Deal size and pipeline stage tell you almost nothing about which deals will actually close. The metrics that do predict closure are behavioral: how fast a deal is moving relative to your baseline, how many people at the buying company are engaged, and whether your champion is going quiet. Track those three and you'll spot a dying deal weeks before it dies.
Most early-stage founders build their first CRM dashboard around the numbers that are easiest to pull: total pipeline value, deal count by stage, average deal size. Those are inventory numbers. They tell you what you're holding, not what's about to convert. A $150,000 opportunity sitting in "proposal sent" for six weeks with one contact isn't worth more than a $40,000 deal moving fast with three stakeholders replying same-day. It's worth less.
Why stage and size are the wrong signals
Pipeline stage measures where a deal is, not whether it's healthy. A deal can sit in "negotiation" for months while the prospect quietly moves on. Deal size measures upside, not probability. Neither one moves fast enough to warn you.
The problem compounds at the seed stage because founders usually have five to fifteen open deals at any time, not five hundred. With that little volume, a stage-and-size dashboard produces false confidence. You look at a full-looking pipeline and assume it's healthy, right up until three deals go quiet in the same week and your quarter collapses.
The three metrics that actually move first
Deal velocity variance. This is how much a specific deal's pace deviates from your historical average time-per-stage. If your typical deal moves from demo to proposal in nine days and this one has been sitting for twenty-two, that gap is the warning, not the raw day count. Deals moving roughly 20% slower than your baseline are already telling you something is wrong, even while every other field on the deal still looks normal.
Engagement density. Count the number of people at the prospect's company who are actively replying, not just cc'd. A deal with three engaged stakeholders who each respond within a day beats a larger deal with a single champion, every time. Buying committees at even small B2B companies now regularly run eight to eleven people. A deal that's still single-threaded after the first month is structurally fragile, no matter how enthusiastic that one contact sounds.
Champion behavior drift. Track response time as a trend, not a snapshot. A champion who replied in two hours during discovery and now takes four days per email is disengaging, even if the words in their emails haven't changed. Meeting reschedules and vague "let me check internally" replies are the same signal wearing different clothes. This is usually the first metric to move, and the easiest one founders ignore because the relationship still feels warm.
What the data says about timing
Deals that close within fifty days of entering pipeline have historically converted at roughly a 47% win rate. Past that window, win rates drop to 20% or lower. That's not a reason to rush prospects. It's a reason to treat deal age itself as a metric, not a footnote. If a deal crosses the fifty-day mark without a clear next step on the calendar, it has already shifted risk categories, whether your CRM flags it or not.
How to actually track this without extra software
You don't need a forecasting tool to watch these three signals. A basic CRM with custom fields does the job:
- Add a "days in current stage" field and compare it against your average per stage, recalculated monthly as you get more closed deals in the sample.
- Add a stakeholder count field, updated every time a new contact is CC'd or joins a call. Anything under three by the proposal stage gets flagged.
- Log the date of a champion's last same-day reply. If it's been more than a week, that deal gets a manual check-in, not another automated follow-up.
Fifteen minutes a week reviewing these three fields across open deals will surface more real risk than any dashboard sorted by deal size.
The 30-day move
Pick your five largest open deals right now. For each one, write down: days in current stage versus your average, number of actively engaged stakeholders, and days since the champion's last fast reply. You'll find at least one deal that looked "fine" on stage and size alone but is actually cooling off on all three signals. That's the deal to call today, not the one at the top of your pipeline report.
Frequently asked questions
What is deal velocity variance?
Deal velocity variance is the difference between how long a specific deal has spent in its current stage and your historical average time-per-stage for deals that eventually closed. A large gap is an early warning sign, independent of deal size or stage label.
How many stakeholders should be engaged in a healthy B2B deal?
For deals of meaningful size, aim for at least three actively engaged contacts by the proposal stage. Buying committees average eight to eleven people at many B2B companies, and single-threaded deals are far more likely to go dark without warning.
Does deal size predict which opportunities will close?
Not reliably. Deal size measures potential upside, not the likelihood of closing. Behavioral signals like engagement density and response-time trends are better predictors at the early stage, where sample sizes are too small for size-based forecasting to mean much.
How long should a deal stay open before it's considered at risk?
Deals that close within fifty days of entering the pipeline tend to convert at meaningfully higher rates than those that drag past that window. Treat the fifty-day mark as a checkpoint, not a hard cutoff, and pair it with the engagement and velocity signals above rather than relying on age alone.
Do I need a dedicated forecasting tool to track these metrics?
No. Three custom fields in a basic CRM, checked weekly, capture the same signal that expensive forecasting software is built to surface. The discipline of checking them matters more than the tooling.