Inquiring marketers want to know. Would you be willing to pay more for flashlight batteries when a winter storm is forecast? Would you pay more for Duracell after seeing its latest ad on TV? And would you put a package of CVS private-label AA batteries in your basket when you’re picking up your prescription?
The quest to determine optimal pricing is hardly new. But its importance is rising rapidly in many sectors as cost cutting to lift margins runs out of steam and as new analytical tools and techniques become part of the pricing conversation.
What’s got the attention of senior executives is the shift from descriptive to predictive analytics.
Descriptive analytics—already in wide use in many ad hoc ways—looks at historical data, helping companies answer such basic questions as what happened, why it happened, and how much it helped or hurt results. Predictive analytics extends those findings using sophisticated statistical modeling, forecasting and optimization algorithms to anticipate the impact of various actions, such as promotions, price changes and advertising, on business outcomes. Predictive analytics is about more than simple linear what-if exercises. It enables complex, dynamic research with multiple variables and often involves flurries of concurrent, low-cost, fast-turn market experiments. Forecasts become more about facts than about hunches (see sidebar).
In fact, there is already an upsurge in interest in predictive analytics in the consumer packaged goods sector, among leading retailers and at top businesses in both the B2C and B2B sectors. These companies see that reviewing past performance is no longer enough to be truly competitive.
Despite the excitement around predictive analytics, executives understand that it is not easy to build analytics capabilities. Yes, at a basic level, businesses can derive some benefit by deploying business intelligence solutions that cover standard reporting, queries and alerts. But to properly harness predictive analytics, companies must think and act in terms of integrated analytics approaches—developing a suite of analytics capabilities, being clear about which metrics to track and which experiments to run, and building governance mechanisms to help ensure that the analytics activities are aligned with the business strategy.
Those approaches cannot be implemented overnight. Executives have to think and act in terms of a suite of analytics capabilities, starting with cleansed, homogenized data that reveals “one version of the truth” (see chart).
Many companies are not yet ready to apply full-strength analytics to pricing, which, compared with other business activities, has been slow to take advantage of analytics developments. In general, pricing is strongly associated with sales, with marketing providing guidance and finance lending support. Few organizations have a dedicated chief pricing officer or executive for whom pricing is a primary responsibility. (This is not necessarily the case in retail, where the merchant role is wholly responsible for initial pricing and the equally important markdown pricing strategy.)
Furthermore, pricing has not always been linked strategically to growth. Accenture’s latest research found that fewer than 29 percent of chief marketing officers and chief financial officers claim to use pricing to achieve their growth objectives.
|The challenges have been more marked in the B2B sector, where pricing has often been a game of chicken in which discounts are offered at the first sign of competitive pressure or indifference from buyers. In such circumstances, salespeople are all too often negotiating with purchasers who come to the table bristling with data on everything from historic price trends to contract terms. Sales teams that lack hard data to push back often buckle rather than risk losing the business.
Click to Enlarge
Despite the challenges, rapid developments in analytics are converging with the uncertain economic recovery to force pricing into the spotlight in a widening range of industries. Indeed, financial executives responding to a recent Duke University/CFO Magazine survey still rank price pressure as a top concern. Most companies have made all the usual performance improvement moves—restructurings, capital spending freezes, SG&A spending cuts and more. So something else is needed.
The sheer complexity of offerings—more SKUs, more product/service and product/product bundles, and more co-branding initiatives—adds pressure for fact-based pricing tactics. So, too, does the array of distribution channels.
At the same time, analytics software is becoming more capable and more accessible. The skills and experience needed to use those tools are becoming increasingly prevalent. More companies are getting control of their data. And high performers have invested more aggressively in data quality assurance and master data management technologies, giving them reliable, consistent information about customers, products, employees and suppliers.
Pricing steps up
There is growing momentum behind the shift from descriptive to predictive analytics in pricing. Retailers have been in the forefront; the shifts in pricing analytics allow them not only to understand the impact of their price changes in the past but to predict what will occur if they do change prices. Better pricing using granular analyses often makes the difference between whether a retailer simply survives or leads its industry.
One US auto parts retailer uses software that analyzes the price elasticity of individual items in each store to identify where volumes can be raised by strategically cutting prices. Conversely, the retailer keeps a list of items it can use to protect and enhance margins. By combining pricing elasticity analytics with store clustering techniques, the company can understand the difference in price tolerance by individual market.
In more competitive markets, it can aggressively reduce prices to maintain or grow share; in less competitive ones, it can preserve margins. Predictive price analytics can help assess the impact of individual price actions before they take effect, whereas in the past, the retailer would have run time-consuming tests in isolated markets—tests that may have become irrelevant by the time the results were analyzed.
In the airline sector, pricing has long been based on sophisticated pricing and yield algorithms that predict how to optimize prices based upon many factors, from day of the week to size of the aircraft. And of course web-based analytics enable e-tailers like Amazon.com to continually run pricing experiments to optimize and fine-tune prices.
In a small but growing part of the B2B world, too, analytics are at the heart of pricing initiatives. At one international aerospace and high-tech manufacturer, predictive analytics are being used to model potential shifts in key variables like raw materials prices and manufacturing defect rates. These analyses are incorporated into future bids and programs in the form of additional risk mitigation items or justifiable cost contingencies.
The pioneers continue to push the frontiers by, for instance, linking price-change analytics to supply chain forecasts. They are also taking advantage of more accessible predictive analytics solutions—using managed services and software as a service, for example.
But there is still enormous potential for business to harness predictive analytics for competitive advantage.
As business leaders start to grasp the allure of predictive pricing, they are asking many more questions. What insights will help us defend against private-label brands? How do we improve our analytics capabilities? What do we need to do to improve the rate at which we “test and learn” about pricing optimization? How can we turn pricing insights into action day in and day out, with process discipline to make our pricing stick?
Those questions are not easily answered when pricing is everybody’s and nobody’s chief responsibility. There is no one blueprint for how an organization should elevate its pricing performance. However, Accenture’s studies of analytics best practices across a range of industries reveal five fundamentals that make a difference.
1. Use the abundant data at your disposal—and sift it for insights
Effective pricing is fact-based. You may not have exhaustive detail on every transaction for every product in your portfolio, but chances are that you have a lot of what you need, whether from your sales force, from your distributors or directly from your customers. This data should be an essential input in setting prices.
One pharmaceuticals giant leveraged its access to unprecedented volumes of transactional prescription data to mine new insights into pricing its products. Before embracing this approach, the company, like many others in its industry, assumed that it could not drive greater consumer market share by using different pricing structures. The assumption was that pricing was already optimized by insurance providers and therefore completely out of the company’s control.
After applying advanced analytics to overhaul its pricing activities, the company was able to quantitatively test the impact of a range of factors that determine its customers’ willingness to pay. Using descriptive analytics, the pharma discovered that customer age and geography were the leading determinants. In fact, the insurance providers had little actual impact on pricing.
Once the company understood the behavior of its various customer segments, it was able to use predictive analytics to determine what each segment’s response to different price points would be. Capitalizing on this information, it tailored its insurer negotiations and direct-to-patient incentives to provide greater coverage to the most promising customers. Not only did this greatly increase the number of patients who bought their medicine as prescribed by their doctor—as opposed to a generic or lower-priced medicine from a competitor—it also significantly reduced the number of ineffective promotions.
Nowadays, there is no shortage of price optimization software available to help sift through data quickly and gain an understanding of the price elasticity of various products. At some point—sooner rather than later—it will be necessary to integrate data across divisional and product silos. And it will be important to ensure that pricing strategy incorporates detailed cost and cost-to-serve data by channel and customer segment to drive the optimization of margins in each segment in which the company participates.
To be sure, using predictive pricing analytics, and basing many more decisions on data, may force you to acknowledge areas where your expectations have been unrealistic. This is to be expected. It’s a rare company that doesn’t have any misconceptions about its products. Looking at the facts takes you out of the domain of hoping and puts you into the domain of statistical confidence.
2. Fit your prices to your strategy—and enforce the rules that turn strategy into action
Essentially, there must be an explicit pricing strategy that is linked to the overall marketing strategy and driven by a deep understanding of customer needs and attractiveness and of the firm’s competitive positioning.
A few leading companies seem to have pricing analytics in their bloodstream. Capital One, the financial services provider, uses data analysis to differentiate among customers based on credit risk, usage and other characteristics, and to match customer characteristics with appropriate product offerings. The company uses a series of experiments to determine the right interest rates and terms—essentially determining the right price for the level of risk.
Strategic pricing decisions can break down into very fine subsegments—within the online sales channel, for example, or by product type or customer. A leading specialty retailer, for instance, priced its products based only on the price sensitivities of its best customers. This deliberate choice screened out the bargain hunters—the unprofitable customers the retailer did not mind losing.
Pricing decisions for a particular product should also change at different points in the product’s lifecycle. And the decisions will change depending on whether your goal is market-share growth, revenue growth or profitability.
A well-aligned pricing plan for a particular product can even work in reverse, helping you spot a miscalculation in strategy. For instance, you may have pegged a particular product for hypergrowth and priced it accordingly. However, if retailers and consumers aren’t responding the way you expect, it may be time to reconsider your strategy for the product and how you’re pricing it.
Making it stick
Of course, it’s one thing to apply analytics and come up with optimal prices; it’s quite another to make sure those prices get enforced. Unclear or unpublished rules on discounting, inappropriate incentives and a poorly trained sales force (including one with less information than the buyers it’s negotiating with) can all undermine your attempts to set a price and have it stick.
Recently, responding to a decline in margins, a building products manufacturer examined the contracting behaviors of its sales force. The company’s strategic intent was to ensure that discounts and service levels always rewarded its most valuable customers.
What the company found, using in-depth descriptive analytics, was that certain sales offices were offering the lowest prices to their high-cost customers simply to meet revenue quotas. By instituting more disciplined, data-driven negotiation guidelines and incorporating margin into the salespeople’s compensation structure, the manufacturer was able to stem this behavior and even recover some lost customers.
Critically, the company also used sophisticated predictive analytics to calculate the risk of pricing particular products and deals higher. As a result, it was able to manage the risks of revenue and profit loss associated with stricter pricing guidelines. The company now estimates that its more strategically aligned approach to pricing has increased profits by more than $100 million.
3. Don’t wait for top management: Realize that pricing experiments can start anywhere
Pricing is one area where operations-level managers can achieve a lot by themselves. Indeed, Accenture’s research shows that decisions about pricing are already made most often at the business unit level. There is no reason why business unit managers can’t take an even bigger role in pushing toward more systematic pricing.
At the same time, pricing improvements can be instituted in waves. It’s counterproductive to pursue everything from Day One. Instead, identify the product categories that provide the majority of your profits and growth potential, and start with them.
So it’s best not to wait for executive management. Any senior manager can set a positive example by making pricing decisions based upon data and careful analysis instead of on assumptions and conventional wisdom. Senior managers can also clearly define pricing processes, roles and responsibilities within their spans of control.
A leading home improvement retailer recently found many of its most innovative pricing approaches coming from its category managers, not from its executive team. The plumbing department, in particular, designed groups of test markets for various products and goals.
Team members would routinely analyze the impact of price changes and develop measures of elasticity to predict customer responses to price shifts. Based on the test results, the team would roll out the price shift to the entire chain or readjust to discover the optimal price point. Of course, while category managers were still responsible for their category’s financial performance, the results of their testing environment more than met the goals of their leadership, increasing margins as much as 3 percent within the first year.
4. Pick your spots with maximum impact in mind
Monolithic approaches will not work: The price that worked for you when you first came to market may not work later in the product lifecycle. For that matter, the price that works in Russia may not work in Ukraine, even if your strategy in the two countries is identical. If it isn’t identical—if you’re trying to maximize profits in Russia and gain a foothold in Ukraine—that’s another reason, apart from country differences, why the prices may diverge.
Best-practice companies are disciplined about their incentives, workflows and selling processes. They practice good governance as they manage deals. They focus on analyzing and predicting where they can raise prices, where they need to be aggressive on price, and how they can use non-price factors in combination, to win deals. The right rules will help eliminate process and decision-making barriers between divisions, products, channels and geographies, thus driving optimal results across the enterprise.
For complex companies with broad product portfolios in highly competitive markets, being flexible isn’t an option; it’s a necessity. Analytics can help them apply pinpoint pricing to drive different outcomes in different places and at different times.
Starting with the area of greatest opportunity is common sense. All else being equal, you would concentrate on a product category that accounts for three-quarters of your sales before one that accounts for 10 percent of sales. You would fine-tune pricing for the products you distribute through Walmart before doing so for the products you distribute through a mom-and-pop shop in Maine.
Figuring out where to begin your pricing work requires a high-level analysis. But that doesn’t mean you can’t invest in lots of small concurrent pricing experiments. The beauty of today’s pricing optimization tools is that they permit rapid test-and-learn cycles at low cost (see sidebar).
5. Start lining up the right levels of analytical talent
Few companies manage analytical talent as a strategic resource. Many companies don’t even have a clear picture of who their analysts are, how many they have or where they reside organizationally. They certainly don’t recognize or manage them as a distinct and pivotal workforce segment that requires its own recruiting strategies, training and development plans, career paths or performance management processes.
At leading-edge pricing companies, the picture is very different. We are seeing a trend toward directors and vice presidents of pricing who recruit lead analysts with doctorate degrees. These professionals are products of college and university business programs where rigorous courses are dedicated to pricing theory and practice. They typically belong to professional societies dedicated to honing analytic skills in pricing.
At one prominent US industrial manufacturer, the head of the pricing organization is a senior executive vice president and a member of the board of directors. At several leading private equity firms, there is an analytics team whose specific role is to improve the pricing practices of the firms’ portfolio companies.
Accenture’s longtime studies show that the most successful companies create a high-performance analytical organization by building four key talent management capabilities. They start by defining analytical talent needs. They search for and find new sources of analytical talent. They develop their analytics experts, identifying the skills they need and how those skills must be built. And they successfully utilize their analysts, figuring out and creating the best possible match between analysts’ skills and the demands of the business.
What distinguishes talent-powered analytical organizations isn’t just the quality of their analytical talent; it’s their ability to unleash their analysts’ talents to maximize and continually expand the company’s analytical capabilities.
Seasoned business executives don’t need to be reminded that pricing has a disproportionate impact on both top and bottom lines. But they do need to recognize that analytical tools, techniques and competencies are starting to come together to transform pricing from a collection of sporadic, tactical and defensive activities to a robust, strategic and assertive discipline that can continually create value.
This is not about bits and bytes. The data and the technologies that turn that data into insights are necessary but not sufficient. It’s critical to have a vision and an operational plan to change the way you think about and execute pricing.
No matter how good a company’s analytic capabilities, it still must institutionalize an ability to bring insight to action at speed in order to create value. The fact-based statistical component of analytics is only half the battle. Driving decisions and actions based on those insights is arguably the hardest part of the battle.
In short, business leaders who grasp the value of predictive analytics can stop looking over their shoulders to see what their competitors are up to—and start taking charge of their own pricing destinies.
And that’s something their shareholders will most certainly want to see.
For further reading
“How to turn data into a strategic asset,” Outlook, June 2010
Sidebar 1 | From descriptive to predictive analytics
Descriptive analytics. This discipline, also referred to as “business intelligence,” enables fact-based decisions in a reactive or historic way. Descriptive analytics can help deliver a report every Monday on how many of a cell phone company’s customers churned or how many insurance claims were filed for a specific medical procedure. When it comes to pricing, descriptive analytics can quickly show you how many new subscribers your company got when you put out a new promotion or how many more air conditioners you sold when you placed a new coupon in the Sunday circular. But while descriptive analytics will certainly help to optimize profits, new analytics capabilities are what will drive revenue growth.
Predictive analytics. This approach helps businesses derive actionable insights from their data. The insights can help business leaders make business decisions that can generate better outcomes and improve performance. Predictive analytics is enabled by sophisticated statistical, forecasting and modeling capabilities. But it is much more than just a “technology solution”—it is a new way of doing business.
Sidebar 2 | The beauty of many quick pricing experiments
When Gary Loveman joined Harrah’s Entertainment (now Caesars Entertainment) in 1998, the casino company priced its slot machines like everyone else in the gaming industry. Management presumed that decreasing payouts—essentially raising the price—would drive some customers to other casinos. A sensible assumption, perhaps, but Loveman—a quantitative type who left a professorship at Harvard Business School to join the company—wasn’t one to assume. He commissioned a study to determine how sensitive slot-machine players would be to a change in payouts.
The company discovered it could make very small adjustments in the frequency of payouts—making them just one-tenth of 1 percent less frequent—without customers even noticing. Harrah’s shareholders, however, would notice. In fact, as Loveman later put it, single insights like this can “ring the cash register literally thousands or millions of times” in a large business.
Just about every company has its slot machines—product areas in which it could get a major boost in profitability by optimizing its pricing strategy.
Sidebar 3 | When 1 + 1 = 3
Some retailers are blending advanced customer analyses with price optimization analytics to attract, retain and grow valuable customer segments. A home improvement retailer combines detailed customer insight data to understand “willingness to pay” by customer segment. When this customer view is crossed with insights into product elasticity, the retailer can target price reductions and promotions that are focused tightly on the products that matter most to each particular segment.
For example, at the beginning of the last economic downturn, the retailer’s customer analyses revealed increasing consumer preferences for lower and more “round number” price points, such as $10 hammers. Additionally, the retailer found high product price sensitivity on items throughout its assortment. Combining these two data points, it undertook a large strategic effort not just to lower prices but to introduce a wider assortment of lower-cost price point goods in key strategic categories. It also led the company to expand its private-label offerings and adopt a design-to-price approach in which product development begins with the analysis of the final price point the company needs to achieve.
About the authors
Kenneth Dickman is the global analytics lead for Accenture Management Consulting Strategy. He works with consumer packaged goods and retail clients to develop and implement growth strategies and value-based transformation programs that build the marketing, merchandising and supply chain capabilities required to enable these strategies. Mr. Dickman also helps clients with innovative growth programs and corporate venturing disciplines, including the conception and launch of new businesses and retail formats. He is based in Chicago.
Jeanne G. Harris is a senior executive research fellow with the Accenture Institute for High Performance in Chicago, where she leads the Institute’s global research in the areas of information, technology and analytics. Ms. Harris was awarded the Lifetime Achievement Award for Women Leaders in Consulting by Consulting magazine in 2009. She is the coauthor (with Thomas H. Davenport and Robert Morison) of Analytics at Work: Smarter Decisions, Better Results (Harvard Business Press, 2010).
John G. Hanson is a Boston-based senior executive and the North American lead for Accenture’s Pricing and Profit Optimization group. His work focuses on pricing strategy and price management solutions primarily in the communications, media and technology industry. Before joining Accenture, Mr. Hanson was the vice president of business consulting at a pricing software vendor.