The question

It’s in Commit. The close date was last week. Is this deal actually going to close?

Everyone in the pipeline review has a feeling about it, and nobody wants to be the one to ask. The rep is confident, the manager is hopeful, and the forecast is counting on it.

Where it sits on the loop

Pipeline, at the point it’s meant to become Won. It’s the job of the deal review and the manager’s 1:1, and it feeds the forecast call: every Commit deal that isn’t real is a hole in the number.

How to think about it

Ask what the buyer has done, not what the rep believes. Five questions, in this order:

Is the close date still in the future? A passed date with no new one is a fact, not a detail.
Who on the buyer’s side has done something recently? Replied, met, sent a document. Being emailed doesn’t count.
Is the person who signs engaged, or only named? A name on a slide isn’t engagement.
Is the next step dated and owned by the buyer? “We’ll follow up” is ours. “Their CFO reviews it on the 14th” is theirs.
What’s the first thing failing? Fix that, not the whole scorecard.

Qualification frameworks are the checklist behind questions 2 to 4. The questions matter more than the acronym.

Common mistakes

Walking every field of the scorecard. Forty minutes on one deal, and the one thing that’s actually failing gets the same time as everything else.
Arguing about the probability. 70% or 80% is opinion. What the buyer did last week is evidence.
Treating rep confidence as evidence. Confidence is useful context; it isn’t a buyer action.

In Summit53

The five questions map onto the opportunity cockpit, and two skills prepare the review for the manager and the rep.

1. Find the Commit deals at risk

In Deals → Opportunities, open Show Filters, tick Commit under Forecast Categories and click Apply Filters.
On the Revenue Engine, look for deal items in the Action Brief, such as a close date that has passed, and for deal signals on Pipeline in the Stage Inspector: Close date has passed, No activity for over two weeks, Weak qualification for a late-stage deal or High-value deal with low health.

See Opportunities and Diagnostics.

2. Ask the five questions in the cockpit

Close date: the signal strip shows N days remaining or, in red, N days overdue, beside the Forecast category.
Buyer activity: Latest activity on the Cockpit tab, and the Activity tab. Engagement Momentum under Deal Health asks Are we actively working this?
The person who signs: Decision Readiness (Can they actually buy?) opens the People tab, where Roles to fill shows who’s missing.
Next step: the deal’s notes and Activity tab. If there’s no dated buyer step, that’s the gap.
First thing failing: the weakest Deal Health dimension, Gaps to close, and the Qualification card (click Details for the Deal Health tab).

The Deal guide button in the header adds priorities, a daily check, scenarios and risk tactics. See the opportunity cockpit, Deal Health Score explained and Deal guide.

3. Decide

If the evidence holds, agree the one action that fixes the first failing thing, and who owns it.
If it doesn’t, move it out of Commit, or qualify it out from Risk Signals. See Risk Signals.

From Claude or Codex

Two skills prepare the same conversation from either side of the table.

For the manager: deal-review-prep builds a deal review deck for one deal: a health verdict, the first constraint, framework scorecards, risks and blockers, stakeholder coverage, coaching questions and next steps, with the tool scores kept apart from its own assessment. Ask, for example, Review the <deal> deal. How healthy is it and should I be worried? or Give me the full picture on the <deal> deal for my pipeline review with the team.
For the account executive: ae-deal-inspection-prep builds a one-page sheet: current facts, honest gaps, the questions your manager is likely to ask and the evidence for your answers. Ask, for example, Prep me for my deal inspection on <deal>. What will my manager ask? or Help me answer questions on the <deal> deal before my 1:1 with my manager.

Both only read: they don’t update the deal, write notes or send anything. If a name matches more than one deal, they ask which you mean. See deal-review-prep and ae-deal-inspection-prep.

When the data is thin

No activity logged isn’t proof of a dead deal. It means nothing was recorded. Ask the rep, then log what happened.
Scores come from notes. Framework scores are extracted from the deal’s notes, including attached meeting notes. Thin notes mean thin scores, not a weak deal.
Unavailable, not zero. The skills report evidence they can’t reach as unavailable, and list it in their sources and limits.
Assessment, not fact. The deal review’s verdict and coaching questions are the reviewer’s judgement on the evidence. The scores beside them are shown exactly as Summit53 returned them.

Do it this week

List every deal in Commit.
For each, answer the five questions in one line each, with a date for every buyer action.
Circle the first question that fails, and write one action and one owner for it.
Move any deal that fails questions 1 and 4 out of Commit until the buyer gives a dated next step.

Troubleshooting

The framework scores haven’t changed after the meeting. Log what you learnt as a Note on the Activity tab; a new note starts a fresh extraction.
Win probability differs between the cockpit and the list. The cockpit shows the probability entered on the deal; the list shows the model’s prediction where there is one.
The skill asked which deal I meant. More than one deal matched the name. Pick one, or give the opportunity ID.
The deck says the signal layer is unavailable. The skill couldn’t read the deal’s signals, for example because of your access. The review falls back to the health and framework scores.

Is this Commit deal real? Inspect one deal on evidence