Financial companies have long depended on a careful mixture of hard data and human expertise to set product prices, but some are experimenting with what they call a more purely quantitative approach: using automated pricing software.
Though only a handful of companies are using this technology, some observers predict it could become standard practice within a few years.
"Before, the process was a little bit beyond throwing darts at a wall and trying to guess what the client wants," said Dan Rozelle, a vice president and the sales finance product manager at BB&T Corp. "It was a combination of science and art."
Using automated pricing software "brings to the table a more formal process and a more scientific database for analyzing pricing," he said.
The Winston-Salem, N.C., banking company began testing automated pricing software last month for its indirect auto loans. The test is expected to conclude this week, and Mr. Rozelle said he hopes to learn the software generated gains in both market share and profits. The full results will not be available until February.
Still, Mr. Rozelle said the software already has revealed some surprising trends. After running the software, he was expecting it to suggest that BB&T raise some of its rates, but in a few areas it called for lower ones.
And the proposed cuts were not minor, he said. "The numbers weren't off by 5 to 10 basis points. This was a decrease of 40 basis points." Also, the proposed cuts were "in areas that I wasn't expecting, but when I looked at it, the changes made sense," because they were areas with more "price sensitivity."
The key to setting prices, according to Mr. Rozelle is having as much information as possible - about the borrower, competitors, and the overall state of the market. He said that BB&T sales executives traditionally assess about eight to 10 factors when setting prices for indirect auto loans.
But the software from Nomis Solutions Inc. of San Bruno, Calif., "enables us to use data from a lot more places" to generate prices that lead to higher profits, he said.
Prashant Balepur, Nomis' director of product management, said that automated pricing was developed years ago in the airline industry to help sell as many seats as possible. Because seats on a flight are the same, the main variable in a sale is the customer, he said, and the software was written to evaluate such factors as why the customer is flying and when - business travelers and last-minute flyers usually are willing to pay more.
In the last five years the idea has spread to retailers, Mr. Balepur said. "Now we are seeing banks look around and say, 'Maybe we should try this, too.' "
When applied to banking, the primary idea remains evaluating who is applying for products and why, in order to figure out what they will be willing to pay, he said. "We have to understand everything leading up to a sale."
Using the Nomis software to price products could raise banks' profits from those products 10% to 20%, Mr. Balepur said.
Bobbie Britting, a senior analyst for consumer lending with TowerGroup, the Needham, Mass., market research unit of MasterCard Inc., said that banks traditionally will not differentiate in pricing among applicants with comparable household incomes, credit scores, and debt-to-income ratios.
However, one applicant might be happy to pay, say, 7% for a loan, while another with similar demographic features "might have been willing to pay even more, maybe 8%," and the difference is money left on the table, she said.
The Nomis software evaluates "lost quote data" - loan applications whose terms are approved by a bank but rejected by customers - because this information can offer insights into prices that may be too high to attract customers, Ms. Britting said.
"Banks have to have good data. That's super important," she said. "Many don't save this information."
Only about 9% of large U.S. banks are using or testing automated pricing, but that number will reach 95% by 2009, according to Ms. Britting. About 7% of midsize banks, and no small ones, have adopted the technology, but 60% of midsize companies and 40% of small ones should be using it by 2009.
"It's very new to financial services," she said. "But once a few companies start to use it, it will catch on pretty quickly."
Gwenn Bezard, a research director at Aite Group LLC, said that several Europe banking companies, including ABN Amro Holding NV, have been using automated pricing technology for several years. However, he predicted that it would be slower to catch on here.
The Netherlands, ABN Amro's home country, is small and has a relatively mature banking market, and there "the only way to grow in retail banking is to become more scientific" about how to approach your customers, he said. "In the United States, it's easier to go down the path of acquiring other banks as a way of gaining share."
Nomis and SAP AG are among the top vendors of automatic pricing software for the financial services market, Ms. Britting said. SAP entered the market in November of last year by purchasing the Scottsdale, Ariz., automated pricing technology vendor Khimetrics Inc.
Cynthia von Hollen, SAP's industry principal for financial services, said that four U.S. banking companies and an Asian one are using the Walldorf, Germany, company's pricing software. (She would not name the customers.)
Banks have long based their pricing strategy on offering "the price we're willing to sell at," but the strategy is inefficient, she said. "By putting a price at which they are willing to sell out there, they have no idea if that's the price customers are willing to pay."
According to Ms. von Hollen, banks can tweak the software to maximize profits, market share, or both. For example, a company that wants to maximize its profits can set the software to come up with the highest price it thinks many consumers will pay. Banks that want to increase share can set the software to deliver lower prices, but not so low that it leaves significant potential profits on the table.
"Getting the price right is a critical first step," she said. "Then you have to see how it fits into your overall strategy."
In the past some bankers have been accused of discrimination for offering different prices to customers who appear similar on paper - which is exactly what automated pricing software does. However, Ms. von Hollen said SAP's system ignores any criteria that could lead to charges of discrimination. "The software doesn't care where you live, or your color, or your gender. It comes up with prices based only on past events."
Using the software "is a paradigm shift for how banks price their products today."
Some users might encounter some resistance to the technology from sales people who are told that they should increase their prices, Ms. von Hollen said. Though an overall boost in sales could be good for the company, individual "commission sales people might not want to raise their rates because their volume may go down."
In this situation, banks may have to "change their compensation plans," she said. "This is truly disruptive technology."
