Try this with your sales team. Ask three reps what has to be true for a deal to sit in "Qualified."
You'll get three answers. Usually something like "I had a good call," "they said they have budget," and "it felt real." All three reps are being honest. None of them is using the same definition, which means the number at the top of your pipeline report is an average of three different opinions.
This is the most common reason a forecast misses, and it has nothing to do with rep discipline. Here's how to fix it.
Most pipelines are named after what the rep did. Contacted. Demo Scheduled. Proposal Sent. Those feel like progress, and they're easy to log, which is exactly the problem — a rep can complete all three and be no closer to a deal than when they started.
A stage should describe what the buyer did. Because the buyer is the only one who can actually move a deal forward.
"Proposal Sent" tells you a PDF left your outbox. "Proposal Reviewed with Economic Buyer" tells you something happened on their side. The first is an activity. The second is evidence.
An exit criterion is the specific, checkable condition that must be true before a deal leaves a stage. Not a feeling. Not a judgment call. Something you could point at in the record.
The test is simple: if two people could look at the same deal and disagree about whether it belongs in that stage, the criteria aren't finished.
Here's what that looks like on a straightforward B2B pipeline:
Qualified — We've confirmed the problem they're trying to solve, who owns the budget, and roughly when they need it resolved. All three, written in the record. Not two out of three.
Discovery Complete — We've spoken with at least one person who will use what we're selling, not just the person who found us. Their current process is documented well enough that we could describe it back to them.
Solution Agreed — They've seen what we're proposing and told us it addresses the problem. Any objection about approach has been raised and answered.
Proposal Reviewed — The proposal has been walked through live with whoever signs. Not emailed. Walked through.
Verbal Commitment — They've said yes and named what has to happen procurement-side before signature.
Notice none of these are things your rep does alone. Every one requires the buyer to have participated.
Get the people who actually work deals in a room — not just the sales leader. Take one stage at a time and ask two questions:
What's the last thing that happened before you knew a deal was really in this stage? The answer is almost always something the buyer said or did. That's your criterion.
What's the most common reason a deal goes backward from here? Whatever they name is usually a criterion you skipped in the previous stage.
Write each one as a sentence a new hire could apply on their first week without asking anyone. If it needs interpretation, it isn't done.
Keep the pipeline short. Five or six stages is plenty. Every stage you add is another definition to maintain and another place for reps to disagree.
A definition nobody enforces is a suggestion. HubSpot gives you a few ways to make the criteria real:
Required properties per stage. In your pipeline settings, you can require specific properties before a deal moves into a stage. If "Qualified" means budget owner is confirmed, make the Budget Owner property required at that gate. The rep can't advance the deal without it.
Write the criteria into the stage description. Each deal stage has a description field, and it appears when a rep changes the stage. This is the cheapest, most underused feature in the whole pipeline setup — the definition shows up at the exact moment someone needs it.
Customize what shows on the board card. Put the properties that matter to your stage gates directly on the deal card so they're visible without opening the record. It makes gaps obvious at a glance.
Build a report on stage duration. Once criteria are real, time-in-stage becomes meaningful. A deal sitting in Discovery for sixty days now tells you something specific rather than just looking stale.
Three things, and the first one is the reason to do this at all.
Your forecast becomes defensible. Not more optimistic — defensible. When "Solution Agreed" means the same thing across every rep, the probability attached to that stage starts describing reality.
Coaching gets specific. Instead of "you need to qualify better," it's "this deal has been in Qualified for three weeks and we still don't know who owns the budget." That's an actionable conversation.
Deals stop going backward without explanation. When a deal slips, you can see which criterion was never actually met — usually two stages earlier.
One warning: expect your pipeline to shrink the first month. Deals will drop back to earlier stages once real criteria get applied, and the total will look worse. That's not a regression — it's the first time the number has been honest. The leaders who survive that month are the ones who explained in advance that it was coming.
Don't redesign the whole pipeline. Pick the stage where deals pile up and die, write exit criteria for just that one, and require the properties that prove it. Live with it for a few weeks before touching anything else.
The goal isn't a perfect pipeline. It's a pipeline where three people describe the same deal the same way.
The RevOps Foundation Scorecard takes about five minutes and gives you an honest read on your CRM structure, data quality, alignment, reporting, and AI readiness — including whether your pipeline is telling you the truth.
Take the RevOps Foundation Scorecard →
Want a second set of eyes on your pipeline? Schedule a call and we'll walk through it together.