Pipeline Exit Criteria: Why 'Proposal Sent' Isn't a Real Stage on Its Own

A financial advisory client's pipeline had a stage called "Proposal Sent" with fourteen deals sitting in it, some for three days and some for four months. Every one of those deals was reported the same way in the weekly forecast, at the same probability, even though the three-day-old ones were warm and the four-month-old ones were almost certainly dead. The stage name described an activity someone had completed, not a state the deal was actually in; and that is the root cause of most pipelines that look healthy in a report and hollow in reality.

The difference between an activity stage and an exit-criteria stage

An activity-based stage is named after something the salesperson did: "Called," "Demo Scheduled," "Proposal Sent." The problem is that completing an activity doesn't tell you whether the deal advanced, because you can send a proposal into a dead deal just as easily as into a live one. An exit-criteria stage is defined instead by a condition that has to be true, verified from the buyer's side, before a deal can move forward. "Proposal Sent" becomes something closer to "Buyer has confirmed budget and reviewed proposal with their decision-maker," a state, not a task checked off.

Writing exit criteria for a typical service-business pipeline

A workable pipeline usually has five to seven stages. A common structure, with exit criteria attached to each stage:

1. New Lead → exits when contact information is verified and a need has been confirmed in a first conversation, not just when a form was submitted.
2. Qualified → exits when budget range, decision timeline, and decision-maker access have all been confirmed.
3. Needs Assessment/Discovery → exits when the specific problem, its cost to the buyer, and the criteria they'll use to decide have been documented.
4. Proposal/Quote → exits when the buyer has reviewed the proposal with the actual decision-maker and given a specific objection or a specific next step with a date.
5. Negotiation → exits when terms are agreed verbally or in writing, pending only paperwork or signature.
6. Closed Won/Closed Lost → a terminal state, with a required reason logged for lost deals.

Notice that most of these exit criteria require the buyer to have done something: confirmed a budget, reviewed with a decision-maker, or given a specific answer, not just the salesperson. That's deliberate: buyer-side action is a far more reliable signal of real progress than seller-side effort.

Why this changes the forecast, not just the CRM

Once every stage has a hard exit condition, a deal sitting in a stage for an unusually long time becomes a visible, actionable signal instead of invisible clutter: it should have moved by now if it were actually progressing. That's what lets a sales manager run a pipeline review by asking "what's the specific blocker on this deal" instead of "how's this one going," which is a much harder question for a rep to dodge with a vague answer.

A stage without an exit criterion is a place deals go to get stuck.

Common design mistakes

• Too many stages, which makes moving a deal forward feel like a chore and encourages reps to skip stages rather than update them accurately.
• Stages defined by internal process instead of buyer behavior: "Internal Review" tells you what your team is doing, not what the buyer has confirmed.
• No required reason code on Closed Lost, which throws away the single most useful data point for improving the earlier stages of the pipeline.
• Probability percentages assigned to stage names that were never validated against actual historical win rates from that stage.

Retrofitting exit criteria onto an existing pipeline

You don't need to rebuild a CRM pipeline from scratch to fix this. Start by writing exit criteria for the stages you already have, even if the stage names stay the same for now. Then audit the current pipeline against those new criteria; you will likely find deals sitting in stages they do not actually qualify for, which is useful information on its own about how the team has been using the stages.

This applies whether the deal cycle is two days or six months

A home services business closing same-week jobs and a law firm running a multi-month intake-to-retainer cycle both benefit from the same underlying discipline, even though their stage names and cycle lengths look completely different. The exit-criteria approach scales down to a short cycle and up to a long one because it's about defining progress by evidence, not by elapsed time or activity count.

We build pipeline stage design into every CRM implementation, including inside NetWebMedia's own CRM, because a pipeline with vague stages produces a forecast nobody can trust, no matter how good the tooling around it is.

← Back to all articles