ICP drift begins with a deal that should not have been closed. It is a customer who is too large, too small, in the wrong industry, or with a use case the product was not built for — but the revenue was real and the team was under pressure. That deal closes. Then another like it closes. The ICP definition expands to accommodate the new deals. The product roadmap adjusts to serve the new customers. The support team builds expertise in new use cases. And at some point the company realizes that its best early customers are not renewing at the same rate, and the new customers are churning faster than the model predicted.
How ICP drift happens
The most common trigger is a growth target that requires closing deals outside the existing ICP. The company has exhausted the easy wins within its defined segment and is now under pressure to grow revenue from new sources. The sales team, which is compensated on closed revenue, is incentivized to close any deal that can close. The deals that can close outside the ICP are typically larger (upmarket drift), or more accessible through existing relationships (industry drift), or pursuing a use case adjacent to the product's core strength (use case drift).
Each form of drift creates a different problem. Upmarket drift means acquiring customers whose success requirements exceed the product's current depth. Industry drift means acquiring customers whose workflow assumptions the product was not built around. Use case drift means acquiring customers who are using the product as a workaround for something it was not designed to do. All three produce the same result: customers who are harder to retain, harder to expand, and more expensive to support.
The diagnostic signals
ICP drift shows up in the data before it shows up in the narrative. The signals: NRR declining in cohorts acquired in the last three to four quarters, time-to-value increasing for new customers, support ticket volume per customer increasing, and customer health scores declining in accounts sourced through specific channels or closed by specific sales team members. The data is almost always present. The interpretation is almost always delayed because no one wants to conclude that the growth strategy is producing the wrong customers.
The correction
Correcting ICP drift requires two uncomfortable decisions: tightening the ICP definition back toward the segment where the product creates the most demonstrable value, and saying no to revenue that falls outside that definition. The second decision is harder than the first. No is easy to say in principle and difficult to execute when a deal is in front of you, the rep has invested three months in it, and the revenue would hit the quarter.
The companies that correct ICP drift successfully build the ICP definition into the sales qualification process at the front of the funnel rather than the deal review process at the back. When the ICP definition is the qualification standard — not the aspiration, but the actual pass/fail criterion for sales investment — the drift is caught before the deal closes rather than after the customer churns.