An Old Yes in Commit Will Cost You at the QBR

Every sales leader I've worked with has a deal like this in the pipeline. It's been open for months, and the team still calls it warm. Nobody has spoken to the buyer in six weeks, yet it sits at the same stage and probability it had the last time anyone looked closely.

Most teams assume an open deal is a live deal, as long as the rep still feels good about it. But a "yes" from May is being carried into August as if it were a fact about today, and that old yes is the deal that slips at the QBR. Below is a simple test for finding these deals, and the one line to say about them in your next pipeline review.

The test I use comes from Eddie Reynolds of Union Square Consulting, who calls it the pipeline truth test in the GTM AI Podcast newsletter. Look at every opportunity open for more than twice your average sales cycle. If your normal cycle is 60 days and a deal has been open for 150, it may still be alive. But the original yes can no longer stand in for current evidence.

Run that on your own pipeline and you'll likely find a few deals in commit that fail it. Nobody is trying to mislead anyone. The deal is just old, and the evidence behind it is older. Those deals have to be re-earned, which means someone goes back to the buyer and confirms what is still true.

Stale Yeses Become Public Misses

If you're an AE, a stale deal in commit is slippage waiting to happen. It sits there looking solid until the week it matters. Then you're explaining in a forecast call why a deal you called for this quarter is suddenly next quarter. The miss is only half the damage; the other half is your manager wondering which of your other calls are built the same way.

If you manage a team, stale deals compound. Five reps each carrying two or three old yeses turns into a number you present with conviction. When it falls apart, it usually falls apart in a QBR, in front of the people who planned around it. At that point you have a credibility problem to repair, and coaching can't fix it after the fact.

AI Polishes Stale Evidence Instead of Fixing It

When I look at a deal past that line, the question I ask is when a person last confirmed what we believe about it. Not when the CRM was updated, and not when a tool summarized it. When did someone hear from the buyer that the problem, the priority, and the intent to move are all still there? A stage stays put until someone changes it, but a buyer's yes starts aging the day it's given.

That matters more now, because a lot of teams are putting AI on top of the CRM and expecting it to make the pipeline more honest. It helps in real ways. It can surface deals that haven't moved, summarize activity, and flag gaps faster than any manager scrolling a report.

But much of a deal never reaches the CRM. It lives in email threads, unrecorded calls, texts, and hallway conversations.

The AI works from the stage, the probability, the notes, and the activity history it can see. If the record says the buyer is interested, the AI treats that as part of the story. It will score the deal, recommend a next step, and roll it into the forecast, quickly and confidently.

What it can't know is whether a three-month-old yes still holds. It doesn't know the champion changed roles, the budget moved, or the priority disappeared. When the information is stale, AI doesn't fix it. It gives it a cleaner presentation, which is why a polished AI forecast can feel more credible than the evidence underneath it.

More reporting won't fix this. Going back to the buyer will. Priorities change, budgets move, and champions leave. A buyer who was enthusiastic in May may have gone quiet in August for a very good reason, and the only way to learn it is to ask.

When the answer comes back and the deal is gone, that call didn't create a loss. It revealed one you already had, early enough to stop planning around it.

Lead With the Last Buyer Confirmation

When a deal open longer than twice your average cycle comes up in pipeline review, say the date of the last buyer confirmation before you say the stage. Your next chance is this week's review. Sorting your open deals by days open takes about five minutes, and the line takes 15 seconds per deal.

"Acme's been open 140 days. Dana last confirmed the project was still funded on our September 3rd call."

If you can't give a real name and date, say so:

"Acme's been open 140 days. I haven't heard from anyone there directly since June. I'm calling Dana this week before I keep it in commit."

You don't have to delete those deals. They just can't hold up a confident forecast until the buyer confirms them again. Old deals then prove they're alive this week, or leave commit before the QBR finds them.