Sales Pipeline Audit · System 03 · deciding what happens next

Fix the Deal, or Fix the Process

A stalled deal with a named cause still needs a decision, and most of the time that decision belongs to the rep carrying it: get a next step booked, find the economic buyer, or let the deal go and take the time back. Every so often, though, the same cause turns up on the fifth deal this quarter, and at that point the finding isn’t about a rep or a buyer anymore. It’s about how the pipeline is run.

The idea in one line: fix the deal first, every time. Only fix the process once a cause has shown up often enough that fixing deals one at a time is treating the same finding as if it were new each time.

Two different judgments

Finding out why a deal stalled and deciding what to do about it are not the same skill, and doing both at once is how a stalled-deal review turns into either a blame session or a pep talk.

System 02 sorted each stalled deal into a cause: no next step, budget or authority never confirmed, indecision, champion turnover, or competitor selected. This system takes that list and asks a separate question of each one — not “what happened,” but “what changes on Monday, and who does it.” A finding without an action attached to it is a status update. It goes back in the CRM notes and nothing about the deal is different next week.

The per-deal decision

Every stalled deal gets one of these, chosen by the cause System 02 found for it.

Book the next step. For anything stalled only because nothing’s on the calendar. This is most of the list on a pipeline that hasn’t been reviewed before, and it’s the fastest fix on it.

Multi-thread. For budget or authority never confirmed, or a single champion carrying the whole deal. Get a second contact, ideally the economic buyer, into a conversation before doing anything else.

Requalify. For indecision or a weak value case. Re-run BANT or MEDDIC properly rather than pushing the same pitch again, and be honest if the answer is that the case was never strong enough to survive internal scrutiny.

Disqualify, and reclaim the time. For a champion who’s gone quiet with no replacement in sight, or a competitor loss the evidence actually supports. Close it, write the real reason, and free the rep to spend that attention on a deal that can still move.

Escalate. For anything large enough, or stuck enough, that it needs a manager’s relationship or authority rather than more of the rep’s time. Use this rarely — it doesn’t scale, and a list where everything gets escalated is a list where nothing actually gets decided.

When the same cause keeps showing up, it stops being a deal problem

Run the per-deal decision on enough stalled deals and a pattern usually surfaces on its own. That pattern is the more valuable finding.

If a fifth deal this quarter stalls with no dated next step, the rep didn’t forget five times. Something about how the pipeline is run makes it easy to log a call and skip the calendar invite — no prompt to do it, no consequence for not doing it, nobody checking. Coaching the fifth rep individually treats a system problem as five separate people problems, and it will produce a sixth deal next month.

The test is simple: has the same root cause appeared across multiple deals, multiple reps, or multiple months? If it’s one deal, fix the deal. If it’s a pattern, the fix moves up a level, to the process that lets the pattern keep happening.

Five ways to fix the process, cheapest first

In order. Stop at the first one that removes the pattern, the same discipline as choosing any other fix on this site.

Fix the definition. Write real entry and exit criteria for the stage where the pattern shows up, so “proposal sent” means one specific thing to everyone using it.

Fix the gate. Require the missing field or the next-step task before a deal can advance to the next stage, rather than trusting it’ll get filled in later.

Fix who owns it. Give the stalled-deal queue a named owner — a sales manager or RevOps, on a schedule — instead of leaving it to whichever rep happens to notice.

Fix the incentive. If reps are measured on pipeline volume rather than pipeline quality, padding the pipeline is the rational response to the measurement. Change what gets counted before blaming who’s counting it.

Automate the gate. Only once the four above have been tried and the pattern survives them. A workflow rule that blocks a stage change without a next step earns its place after the definition, the ownership, and the incentive are already right.

How to write up a decision

The shape of a decision

Finding: the deal or the pattern, and the cause behind it

Level: one deal / a pattern across several

Decision: book, multi-thread, requalify, disqualify, escalate — or, for a pattern, which of the five process fixes

Change: what’s different by name, and who owns it

Expected: what should be different, measured how, by when

Reversal: what result would tell you the diagnosis was wrong

A pattern-level decision changes the pipeline for every deal after it, so hold it to the same reversal discipline as a single deal’s decision — a gate that quietly makes the CRM slower to use is its own kind of cost.

A worked example

Closing out the 14-person services firm from Systems 01 and 02. Of the 14 stalled Proposal-stage deals: 6 get a next step booked this week, 5 get a multi-threading push, 1 gets requalified, 1 gets disqualified with a real reason recorded, and the champion-turnover deal gets a 30-day watch before a decision either way.

The pattern-level finding is the loss-reason field: 22 of 31 closed-lost deals this period carry “Other” with no note. That’s not five reps forgetting — it’s a field with no required options and no consequence for skipping it. The fix is rung one and two together: a short controlled list of real reasons, made required before a deal can close as lost, with an optional note for anything larger than a small deal. No new software, no dashboard, no rule about who’s allowed to close a deal.

Reversal line: if the “Other” rate is still above one in five closed-lost deals a quarter after the field becomes required, the field wasn’t the problem — reps are picking the fastest option in the new list too, and the next place to look is whether closing a deal quickly is rewarded over closing it accurately.

Systems 01 to 03 are one method in three passes: reconstruct what the CRM actually shows, name why each stalled deal is stalled, then decide what changes for the deal and for the pipeline it sits in. The complete system assembles them into something you can run start to finish on your own pipeline, with the evidence to gather first and a master prompt to paste into whatever assistant you already use.

Next in the Sales Pipeline AuditThe Sales Pipeline Audit 12 min read + a few hours to run