Deal Age Is a Better Forecast Signal Than Stage. Most CRMs Don't Even Track It.

Your pipeline review opens the same way it always does. Someone pulls up the CRM, scans the stages, and the conversation becomes a negotiation about which column things belong in. Proposal at £180k. Verbal at £95k. Negotiation at £240k. Everyone nods. You close the call with a number that feels about right. Six weeks later, two of those deals slip, one goes dark, and the VP of Finance asks why the forecast was off by thirty points. The answer is almost always the same: stage was lying, and nobody looked at age.

Stage is a story. Age is evidence.

Stage is a rep-reported field. In most CRMs it moves forward when the rep decides it should move forward, ratified by a manager who heard a confident update on a Friday call. There is no forcing function. Nothing verifies that the buyer has done anything. A deal that a rep dragged into "Proposal" eight weeks ago because the deck went out can sit there indefinitely, wearing the same stage badge, adding exactly the same notional value to your weighted pipeline, while the actual buying process has stalled completely.

Age, by contrast, is an event-log question. When did the buyer last take a meaningful action? Not when did the rep send something. When did the buyer reply, attend something, share a document, introduce a new stakeholder, or accept a next step? That timestamp is either recent or it isn't. You can't talk your way around it.

The distinction matters because close rates do not degrade uniformly across stages. They degrade by age within stage. A deal sitting in Proposal at day 14 behaves very differently from one sitting in Proposal at day 60. If your CRM is surfacing both of them at the same weighted probability, your forecast is off by design.

The numbers, specifically

Based on patterns across the sales organisations I've worked in and advised, here is what age-gated close rate data tends to look like. These are directional benchmarks, not a controlled study, but they are close enough to what you'll find in your own export to be worth acting on.

Read the last row slowly. A deal your rep has logged as a verbal commit, with a 31-day-old last buyer action, closes at roughly the same rate as a fresh Discovery deal. You are not forecasting a commit. You are forecasting a ghost.

The gap between age bands widens as stage advances, which is the counterintuitive part. You might expect that a late-stage deal would be resilient to a slow patch. The reverse is true: a buyer who has been silent for five weeks while supposedly preparing to sign is the highest-risk deal in your pipeline. The momentum that carries late-stage deals to close is fragile. It disappears faster than it appears.

The two CRM fields most RevOps teams are ignoring

If you pull up your CRM's field list right now, you probably have: Stage, Close Date, Amount, Owner, Created Date. Some of you have a Stage Entry Date field, which is mildly useful but still tells you nothing about buyer behaviour. Here are the two fields worth adding or surfacing.

Last Buyer Activity Date. Not last activity. Buyer activity. The distinction is critical. Reps log outbounds, emails, notes. That activity does not tell you the buyer is engaged; it tells you the rep is trying. You want a field that captures inbound-side actions only: replies, attended meetings, shared documents, accepted calendar invites, introduced new contacts. Most CRMs can derive this from sequence reply data or meeting logs if someone configures it. If yours can't, a rep-maintained date field with a clear definition is better than nothing and takes about ten minutes to implement.

Days Since Last Buyer Action. A calculated field. Literally today minus Last Buyer Activity Date. Surfaces in your pipeline views as a number. Anything over the age threshold for that stage turns red. This is not a sophisticated build. It is a few hours of RevOps time, and it changes every pipeline conversation you have from that point forward.

Once these fields are live, you age-gate your stage probability weights. Deals that breach the threshold for their stage get a reduced probability applied automatically, regardless of what the rep logged. Your forecast becomes a function of observed buyer behaviour, not rep optimism.

If you want to run the numbers on what this does to your coverage ratio before you build it, the Pipeline Coverage & Gap Calculator will give you a quick read on how much of your current pipeline survives an age-gating filter, and by how much you'd need to build to compensate.

The threshold benchmarks worth piloting

If you're starting from scratch, these are the age thresholds I'd test first. They are not universal. Your cycle length, deal size, and segment will shift them. But they are a defensible starting point.

Negotiation and Verbal tighten considerably. A deal in negotiation where the buyer has been quiet for three weeks is not in negotiation. It's stuck, and calling it something else doesn't help you plan.

Once you have six months of data, run the same close rate analysis I described above against your own numbers and adjust accordingly. The Win Rate & Forecast Accuracy Tracker makes this straightforward if you're pulling from a CRM export.

What changes on the forecast call

The shift in conversation is the real win here. When age is visible, the question stops being "what stage is this?" and becomes "when did they last do something?" Reps cannot talk around a 47-day-old buyer action with a confident tone of voice. The number sits there. Managers stop approving deals up the funnel on vibes and start asking for a concrete account of buyer engagement.

That is not a technology change. It is a discipline change, enabled by a tiny configuration change. Most forecast accuracy problems are not data problems. They are accountability problems dressed up as data problems.

The stage column will keep lying to you until you give it something honest to sit next to.