When was the last time you verified the “idealness” of your ideal customer profile (ICP)? It’s a topic Launch Marketing CEO, Chris Leger, discussed last week on a webinar. Letting your targeting wander off the prize can slowly eat away at budget, sales’ confidence and down funnel results.

To break it down, there’s who the company believes it serves. This is talked about in decks and board meetings, but there’s also what your filters are letting in and the deals being chased that live outside the lines.

The ICP in your deck is a statement of intent. The operating version lives somewhere else: in the lead scoring model, the ad audiences, the routing rules, the sequence entry criteria, and the comp plan. Whatever those five things will accept is who gets sold to. When the deck and the systems disagree, the systems win. They get consulted a thousand times a day. The deck gets opened at planning.

The gap between those two is ICP drift, and I’ve watched it play out hundreds of times in 25 years.

Nobody drifts on purpose. It shows up one reasonable decision at a time. The quarter is short on volume, so you widen the lookalike. A warm inbound comes in off-profile and a rep works it anyway, because it’s warm. Scoring gets loosened after a slow month and never gets tightened back. A new channel arrives with its own idea of an audience. Every one of those calls is defensible on the day it’s made.

One root cause is budget. You have a lead number to hit and not enough money to hit it inside the tight definition, so the parameters get wider. Here’s the part that should bother you: I’d bet your sales team would rather you spent that same money going deeper on the accounts that actually convert, but maybe that connective tissue between sales and marketing is fractured or aligned around the wrong metrics?

Drift Isn’t Automatically Bad

If your persona is expanding because you’re gaining real traction in a new vertical, that’s growth. Take it. The condition is that you’ve built the content and the sales motion to serve that audience once it shows up. Deliberate expansion is strategy. The problem is the drift nobody chose, that nobody measured, and that nobody owns.

The Tells

• Lead volume is up and conversion is flat.
• Performance is dwindling quarter over quarter and no single thing explains it.
• You find yourself changing the audience instead of changing the messaging. That one catches more teams than any other.
• Your closed-won list has accounts in it that you would not have targeted eighteen months ago.

Where AI Comes In

AI didn’t create this problem. It made it cheaper to sustain and much harder to see. Lookalikes, intent signals, auto-built segments and agentic outbound will all happily widen your definition of a good fit and never mention that they did. The volume looks healthy for two or three quarters. Then win rates, cycle length and retention start sagging and nobody can trace why, because the drift happened at the input layer where nobody is looking.

Agents scale whatever definition you gave them. Hand one a drifted ICP and you now reach your entire market badly, at speed, with your logo on it.

Three Things to Do About

1. Define the ICP on retained revenue, not closed revenue. Your won-then-churned accounts are drift evidence sitting in plain sight. Review this quarterly looking back 18-24 months.
2. Put the definition into the systems, not the deck. Scoring, audiences, routing, suppression, sequence entry, comp. If a system accepts an account your ICP excludes, that system is defining your real ICP.
3. Instrument it. One number on the executive dashboard, trended: the share of new pipeline and closed-won that sits inside the current definition.

Tools will scale whatever definition you give them. The definition is the goal; having a game plan to surround those long-term perfect customers is the win. We’d welcome the conversation and opportunity to help you dig into where your ICP may be drifting. Reach out, let’s talk.