Sales Pipeline Stages: A Complete Guide with Templates

A sales pipeline stage is a point in your deal process defined by what the buyer has done, not by how your rep feels. If you can't write down the single observable fact that moves a deal from one stage to the next, you don't have a pipeline; you have a mood ring that the whole company forecasts off of.

This guide gives you the canonical stages, the exit criteria most teams skip, and a template you can paste into any CRM. The one opinion baked into it: you need fewer stages than your CRM's default template suggests, and every stage you add is a place a deal goes to hide.

The canonical stages, and what each one is for

Almost every B2B pipeline is a variation on the same spine. The names change; the jobs don't. Here is the full set, with the buyer action that defines each and the exit criterion that lets a deal leave it.

Stage	Buyer action (what they did)	Exit criteria (all must be true to advance)
Lead / Prospect	Showed up — filled a form, replied to outbound, matched your ICP	Contact reachable; fits your ideal-customer profile on the two or three firmographics that actually predict a fit
Qualified	Agreed there's a problem worth a conversation	You've confirmed a real need, a rough budget range, and that you're talking to someone with a path to the decision
Discovery	Gave you access to the people and facts	You understand their use case, current solution, timeline, and the metric they're trying to move; buyer agreed to a next step
Proposal	Asked to see what it costs and how it works	A specific offer — scope, price, terms — is in the buyer's hands, and they've acknowledged receiving it
Negotiation	Started haggling scope, price, or terms	Verbal agreement on commercials; only signature, legal, or procurement steps remain
Closed Won	Signed	Contract executed and the deal is bookable revenue
Closed Lost	Said no, or went dark past your recycle threshold	Reason recorded from a fixed list; recycle date set if the deal can come back

Read that exit-criteria column again, because it is the entire article. Everything else is packaging.

The part every team skips: exit criteria

Most pipelines define stages by name and stop there. "Discovery" sits in the CRM with no definition, so it means whatever the rep dragging the card believes it means. One rep's Proposal is another rep's Discovery. Now your forecast is an average of private definitions, and nobody can tell you why the number moved.

An exit criterion fixes this. It is a fact, verifiable by someone who wasn't on the call, that must be true before a deal advances. Not "the rep feels good," but "a scoped proposal exists and the buyer confirmed they got it."

Write exit criteria as things a skeptic could check. Compare:

Weak: "Deal is qualified." (By whom? On what basis?)
Strong: "Confirmed a named business problem, a budget range, and a contact with a path to the decision-maker."
Weak: "In proposal." (A draft in your Google Docs? A number said out loud?)
Strong: "A written proposal with scope, price, and terms has been sent, and the buyer acknowledged receipt."

The test for a good exit criterion: could two different reps, looking at the same deal, disagree about whether it's met? If yes, it's not a criterion yet; it's a vibe. Tighten it until the answer is no.

This is also the discipline that makes a CRM worth paying for. A pipeline without exit criteria doesn't need software; a spreadsheet holds vibes just fine. The reason to enforce stages in a tool is so that "advance to Proposal" can require a field to be filled or a document attached. If you're shopping for one that can hold you to that, the best CRM for small business weighs which tools let you gate stage progression versus which just let you drag cards around, and our Pipedrive review covers a tool built specifically around a stage-and-activity model.

How many stages you need (fewer than you think)

The destackd position: the default CRM template ships with too many stages, and each extra one costs you more than it gives.

Every stage you add is a place for a deal to sit and look busy without moving. A rep who doesn't want to admit a deal is stuck can nudge it from "Discovery" to "Deep Discovery" to "Technical Validation" and generate the appearance of progress across three stages while the buyer does nothing. More stages means more surface area for that theater.

More stages also means more disagreement about where a deal belongs, which is the problem exit criteria were supposed to solve. If the gap between two adjacent stages is so thin that reps flip a coin, merge them.

The working rule: you need a stage only if a deal can be objectively in it and not the one before it, and the distinction changes what the rep does next. If two stages prompt the same next action, they're one stage wearing two hats.

For most B2B teams, that collapses to something like:

Qualified — worth spending time on
Discovery — actively working, we understand the deal
Proposal — the offer is out
Negotiation — closing the commercial gap
Closed Won / Closed Lost

Lead/Prospect often lives better in your marketing tool or a pre-pipeline holding pen than as a formal stage; a lead nobody has qualified isn't a deal yet, and parking un-worked leads in the pipeline inflates it with noise. Add Negotiation as its own stage only if your deals have a distinct haggling phase; plenty of smaller deals go from Proposal straight to signed, and a Negotiation stage those deals skip is just clutter.

Resist the "stage per sales-methodology step" temptation. Your qualification framework can live as fields inside Discovery. It does not need to be five stages that turn your board into a status-theater generator. This is the same over-tooling reflex that inflates the whole stack, and the cure is the same too: if you already run a stack, a sales tech stack audit applies the identical logic to your tools, and the minimum viable revenue stack is the build-from-scratch version.

Stage-based forecasting, and its limits

Once stages have real exit criteria, you can forecast off them. The basic mechanism: assign each stage a historical win rate, multiply by deal value, and sum. A deal in Negotiation counts more than one in Discovery because deals at that stage have historically closed more often.

This works, but only as well as your inputs, and it fails in specific, predictable ways:

Stage-weighted forecasting assumes your stages mean the same thing they did last quarter. If reps have quietly gotten looser about what "Proposal" means, your historical win rate for Proposal is now attached to a stage that admits weaker deals. The number drifts and nobody knows why. Exit criteria are what keep the definition, and therefore the forecast, stable over time.
It ignores time-in-stage. A deal that's been in Negotiation for ninety days is not the same bet as one that entered yesterday, even though stage-weighting treats them identically. Track age-in-stage alongside the weight, and set a threshold past which a stalled deal gets flagged rather than counted at full weight.
It rewards stuffing. If reps are measured on pipeline coverage, they're incentivized to advance deals to keep the weighted number up. That's the theater again, now with a financial motive. The defense is the same: advancement requires a checkable fact, not a drag of the mouse.

For an early-stage team, stage-weighted forecasting in the CRM is enough, and a dedicated forecasting tool is a solution to a problem you don't have yet. It earns its place when you have enough deals per quarter for the weights to be statistically meaningful and enough complexity that a spreadsheet can't hold the picture. Clari vs Gong covers that tier, including when you're still below it. And if your team is big enough to be asking who owns pipeline definitions and forecast hygiene in the first place, our RevOps guide covers the function that should.

The copy-paste template

Here is a pipeline you can drop into any CRM today. Adapt the exit criteria to your motion; the point is that each one is a fact, not a feeling.

Stage 1: Qualified

Advance when: a named business problem is confirmed, a budget range is known, and you're engaged with someone who has a path to the decision.
Required fields: problem statement, budget range, primary contact + role.

Stage 2: Discovery

Advance when: you understand their use case, current solution, timeline, and the metric they want to move; a concrete next step is booked on the calendar.
Required fields: success metric, current solution, timeline, next meeting date.

Stage 3: Proposal

Advance when: a written proposal with scope, price, and terms has been sent and the buyer has acknowledged receiving it.
Required fields: proposal document attached, amount, expected close date.

Stage 4: Negotiation

Advance when: commercial terms are verbally agreed and only signature, legal, or procurement remains.
Required fields: agreed amount, blockers-to-signature list.

Stage 5: Closed Won / Closed Lost

Won: contract executed.
Lost: reason selected from a fixed list; recycle date set if reopenable.

Three rules that keep this template honest after you paste it:

A deal can skip forward but not backward without a note. Skipping ahead is fine, since some deals arrive ready to buy. A deal moving back a stage is a signal worth capturing: what changed?
No stage without an exit criterion. If you add a stage later and can't write its criterion, that's your answer about whether it should exist.
Closed Lost reasons come from a fixed list, not a free-text box. Free text is where "lost reasons" go to become un-analyzable. A short dropdown you can count is worth more than a rep's paragraph.

That's the whole system. Fewer stages than the default, a checkable fact at every boundary, and a forecast that inherits its trustworthiness from those facts. The goal isn't a prettier board. It's a pipeline where anyone can look at a deal and know exactly what has to be true for it to move, without asking the rep how they feel about it.

Frequently Asked Questions

How many sales pipeline stages should I have?

For most B2B teams, four to six: Qualified, Discovery, Proposal, optionally Negotiation, and Closed Won/Lost. Add a stage only when a deal can be objectively in it but not the one before, and the distinction changes what the rep does next. If two stages trigger the same action or reps can't reliably tell them apart, merge them.

What are exit criteria in a sales pipeline?

Exit criteria are the specific, verifiable facts that must be true before a deal advances to the next stage, for example, "a written proposal with scope and price has been sent and acknowledged." They replace subjective judgment with a checkable standard, so two different reps looking at the same deal would agree on where it sits. They are the single most-skipped and most-valuable part of a pipeline definition.

What's the difference between a lead, a prospect, and an opportunity?

A lead is an unqualified contact who showed interest; a prospect is a lead who fits your ideal-customer profile but hasn't been worked yet; an opportunity is a qualified deal that's earned a place in the pipeline. The practical takeaway: keep unqualified leads out of the pipeline itself so they don't inflate your numbers. Promote a contact to an opportunity only once it clears your Qualified exit criteria.

Should "Negotiation" be its own pipeline stage?

Only if your deals have a genuinely distinct haggling phase. Many smaller or transactional deals go straight from Proposal to signed, and for those teams a Negotiation stage is just clutter that most deals skip. If your commercial back-and-forth is substantial enough to change what the rep does, make it a stage; otherwise fold it into Proposal.

How does stage-based forecasting work?

You assign each stage a historical win rate, multiply it by each deal's value, and sum across the pipeline, so a deal in Negotiation counts for more than one in Discovery. It's only as reliable as your stage definitions, so it depends on real exit criteria to keep those definitions stable. Track time-in-stage alongside the weights, because a deal stalled for months is a worse bet than the raw stage weight implies.

Can my CRM enforce pipeline stages automatically?

Most modern CRMs can require specific fields or attachments before a deal advances, which is the mechanism that turns exit criteria from a document nobody reads into a rule the tool enforces. This gating is worth checking for when you choose a CRM, since it's the difference between stages that mean something and cards you can drag anywhere. A pipeline without enforcement is just a spreadsheet with nicer colors.