Pipeline
“No activity in thirty days” is the report nobody trusts.
Because it is wrong in the way that costs you the room. It flags the deal that moved to contract yesterday and nobody has phoned since, and the rep says — correctly — that the report does not understand their deal. After that, nobody reads it.
Two clocks fixes it.
The rule
Whichever moved last is the one that counts.
A deal is stale when both the stage it sits in and the last time the customer heard from you are older than the threshold. Move a deal forward and the clock restarts, even if nobody has called. Speak to the customer and the clock restarts, even if the deal has sat in one stage for a month.
Endings are excluded by definition — a won deal is not ageing.
- Okonkwo Industries41 days ago34 days agoStaleBoth clocks stopped. 34 days against a threshold of 14.
- Halvorsen Group2 days ago22 days agoNot staleMoved stage this week. The clock restarted.
- Meridian30 days agoyesterdayNot staleSitting in one stage a while, but the customer is live.
- The threshold is yours, and can differ by stage —
A workspace default, with any stage able to override it. Negotiation and discovery do not go stale at the same speed.
- The boundary day is counted the way people count it —
A deal that entered the stage exactly fourteen days ago reads fourteen, not thirteen. Which sounds trivial until somebody's threshold is fourteen.
- One rule, not one per screen —
The nightly nudge, the assistant, the dashboard card and the API all ask the same question, so two screens cannot disagree about which deals are stale.
The nudge
Once a week, to the owner, and never to a league table.
An overnight sweep finds what has gone stale and tells whoever owns it. The cooldown is seven days — chosen to sit between daily noise and a reminder nobody remembers receiving.
It is deduplicated per deal, not per person, and that is deliberate: had it been per person, reassigning a deal would immediately spam whoever inherited it.
A deal with nobody on it is skipped and counted rather than escalated, because an unowned deal is an administrative problem a reminder cannot solve.
- An unowned deal is counted, not escalated —
A deal with nobody on it is skipped by the nudge and reported in the sweep's own numbers. It is an administrative problem a reminder cannot fix, and pretending otherwise just sends mail to the wrong person.
- One bad record never stops the sweep —
Each deal is handled on its own, so a single malformed row cannot silence the whole night's reminders — the failure mode that makes people stop trusting overnight jobs.
›Does an email to the customer count as contact?
Not on the deal today — and this is the limit worth knowing. Automatically captured emails and calendar events attach to the contact, and staleness looks at what is attached to the deal. So an exchange with the customer does not currently stop a deal ageing.
What does count: an activity logged against the deal, and a call recording tagged to it. If your team works mostly by email, expect this to read fast until that is connected.
›Can different stages go stale at different speeds?
Yes — the workspace sets a default and any stage can carry its own. Discovery and negotiation rarely deserve the same clock, and a single number across a long pipeline is why most staleness reports get ignored.
›Does the manager get told?
No. The nudge goes to the owner. There is a screen a manager can open when they want to, and nothing arrives in their inbox about somebody else's deal.
›Can a rep turn the nudge off?
It is a notification like any other, subject to the same per-person settings.
›Where can I see staleness besides that one screen?
Nowhere, today. The rule runs everywhere but the badge appears only on the manager's triage screen — the deal list and the board show a plain day count with no threshold applied. It is on the list to surface properly.