Opportunity Stage: Exit Criteria That Describe What the Buyer Did

An opportunity stage is where a deal sits in a defined sales process. The stages themselves are only labels; what gives them meaning is the exit criteria attached to each, the specific conditions that must be true before a deal is allowed to move on. Done well, a pipeline becomes readable at a glance and comparable across salespeople, conversion between stages can be measured, and a forecast can be challenged with evidence rather than tone of voice. Done badly, stages become a rough proxy for how optimistic the deal owner feels, and every report built on them inherits that optimism.

The buyer-action rule

Every exit criterion should describe something the customer did. Proposal sent is a seller action and can be achieved on a dead deal by a diligent salesperson. Proposal reviewed with the budget holder and a decision date agreed is a buyer action and cannot. Applying this single rule to an existing pipeline usually moves a noticeable share of deals backwards, which is uncomfortable and precisely the point.

Keep the number small

Four to six stages per sales motion is enough for most businesses. A stage earns its place by implying a different next action; stages that do not simply create disagreement about where a deal belongs, and that disagreement destroys the conversion data the stages were built to produce.

Separate pipelines for separate motions

A short self-serve motion and a long enterprise one have different steps and different buyer actions. Forcing them into one set of stages produces a pipeline where half the deals skip stages and the other half sit in stages that do not apply, and the reporting describes neither.

A worked stage design (illustrative)

Qualified: the buyer has confirmed a problem in their own words and agreed a next meeting. Discovery: the buyer has described how the process works today, what the problem costs, and who else is involved. Evaluation: the buyer has seen a demonstration tailored to what they described and confirmed it addresses their situation. Proposal: the proposal has been reviewed with the person controlling the budget and a decision date is agreed. Negotiation: commercial terms are being discussed and the approval and paperwork steps are mapped with dates. Five stages, each defined by something the customer did. Notice that a salesperson cannot advance a deal through any of them alone, which is the property that makes the pipeline worth reading.

The first use is diagnosis. When a business converts poorly, the aggregate win rate says nothing about where the problem is. Stage conversion localises it: deals that never leave discovery point at qualification or at the value case, deals that stall at proposal point at the decision process, and deals lost in negotiation point at commercial terms or at a competitor. Each of those has a different remedy, and without stages the team will try all three at once.

The second is forecasting. A stage with credible criteria carries real information about likelihood, which is what makes a forecast reviewable. The third is coaching: a manager can look at a rep whose deals consistently stall at one stage and know which conversation to work on, rather than offering general encouragement.

Seller-activity stages

Naming stages after what the seller did is the most common error and the most damaging, because it makes a busy quarter look like a productive one. Every stage reachable by unilateral effort is a place where dead deals accumulate while appearing to progress.

Probabilities nobody derived

Default probability percentages are round numbers supplied by a software vendor who has never seen your business. Multiplying deal values by them produces a weighted pipeline with a decimal point and no predictive content. Either calculate the actual historic conversion from each stage or do not publish a weighted figure at all.

Stages without expected durations

A deal that has sat in evaluation for five months is functionally lost, but it is technically open, so it keeps its value in the pipeline and its place in the forecast. Setting an expected duration per stage and reporting anything that exceeds it turns those deals into a decision rather than a permanent presence.

Look at conversion between adjacent stages rather than at the win rate. A pipeline can look healthy in aggregate while one transition converts at a fraction of the others, and that transition is where the process is failing. Look also at the distribution of deals across stages: a pipeline heavily loaded at the front is a qualification problem waiting to happen, while one loaded at the back with old deals is a closing problem already in progress.

Pay attention to time in stage as well as count. Two deals in evaluation are not equivalent if one entered last week and the other four months ago. And treat backwards movement as information rather than embarrassment: frequent regression from a particular stage almost always means its exit criteria are too easy to satisfy and should be tightened.

These four are all ways of describing where a deal stands and they are not interchangeable.

The distinction between stage and forecast category is worth holding on to. Stage describes how far a deal has progressed; forecast category describes whether the owner expects it to close in this period. A late-stage deal can sit outside the commit if the buyer's timing has slipped, and collapsing the two loses that information entirely.

Stage discipline is what makes every other pipeline measure trustworthy. Start with what a CRM is, then look at lead management software, sales automation for keeping deals moving, AI CRM for spotting stalled deals, CRM versus Excel for why stage timestamps cannot live in a spreadsheet, and sales use cases.