Wells Is Exhibit A of Employee Incentive Failures
"You get what you incent." That's the common theme behind most financial scandals, including Wells Fargo's $185 million settlement with federal regulators and the Los Angeles city attorney over allegedly unauthorized account openings. The improper practices, known as "bundling" and "sandbagging" by Wells employees, were fueled by an incentive system that paid employees based on the number of accounts opened.
Incidents like these are an inevitable byproduct of modern management's attempts to "reverse engineer" corporate performance by breaking broad corporate goals into component parts and pushing those parts down the organization structure, layer by layer. Those key performance indicators, or KPIs, become the biggest component of employee performance evaluations and the key to how much money individual employees make.
It seems like a perfectly reasonable way to translate broad strategy into individual performance goals. But as KPIs move down the management structure and are embedded in compensation decisions at lower and lower levels, they become increasingly narrowly focused, more and more detached from broader goals. Keep in mind that front-line employees have no real input into their own goals, which are set for them, but receive constant feedback on "how they're doing" on KPIs communicated from the layer above.
It's no wonder they resort to gaming the KPIs for personal advantage—whether to make more money or avoid being fired. From their perspective, the whole system looks like a game, unrelated to their day-to-day experience with customers. That's clearly what happened at Wells Fargo. By the time the corporate goal of maximizing the number of product relationships per customer—the sacred "cross sale ratio" that had been Wells' best calling card with equity analysts for years—made it to the front line, it incented those employees to create bogus accounts to drive up their KPI numbers. Over 5,000 Wells employees ended up losing their jobs when all this was uncovered. One hopes that more than one or two of these were senior managers, although Wells is not telling.
This kind of incentive-driven bad behavior happens at all levels in modern corporations, not just in front line posts. Frequently it is committed by a CFO, whose compensation is based heavily on "delivering the numbers," bending the accounting rules. Or, as is alleged to have happened a few weeks ago at Fiat Chrysler, where a "tough-minded" sales executive is said to have falsified sales numbers. As one Chrysler dealer said: "He gives us a number, and we have to figure out how to hit it." But KPI-driven misbehavior is particularly dangerous at the front-line customer-facing level, where the emphasis should always be on doing the right thing for customers and building trust and customer loyalty through positive experiences.
How can managers be sure that the goals they are setting for the people working for them (and the goals that they are receiving from the level above) don't encourage bad behavior? One way is to abandon numerical goals entirely and return to the days of "gestalt" performance and compensation decisions by direct managers. In theory, giving managers complete discretion would allow them to value all contributions by individual employees, like teamwork, pleasing customers, etc. in a way that maximizes the overall performance of their unit. Direct managers would love this. Unfortunately we have seen over time that abandoning all top-down guidance can encourage favoritism, discrimination and other bad outcomes.
Another option, which has been frequently tried, is to "reverse engineer" a customer satisfaction rating system like the Net Promoter Score (NPS), and use individual NPS-derived individual metrics to balance operational and financial KPI metrics in the hope that this will avoid bad behavior. This may work a little better, but it leaves the operational KPI incentives in place while allowing front-line employees to game the NPS process, often by working around the compliance infrastructure to get a good "review" from an impatient customer.
So what approach would work better for both employers and employers? How about empowering front-line employees with goals that inspire them rather than attempting to micromanage them with sterile KPIs? How about focusing them on doing things that improve their customers' financial lives?
The Center for Financial Services Innovation (CFSI) has come up with eight key indicators of financial health, which can provide a starting place for rethinking front-line KPIs and compensation practices. CFSI believes an individual is financially healthy when he or she can:
- Spend less than income
- Pay bills on time and in full
- Have sufficient liquid savings
- Have sufficient long-term savings or assets
- Have a sustainable debt load
- Have a prime credit score
- Have appropriate insurance
- Plan ahead for expenses.
Why not design a system that incents front-line employees to help customers improve their financial health and measures how employees are helping customers reach these goals? KPIs tied to real and tangible customer goals like these would unleash the creativity of employees, improve employee engagement and reduce turnover. And there is little doubt that the best way for a bank to grow, delight its customers and reward its shareholders over time is to help customers become and stay financially healthy.
Defenders of the current approach to front-line employee compensation view bad behavior as an "unintended consequence" of efforts to find the perfect system for incenting profitable behaviors. Their solution to things like the Wells situation is implementing better and more detailed operational and financial KPIs to try to manage every aspect of employee behavior through incentives. But that approach, oddly similar to the regulators' quixotic quest for perfect risk-based capital weightings, misses the point that human beings really can't be micromanaged into perfect behavior. They can, on the other hand, be led to create value through active customer engagement in support of a clearly understood and socially valuable set of goals. As for the objections that will come from consultants who have built their careers trying to turn front-line employees into living algorithms, who cares? Their efforts over the years to micromanage employee performance have left bank customers alienated and bank reputations damaged. It's time for a new approach.
Todd Baker is a senior fellow at the Mossavar-Rahmani Center for Business and Government at the Harvard Kennedy School. He is managing principal of Broadmoor Consulting LLC.