voice response and Web self-service at no extra charge to the customer. Over the next two years, most clients migrated their bill payment methods to a channel that was more convenient for them -- and more cost-effective for the bank. Price Optimization: Part of the Analytics Family One of the technological offspring of profitability analytics -- and an underlying cornerstone of any sound profitability strategy -- is price optimization. Like profitability analytics, price optimization strategies and their emerging software solutions have expanded their reach enterprisewide, touching upon multiple departments and reaching to the highest and lowest levels of the business. 3 Components of Pricing Model
The Three Components of the Pricing Model: Many organizations choose to focus on one or two of the three components and can still achieve valuable results without having to tackle the entire process. Originally born out of the transportation and airline industries, price optimization strategies and solutions have developed concepts and expanded their functionality to cut a broad swath across a multitude of verticals. The importance of price optimization, albeit due to illegal actions, was brought to light in August of this year, when British Airways and Korean Air were fined $300 million apiece after admitting they conspired to fix prices on international flights. Both pleaded guilty to antitrust conspiracy charges after admitting that they colluded with rivals over cargo rates and fuel surcharges. Fares were increased in response to rising oil prices, leading to higher costs for international shippers and passengers. Today's price optimization solutions are a far cry from the pricing books that sales and service reps have traditionally used over the years. Despite the emergence of these new tools, the market is still a maturing one, and most companies today do not fully understand the influence that pricing has on customer profitability, says Jon Utterback, a specialist leader in Deloitte's pricing center of excellence. A recent study by the Business Performance Management Forum (BPMF) supports this, finding that despite the implementation of big-ticket technology infrastructures, the majority of respondents make high-level pricing decisions and revenue forecasts based on internal data generated and communicated through outdated spreadsheets. According to Diganta Majumber, director of the BPMF, one of the executives surveyed talked about the significant dollars his company had spent on technology, but likened the forecasting capabilities to that of an abacus. Close to 40 percent of respondents said they could boost company revenue by double digits with better data and improved analysis and forecast alignment. This realization is turning into hard dollars for a market that is bursting at the seams. According to Rob DeSisto, vice president of CRM at Gartner, "through 2009, price optimization technology will have the greatest impact on improving the top-line revenue and profitability of any business application." The recent upsurge in the adoption of price optimization solutions is due to the availability of abundant CRM data coupled with the improvement in data quality practices, which is laying the foundation for success. "Organizations are becoming more comfortable using data to make automated decisions," DeSisto says. "It's all about the data: Dirty data in equals dirty data out." The traditional disconnect between pricing and CRM was a result of siloed data, Utterback says. CRM's strengths in lead generation, contract renewal, lead qualification, and order processing have been well integrated, but the price development and quotation approval necessary for a sound pricing strategy have typically been handled outside of CRM systems using pricing books, Excel models, and manual quotes. "I've had clients crunch numbers in Excel and then actually write the price down on a piece of paper during a sales meeting," Utterback says. "It's the most unsophisticated way to draft a proposal." The applications themselves parallel the functionality inherent in profitability analytic solutions, albeit at a more specific level, and allow marketing managers and sales reps to exploit CRM-based transaction history, customer behavior, and market information. That data can then be cross-referenced against metrics such as product data, historical pricing models, rebates, trade promotions, freight and terms costs, invoice fees, and debts to optimize profit margins, measure customer elasticity, and to deliver the optimal combination of price versus related costs. Companies can refocus their sales efforts on highly profitable customers who are willing to pay a premium price, capturing as much "share-of-wallet" as possible with price-insensitive customers, Utterback says. The alternative is to increase price or eliminate unprofitable volume sold to smaller, lower-value customers who receive better prices than larger, higher-value customers. While it's smart to think of price optimization as a complement to (and tactical implementation of) profitability analytics, its scalability and strategic implications cannot be overlooked, Utterback says. This is especially true now that these solutions are coming with embedded predictive capabilities, allowing corporate to determine pricing strategies based on metrics such as product line and product/SKU metrics, and allowing C-level executives to plan sales forecasts. "These solutions have broad implications for both managerial and corporate employees alike," Utterback says, though he warns, just like profitability solutions in general, the key lies in the methodology, not in the technology. "The key to determining who your most profitable customers are lies with which metrics you're measuring and what methodology you're using. Profitability analytic solutions are simply the vehicle to do that, and will help get us there."
CRM Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues

Related Articles

Welcome to the Customer Data Revolution

Predictive Analytics World '09: A former Amazon.com executive shares the steps required to manage today's customers.

5 Ways to Cut Costs with Predictive Analytics

Predictive Analytics World '09: With a goal of operational excellence, you can target the right customers and save money along the way.

SAS Institute Gets Some Bright IDeaS

The analytics and statistical analysis powerhouse acquires an on-demand price optimization vendor -- and builds a new practice around it.

Buyer's Guide Companies Mentioned