The Downside of the Data-Driven Decision
The law of unintended consequences states that for every action, there will be three unintended consequences, and one of them will be particularly unpleasant. The movement in banking toward increased use of analytics and data-driven decisions is a case in point.
The Los Angeles Times recently reported that at Wells Fargo, where employees are pushed to sell eight products and services to each customer household, staff have opened hundreds of unneeded and unrequested accounts for customers, ordered credit cards without customers' permission and forged client signatures on paperwork.
"One former branch manager who worked in the Pacific Northwest described her dismay at discovering that employees had talked a homeless woman into opening six checking and savings accounts with fees totaling $39 a month," the paper reported. After the article was published, 70 more former and current Wells Fargo employees and customers came forward with similar stories of sales of unwanted and unneeded products at the bank.
Wells Fargo disputed the tone of the LA Times article, without sharing any specifics. "Wells Fargo works hard to create a supportive, caring, ethical culture," Oscar Suris, executive vice president of corporate communications, wrote in a letter to the editor.
Ron Shevlin, senior analyst at Aite Group, thought the article was a witch hunt. "I also found it somewhat ironic that a newspaper that's part of an industry that's dying because they missed the boat on ad sales moving to other channels is criticizing a company in another industry for its sales tactics," he says. "Maybe if the paper was a little better at sales tactics itself it wouldn't be in an industry on the verge of collapse."
But the article raises a good question: Does rigid adherence to quotas, analytics and data-driven decisions have a downside, such as bringing out the worst in employees?
Wells Fargo doesn't deny the existence of its "Eight is Great" cross-sales program, which has been in effect for many years. The bank's customers currently have an average of more than six Wells Fargo products per household. The industry average is 2.3.
"If you're pushing product at a customer just to get him up to eight products, he may not be happy or satisfied, though he has lots of products," Shevlin observes. "Do you want your customers to be happy, to be profitable, or both?"
On the other hand, Shevlin defends the bank's right to rigorously cross-sell and upsell. "What industry isn't aggressive in its sales tactics?" he says. "You go to a car dealer and they want to sell you the most expensive car possible. You go to a realtor who wants you to buy the most expensive house you can buy. Every industry and salesperson is aggressive and assertive, that does not make them unethical."
It doesn't make such practices right, either.
"Wells Fargo does put a lot of pressure on its employees," comments Ted Triplett, chief marketing officer at Insight Ecosystems, an analytics consultancy in Lincoln, Neb. "I was a Wells Fargo customer and every time I'd go in to make a transaction they would always ask, do you need a home equity line? Employees may have felt pressure to do things they shouldn't do."
However, he doesn't believe Wells Fargo encouraged the opening of unwanted accounts. "I think employees made those decisions on their own, right or wrong," he says.
Banks generally need to improve their cross-selling, in Triplett's view. But, he says, "they need to do it the right way, by building relationships I call it relationship selling instead of cross-selling."
Wells Fargo says what it's doing is in the customers' best interest. "We want to build lifelong relationships with customers," says Mary Eshet, corporate media relations manager at the bank. "If the average household uses 14-16 products, then we believe there's more value for the customer to have much of their banking relationship with one company. We know them better, we can understand their needs better, it's more efficient for the customer and we can help them throughout the financial stages of their life if we understand their entire financial picture."
The bank's focus on the eight-products-per-household metric is an example of the type of "data-driven decision" that is in fashion with Big Data devotees in banking and everywhere else. At the Banking Analytics Symposium in October, an executive at Key Bank told of a regional manager who wanted to keep a branch open, arguing that he had a plan for "turning" the branch. The analytics group replied that the branch would only be kept open if performance improved four-fold within a few months. The goal was known to be unattainable and the branch was shut down.
"A bank like that scares me because banking is still a personal relationship business," Triplett says. "I don't care how much technology you have involved, people still want to talk to people."
Analytics don't always tell the full story, he observes. For instance, analytics might tell a bank its ATM usage is up month over month for the past year, but it doesn't explain why. Are the ATMs conveniently located for customers? Or is it just that all the nearby branches have closed and customers are desperate to access their money? That insight would only be obtainable through surveys or personal conversations with customers, Triplett says. "You'll never replace the relationship with technology, you have to have a combination of both."
A better use for analytics might be to tailor a sales quota to individual customers or segments.
"If they were concerned about their customers, they could conduct customer analytics to see what the potential of each customer is," Triplett says. "You could look at behaviors and determine that this customer has no more potential to buy more products and services, based on the amount of income they have." Another customer, perhaps older and wealthier, might need more than eight products.
And there's always room for human judgment in such metrics-driven initiatives.
"Combining data with experience and judgment will produce a better decision than the data by itself, 99 times out of 100," says Shevlin. "Maybe 96 times out of 100, because four times out of 100 your experience prevents you from seeing something different in that situation. The idea that we're going to take someone out of school who's 22 years old and has no experience and hand him data and brilliant decisions are going to come out of it is ridiculous."
Rather than beingfullydata-driven, organizations need to understand which decisions could benefit from more data, and what that better data is, he points out. "Everybody talks about Big Data like it's a big savior," he says. "More data is not the panacea. Figuring out which data will help you make better decisions or understand the situation clearly is what's needed."