It's no secret that consumers vary enormously in the way they buy and use financial products. Frequently, however, industry pricing structures do not fully take these variations into account.
By applying consumer modeling techniques, in conjunction with targeted in-market testing programs, banks, brokerage firms, and other financial institutions can reprice their current product portfolios based on actual customer usage.
Usage-based, or "relationship," pricing can quickly and dramatically improve the profitability of low-profit customers without endangering existing relationships with high-profit customers. What follows are excerpts from a Mercer Management Consulting report.
Most financial service companies, including retail banks, credit card companies, small business banks, and brokerage firms, serve a wide spectrum of customers who differ along a variety of dimensions-the types of financial products they buy, the way they use those products, the channels through which they access the products, the levels of service they demand, their creditworthiness, their sense of loyalty to a particular provider or brand, and so on. As a result, they vary enormously in their revenue potential, their cost-to-serve, and, ultimately, their profitability.
Recent Mercer analyses reveal, in fact, that the usual "80/20" rule of customer profitability-80 percent of profit comes from 20 percent of customers-often turns into a "110/20" rule in financial services. The top 20 percent of the customer portfolio generates all the profits, in effect cross-subsidizing the remaining 80 percent, who are either marginally profitable or downright unprofitable. The dramatic skew in customer profitability has become a strong motivator for financial service executives to find ways to improve the profitability of their weakest performing customers. Although it is practically impossible to prevent customer profitability skews from ever arising, they can today be managed more aggressively and effectively than ever before. Highly sophisticated models of customer purchasing patterns and behavior can be used to segment customers according to their usage and economics. Armed with this knowledge, companies can develop highly targeted "relationship pricing" structures that directly link prices to each customer's buying and usage behavior-i.e., their relationship value.
With relationship pricing, the targeting mechanism is embedded in the pricing structure itself. Individual customers "activate" the appropriate pricing action based on their actual usage of the product or service.
Effective repricings
The bottom-line payoff from relationship pricing can be swift and substantial. Relationship pricing opens three sources of added profit from currently low-profit customers:
n The bulk of the gains typically flow directly from raising the prices actually paid (without instigating major behavioral changes).
n Additional gains flow from using repricing to change service usage patterns (e.g., reducing transactions, using less expensive channels).
n A final, if relatively modest, profit boost comes from the attrition of unprofitable accounts that refuse to either modify their behavior or pay more realistic prices (see chart).
Typically, these actions can recoup 60 to 85 percent of the total profits forfeited in the low-profit segments. (The remaining 15 to 40 percent of lost profits is beyond the reach of repricing, due to non- negotiable pricing terms with existing customers, or simply beyond the limits of a repricing program's effectiveness.).
The big challenge in relationship pricing is to increase prices to the point where unprofitable customers become profitable while leaving profitable customers relatively unaffected.
Given the complexity of the problem, what is the best way to develop and introduce relationship pricing structures? Mercer has successfully used a four-stage approach across a range of financial services products Our approach draws on state-of-the-art tools and techniques that we have developed for understanding customer profitability and for developing pricing policies based on hard consumer behavior and service usage data. Using our approach, managers are able to base complex repricing decisions on concrete product profitability data along with grassroots consumer information derived through in-market tests. No longer do they need to fall back solely on professional judgment or the emulation of competitor actions in setting prices.
n Customer Profitability Analysis and Segmentation
The first step is to develop a clear and accurate understanding of the relationship between customer usage behavior and profitability. A sophisticated customer profitability algorithm needs to be developed that assigns cost-to-serve, revenue, and profitability estimates to each customer on the basis of actual recent usage history. Applying this algorithm enables the profitability skew to be documented, revealing the size of the repricing opportunity.
See the Forest
Since customers may be unprofitable in one product (e.g., savings) but use other products (e.g., mortgage) that make the overall relationship profitable, it is important to examine total relationship profitability before targeting customers for repricing
n Development and modeling of repricing alternatives
The overall screening process begins with a qualitative feasibility review of a large range of repricing "themes." For example, a theme may be "provide incentives for shifting to cheaper channels" or "reduce overall transaction volumes." Feasible themes are then translated into specific pricing structures and levels, providing a set of alternatives for further testing.
At this point, we use financial modeling to gauge the likely profitability impact of the different alternatives. Using estimates of price elasticities, we integrate forecasts for all major behavioral drivers of cost (e.g., branch deposits) and revenue (e.g., interest income, fees) into an overall pro forma P&L model.
n In-market tests provide reliable estimates of price elasticities and behavior changes under real market conditions. They can be used to gauge the impact of both the pricing structures (e.g., attaching fees to research requests) and the price levels within the structure (e.g., two dollars per research request, waived for 12 commissionable trades/year). Moreover, because the tests can be isolated to specific submarkets, they can be conducted quickly and with relatively low risk and can, to a degree, be shielded from competitors.
Experimental design principles help determine the optimal number and type of test cells for achieving the best learning in the most efficient manner. . We have found that by applying scientific principles test cells can be limited to several thousand customers, far fewer than the tens of thousands that are frequently required.
n The rollout of the optimal repricing policy should follow the in- market testing very swiftly-within two or three months. The organization will be more efficient in making the changes necessary if it is able to ride the momentum developed during the testing phase.
It will, of course, take time to translate the learnings from the in- market testing into the general rollout and to prepare the infrastructure for across-the-board repricing. Nevertheless, the pressure for a swift rollout must be maintained. Otherwise, other priorities may begin to siphon away the organization's focus and resources.
Aid shareholder value
We would argue that in today's fragmented and heavily contested marketplace, decision making must seek to balance share optimization with profit optimization. Otherwise, financial institutions will find themselves expending enormous resources chasing unprofitable customers, ultimately eroding their bottom line and their shareholder value. The shift to a more balanced view of the marketplace will require a great deal of communication and education within an organization, but, as we are seeing with relationship pricing, the benefits can be great.
Mercer Management Consulting's Stephen Kutner is a vice president based in Boston; John Cripps is a principal based in San Francisco; Michael Smith is a vice president in charge of the firm's San Francisco office; and Corey Yulinsky is a vice president based in New York.
