The five pipeline stages a small sales team needs
Too many stages invent precision you do not have; too few hide where deals die. Five stages, and transition rules two people would agree on.
· 5 min read
Why the number of stages is the real decision
A pipeline is not a picture of your sales process. It is a set of buckets you agree to sort deals into, and the only reason to have buckets at all is so that a question like "where are we losing people?" has an answer somebody can act on this week. That makes the number of stages a design decision rather than a description of reality, and it is the decision small teams get wrong in one of two opposite directions.
Too many stages produce false precision. A pipeline with eleven stages implies you can reliably tell technical evaluation from commercial evaluation, and in a three-person team you usually cannot — so deals get filed by whoever last touched them, two people file the same situation differently, and the stage report becomes a record of filing habits rather than of deal progress. Too few stages hide the opposite problem: a pipeline with only open and won cannot tell you whether deals die before a proposal exists or after one has been sent, and those are two completely different problems with two completely different fixes.
The five stages, and what each one actually means
Lead means someone whose contact details you have and whose interest you have not tested. It is a holding area, and it should feel like one. Qualified means you have had a real conversation and established that there is a need you can serve, a budget that could plausibly cover it, and a person who can either sign or walk the decision to whoever signs. Proposal sent means a written number is in their hands — not discussed on a call, not promised by end of week, actually sent.
Negotiation means they have responded to the proposal and are arguing about something: price, scope, timeline, terms. The distinction from proposal sent matters more than any other boundary in the pipeline, because silence after a proposal and pushback on a proposal look similar in a spreadsheet and mean opposite things. Closed is won or lost, recorded either way, with the reason attached. Five buckets, each one answering a question the previous one cannot, and none of them requiring a judgement call that two reasonable people would make differently.
Transition rules: the test two salespeople would both pass
A stage definition is worth having only if it is a test rather than a feeling. "Seems interested" is a feeling. "Has told me what they are currently spending on this, or told me they have no current spend and a budget owner" is a test. Write the test down for each of the four boundaries, in one sentence, in language specific enough that a new hire in their second week reaches the same conclusion as the person who has been there three years.
The practical way to find where your definitions are vague is to take ten live deals, have two people stage them independently without conferring, and compare. Every disagreement is a boundary whose definition is missing a word. This is unglamorous and it takes an hour, but it is the difference between a pipeline you can reason about and a pipeline that produces a confident-looking number nobody in the room actually believes. Do it once when you set the stages up, and again the first time two people argue about a forecast.
What a stage is not
A stage is not a task list. The temptation is to name stages after things your team does — "demo booked", "site visit done", "samples sent" — because those are visible and easy to tick off. The problem is that activity stages measure your effort rather than the buyer's progress, and a deal can accumulate four completed activities without moving any closer to a decision. Track activities as activities; keep stages for what the buyer has committed to.
A stage is also not a probability. Assigning seventy per cent to negotiation and thirty per cent to proposal sent feels like forecasting, but the percentages are usually invented once and then never checked against what actually closed. If you want probability, derive it from your own history — what share of deals that reached negotiation last year were eventually won — and accept that with a small number of deals the figure carries real uncertainty. A number you calculated from your own closed deals is worth something. A number you copied from a template is decoration.
Reading the pipeline: where deals actually stop
Once the boundaries are tested rather than felt, the pipeline starts answering questions. A pile-up at the lead boundary means you are generating contacts nobody is calling, which is a capacity or a priority problem, not a lead-quality problem. A pile-up between qualified and proposal sent usually means proposals are slow to produce, and the fix is a reusable structure rather than more discipline. A pile-up at proposal sent means proposals are going to people who cannot decide, or answering a need the buyer does not actually have.
The useful habit is to look at the boundaries rather than the totals. Total pipeline value is the number everyone quotes and the least informative one available, because it moves when a single large deal is added and tells you nothing about whether anything is progressing. Counting how many deals crossed each boundary in the last two weeks is a smaller number and a far more honest one — it is the only view that distinguishes a pipeline that is growing from a pipeline that is merely accumulating.
When to add a stage, and why the answer is usually not yet
There is a legitimate reason to add a stage: a boundary is hiding a distinct failure mode that you keep needing to diagnose by hand. If deals routinely stall specifically while waiting for a security review or a legal redline, and you find yourself opening records one by one to work out which ones are stuck there, that is a stage asking to exist. The test is whether the new stage would change a decision, not whether it would describe reality more finely.
Everything else is best resisted, because each extra stage costs accuracy in a way that is easy to underestimate. Every boundary is another judgement call, every judgement call is another chance for two people to file the same deal differently, and the cost compounds quietly — the pipeline keeps producing a report, so nothing looks broken. A five-stage pipeline that everyone stages identically is more useful than a nine-stage pipeline that is technically more descriptive and practically a record of who touched what last.
Common questions
Should a lead that goes quiet move backwards through the stages?
No — move it out rather than back. Reversing stages destroys your ability to measure how long deals take, because the same deal then crosses the same boundary several times. Either close it as lost with the reason recorded, or park it somewhere outside the active pipeline. A pipeline that only moves forward is a pipeline whose stage timings mean something.
What if a deal skips a stage entirely — a referral who wants a quote immediately?
Record it as passing through the stages it genuinely passed, on the same day if that is what happened. A referral who arrives pre-qualified really did clear the qualified bar; there is no need to invent a waiting period. What you should not do is leave the stage blank or backfill it later, because a same-day transition is real data about how referrals behave.
Do the same five stages work for a business selling to consumers rather than companies?
The authority question mostly collapses, since the buyer is usually the decision-maker, so qualified becomes a test of need and budget alone. The five buckets still hold, but if your sales cycle is a single conversation the whole pipeline may not earn its keep — pipelines pay off when a deal spans multiple contacts over days or weeks.
How often should the stage definitions themselves be revisited?
Rarely, and only with a specific reason. Changing definitions breaks comparison with everything recorded before the change, so treat it like a schema migration rather than a tidy-up: note the date the definition changed, and expect that stage-timing figures either side of that date are not directly comparable.
Related pages