Book a demo
← Blog · Research

Why deals go quiet

We read 11,400 closed-lost deals from twelve sales teams. The first quiet signal showed up 23 days, on average, before anyone marked the deal at risk.

Sofia Lindqvist
Head of Machine Learning
Nov 12, 20269 min read

Ask a sales manager how a deal was lost and you’ll hear a story about the last week: procurement stalled, budget froze, a competitor came in low. Ask the email, calendar and call history and you get a different story, one that starts much earlier and much more quietly.

The data

Twelve B2B sales teams gave us read-only access to two years of closed deals: 11,400 lost and 8,900 won, with a median size of $52,000 and a median cycle of 74 days. For each deal we rebuilt the timeline of every email thread, meeting, call transcript and CRM edit.

Then we asked a simple question: when did this deal first start to look different from deals that closed? Not when the rep marked it at risk. When the data did.

Six patterns

Six patterns explained most of the early difference. Here is how many days before the loss each one typically appeared:

Median days before the loss that each pattern first appears

  • Reply rhythm breaks23 d
  • Next meeting disappears19 d
  • Champion changes role17 d
  • Hedging on calls14 d
  • Signer never shows up12 d
  • Close date moves twice8 d
11,400 closed-lost deals, 12 teams, 2024–2026.
  • Reply rhythm breaks. The buyer answers slower than their own normal. Median first appearance: 23 days before the loss.
  • Next meeting disappears. No future meeting on any calendar, or a recurring call quietly cancelled.
  • Champion changes role. New title, advisor role, or simply stops joining calls.
  • Hedging on calls. “Circle back”, “after planning”, “security will want a look”.
  • Signer never shows up. The economic buyer is named but never in a meeting.
  • Close date moves twice. By the second push, most of these deals are already gone.

The order matters

The patterns don’t arrive at random. In two thirds of lost deals, the reply rhythm broke first, and the meeting disappeared within a week. The date moving was almost always last, which is exactly when most teams start paying attention.

By the time a close date moves twice, the deal has usually been drifting for three weeks.

The single strongest predictor wasn’t any one pattern. It was two patterns within ten days: a broken reply rhythm and no next meeting together lost 71% of the time, against 22% for the average deal at the same stage.

We also checked how often the patterns showed up in deals that closed. Each one alone was common: a third of won deals had a slow week of replies. What separated lost deals was how they stacked.

What to do on Monday

  1. Measure against the buyer, not a rule. “No reply in seven days” flags the slow repliers who were always fine, and misses the fast ones who just went quiet.
  2. Watch for pairs. One pattern is noise. Two within ten days is the moment to act.
  3. Send the thing they asked for. In our data, a follow-up that delivered something the buyer requested got a reply 72% of the time. “Just checking in” got 19%.
  4. Say it out loud early. Teams that discussed drifting deals in week six recovered twice as many as teams that waited for week ten.

That is, more or less, what Riptally Signals and the Monday radar do for every open deal. But the rules work with a spreadsheet too. Start with the pairs.

Keep reading

Ready when your pipeline is

Don't letit drift

Book a radar review

A 20-minute walkthrough on your own last 90 days. No slides. You keep the report.