The gap between the go-to-market you designed and the one that's actually running.
Pull up any company's CRM and you'll find two businesses inside it.
There's the one they designed: the pipeline stages someone drew on a whiteboard, the lifecycle rules set up on day one, the playbook in the sales deck. Then there's the one that actually runs, the thing reps and marketers and CS people really do between nine and six.
GTM Drift is the distance between those two.
We started calling it that because teams kept describing the symptom and had no word for the thing. A forecast that's fiction. A handoff that never happens. A motion the founders could do in their sleep that nobody else can repeat. All the same problem, wearing different clothes. You can't fix what you can't name.
Here's why it stays hidden. A CRM is built to record what happened. It has no idea what was supposed to happen. So it shows you the real motion in lovely detail and never once flags that the real has drifted a mile from the plan, because nobody ever told it the plan. You can have spotless dashboards and total drift at the same time. That's the part that catches people out.
It isn't a sign of a bad team. It's the default state of any go-to-market run by busy people and never re-checked.
The map gets drawn once and never redrawn, so your stages still describe the product you sold three years ago. Process is voluntary and admin is friction, so reps work the deal, not the record. The handoffs between marketing, sales and customer success belong to no one, so that's where the gap collects. Improvements decay: you fix something at the sales kickoff, everyone holds the line for a few weeks, then it slides back and nobody checks whether it stuck. Meanwhile the tooling piles up until half your workflows contradict each other or quietly do nothing.
Put all of that together and the motion you describe to the board slowly parts ways with the motion in the data.
Once you start looking, it's everywhere, and the same shapes turn up across wildly different businesses.
Start with the handoffs. They're the worst of the lot, because they belong to no one. The most expensive is the moment a deal closes: sales is finished, customer success hasn't started, and the new customer's first experience of you is silence. Lay your early churn over the accounts that got no proper handoff and the two tend to sit right on top of each other. The earlier handoffs, marketing to SDR and SDR to AE, leak the same way; they're just cheaper to ignore.
Then the forecast theatre. Close dates that shuffle to next month in a Friday ritual. Deals parked in 'Negotiation' for months because nobody wants to mark them lost. A late stage that's quietly become a junk drawer. The numbers get presented with a straight face while everyone half knows they're a story.
There's the fast response trap. Everyone agrees on replying quickly, and the team genuinely is quick, on the low intent signups. The high intent demo request routes to a senior rep who's always on a call, so your best leads wait the longest. Fast at the wrong things.
There's the people gap. Your best rep and your worst run completely different motions, and the spread is wider than anyone says out loud. Worse, the people who sell best often log the least, so the motion that actually wins is invisible and can't be taught. You don't have a single sales process. You have as many as you have reps.
There are the words nobody agreed on. Marketing's 'qualified' and sales' 'qualified' are different fields meaning different things, so the endless SLA row is really a vocabulary row no one has named. 'Lead status' means three things to three teams, and the dashboard everyone trusts is built on top of all three.
And there's the plumbing. Workflows nobody has opened since the person who built them left. Automations sending contradictory emails to the same contact. Fields whose names stopped matching their contents years ago. The re-engagement campaign that was meant to wake dormant leads and quietly stopped firing months back.
None of this is exotic. It's sitting in your CRM right now, and most of it has never been counted.
You can't fix what you can't see, and drift is invisible by default. The method is the same whatever you sell.
Write down the motion you think you run. One page. What's meant to happen at each stage, who owns each handoff, what 'qualified' actually means. If three of your leaders give three different answers, that's your first finding, and you haven't touched the data yet.
Then reconstruct the motion you actually run from what's already sitting in the CRM, and compare the two. The gap is your drift. Rank it by money, not by how big the gap looks. A slow response on your best inbound costs you more than a blank field on deals you lost months ago.
AI doesn't fix a drifted go-to-market. It runs the one you actually have, faster.
Point an agent at a routing rule that's already adrift and you get the wrong leads chased at machine speed. The teams who'll get the most out of the next few years aren't the ones with the most AI. They're the ones whose real motion is close enough to the one they intended that automation amplifies the right thing. So the order matters: see the gap, close it, then point the robots at it.
We built a product for exactly this problem. Kusabi reads everything a CRM already logs, reconstructs the motion that's actually running, and shows you the gap against the one you meant to run, ranked by what it's costing you.
But the idea matters more than the tool. A whiteboard and one honest afternoon will show you most of your drift. The hard part was never the software. It's deciding to look, and being willing to find that the business you're running isn't quite the one you designed.
If you read this and a few uncomfortable examples come to mind, good. Now you've got a name for it.