Economics & Markets
More Upsells, Lower Loyalty
Wells Fargo wanted eight products per household; the quota that chased them cost the bank its customers’ trust.
2026-08-081 min read
An upsell quota changes what a salesperson is optimizing for, and the change is invisible on the dashboard that matters to leadership. Products-per-customer is a real measure of depth when customers choose the products, and a measure of pressure when they do not. The two look identical in a quarterly report. That is the trap: the metric keeps rising while the thing it was built to proxy — a customer who wants more of you — quietly inverts. By the time churn shows the damage, the incentive has been running for years and the sales culture has reorganized itself around it.
Wells Fargo built its retail strategy on cross-selling, with an internal goal of eight products per household that the bank repeated often enough to become a slogan. Branch employees carried daily quotas. When the quotas outran what customers actually wanted, staff opened millions of accounts and cards that customers had never asked for, and the bank kept reporting a cross-sell ratio that looked like loyalty. The reckoning arrived in 2016: regulators fined the bank, the chief executive resigned, and the head of community banking departed. Regulators later capped the bank’s total size for years — a penalty aimed squarely at growth, imposed on a strategy built to produce it.
The instructive part is not that employees cheated but that the metric could not tell the difference. A products-per-customer number rewards the sale and never sees the resentment, so it will keep climbing right up to the moment the relationship breaks. Measuring whether customers use what they bought closes that gap, because an unwanted product shows up immediately as a dormant one.
Key insights
Products-per-customer measures depth when customers choose and pressure when they are pushed; the two are indistinguishable in a quarterly report.
Usage-after-purchase is the cheapest available check, because an unwanted product goes dormant immediately.
The lag between incentive and churn is long enough that the strategy looks like it is working for years.
Why it matters
A rising cross-sell number can mean depth or pressure, and the report cannot tell you which — so it keeps reassuring you while the base erodes.
Quota-driven selling reorganizes the sales culture around itself, which means the damage outlives the incentive that caused it.
Use this tomorrow
1Pull the last quarter’s add-on sales and count how many of those accounts have not touched the added product in thirty days; if it is more than a quarter of them, your number is measuring pressure.
2Ask three of your salespeople what they would have to stop doing to hit next quarter’s quota, and write down every answer that involves a customer.
Go deeper
The failure is a textbook case of Goodhart’s law: when a measure becomes a target, it stops being a good measure. Cross-sell ratio worked as a proxy for relationship depth precisely while nobody was compensated on it. Attaching quotas to it severed the link between the number and the thing it stood for, without changing the number’s appearance. Every metric promoted to a target needs a second, un-incentivized measure watching it from outside.
There is a counterweight worth naming: some genuinely useful products do need a nudge, and a bank that stopped recommending anything would leave customers under-served. The distinction is whether the customer would keep the product if you asked them again a month later. Firms that survive the shift usually replace sale-count quotas with usage-weighted or retention-weighted ones, which move more slowly and are much harder to game. The cost is a noisier short-term number in exchange for one that still means something in three years.
