Your top rep's closed-won history is a coaching manual. It's not in a document anywhere — it's encoded in their activity data. The pattern of how they run discovery, how often they multi-thread early, how quickly they advance deals from first meeting to business case, how they maintain engagement during procurement delays — all of that is observable. The question is whether you've extracted it and turned it into something your developing reps can learn from.
Most sales coaching is based on conversation observation: ride-alongs, call reviews, pipeline walk-throughs where a manager asks a rep to explain their deals. That's valuable but it has two limitations. First, it's a sample — you're observing some of a rep's work, not all of it. Second, it's reactive — you're coaching after the fact, based on what you heard or what the rep told you, not on what the deal activity data shows.
Deal activity patterns give you a different view. They show you systematically, across all of a rep's deals over time, where their behavior diverges from your winning pattern — before the quarter ends and while there's still time to intervene.
Building the Baseline: What Your Closed-Won Cohort Looks Like
The starting point is understanding what your historical closed-won deals actually look like, behaviorally. Not in terms of deal characteristics (vertical, company size, ACV) but in terms of the rep's activity pattern: what activities were happening, at what frequency, with how many contacts, at what stages.
When you look across a closed-won cohort, certain patterns tend to emerge consistently in B2B SaaS selling:
- Multi-threading typically starts earlier than reps think it should. Deals that close often have a second stakeholder engaged before the business case is formally presented.
- Response latency from the prospect is generally below some threshold in the 21 days before verbal commit. Deals where response time was already degrading heading into final stages have a substantially lower close rate than deals with stable or improving response latency.
- There is usually a characteristic meeting cadence in the final 30–45 days — not just contact frequency but the nature of the contacts. Late-stage winning deals tend to have shorter, more specific meetings (pricing reviews, security reviews, procurement calls) rather than exploratory discovery calls.
- Email thread depth — the average length of email chains per stakeholder — tends to be higher in winning deals than losing deals. Longer threads indicate more back-and-forth engagement, which correlates with a more active buyer.
These patterns will be specific to your own closed-won cohort. Industry averages are useful context but your baseline should come from your own historical wins, because your sales motion, your product complexity, and your typical buyer profile all shape what "normal" looks like.
The Five Behavioral Signals Worth Watching
1. Single-Threading Past the Business Case Stage
A rep who reaches the business case stage with only one active contact is carrying material deal risk, regardless of how strong the champion relationship feels. Deals that close at this ACV typically require sign-off from at least two organizational functions. If your rep hasn't found a path into procurement, IT, or a second executive sponsor by the business case stage, that's a coaching conversation about stakeholder mapping — not about the deal itself.
2. Declining Contact Frequency Without an Identified Reason
Contact frequency naturally varies by deal stage. What's a red flag is when contact frequency drops sharply from one stage to the next without a corresponding activity note explaining it. A rep who had weekly calls with a prospect during discovery and is now going two or three weeks between contacts mid-way through the sales cycle — and whose CRM notes don't explain the gap — is either losing the deal's momentum or has lost track of it.
3. Long Stage Duration on Mid-Funnel Deals
Stage duration is one of the most underused coaching signals. When a deal has been in "Solution Reviewed / Proposal Out" for 2x the historical average for its ACV segment, the question isn't "when is this closing?" It's "what's blocking it?" A manager who surfaces that question at week 4 of a stall has six weeks to intervene. One who surfaces it at week 10 has a slipped deal.
4. Decreasing Email Response Rate
If a prospect who was replying to emails within 24 hours in the first six weeks of a deal is now taking five to seven days, that's a signal worth noting. It doesn't mean the deal is dead — buyers get busy, budget reviews happen, other priorities emerge. But it does mean the rep should be actively working to re-establish engagement, not waiting. A rep who doesn't notice or act on this pattern is a coaching target for buyer engagement strategy.
5. No Activity in the Two Weeks Before Forecast Close Date
This one is simple and often invisible. A deal in commit with a stated close date two weeks away, where the last logged activity was 12 days ago, is at risk. Not because the deal is necessarily lost — but because close dates with no recent engagement rarely hit on the stated date. This is either a CRM hygiene issue (the close date is stale) or an engagement issue (the rep has backed off too early). Either way, it's worth a conversation.
Coaching on Patterns vs. Coaching on Deals
The shift this framing enables is moving from deal-by-deal coaching ("tell me about the Harkon opportunity") to pattern coaching ("here are three deals in your pipeline where contact frequency has dropped in the past two weeks — let's talk about what's going on in each"). The pattern framing is more efficient and more diagnostic. It lets you test whether a behavior is deal-specific or systematic across the rep's book.
A rep who has five deals stalled in the same stage at the same time probably has a specific skill gap at that part of the sales motion — advancing from business case to procurement engagement, for instance. That's a different coaching conversation than a rep who has one stalled deal for an identifiable deal-specific reason. Activity data lets you distinguish between these cases.
The Caution: Pattern Recognition Is Not Mind Reading
This is not to say that activity signals replace understanding why a deal looks the way it does. A rep might have low contact frequency on a deal for a completely legitimate reason — the prospect is at a conference, the company is in a board meeting cycle, the rep deliberately slowed outreach because the champion asked for space during an internal budget review. Activity patterns are a prompt for a conversation, not a verdict.
The coaching signal is a question, not an answer: "I'm seeing this pattern on three of your deals — what's happening?" The answer usually reveals something useful either about the deal specifically or about the rep's process more broadly. The worst outcome is that you asked and everything is fine. The best outcome is that you caught a fixable problem with two weeks to spare.
Over time, comparing where deals end up against where the behavioral signals pointed — did the stalled deals that got a coaching intervention recover? did the ones with declining response latency slip as predicted? — builds your own calibration for which signals in your specific environment actually matter. That feedback loop is how coaching gets sharper.