Sales pipeline stages: the six stages, and the fact that has to be true to enter each one

Sales pipeline stages vs sales process steps, and pipeline vs funnel

Sales pipeline stages describe where a deal is; sales process steps describe what the seller is doing. The two get written with the same word and they are not the same object, which is why so many sales pipeline setups end up with a stage called "follow-up" that nobody can enter or leave on a rule.

A stage is a position. A deal is either in it or not, and the test should be a fact somebody else could verify from the record: a meeting happened, a proposal was sent, a budget holder is named. An activity is something a person does, and several activities can happen inside one stage without moving the deal anywhere. If you want the activity view, the seven things a seller actually does from first contact to close, that is our guide to sales process steps. This page is the other half: where the deal sits, what has to be true to get there, and what each position predicts.

Pipeline and funnel are also not the same thing. A funnel is a marketing view of a population getting smaller: visitors, leads, customers, drawn as percentages. A pipeline is an operational list of named deals with a value and a date, and its shape is not necessarily a funnel: a healthy pipeline can bulge at negotiation if three large deals are all waiting on procurement. The short definition lives in our sales pipeline glossary entry; we will not repeat it here.

The six sales pipeline stages, their entry criteria and their CRM fields

Six sales pipeline stages cover almost every B2B pipeline: a verified record, a qualified deal, a held meeting, a sent proposal, an active negotiation, and a closed outcome with a coded reason. Each row below gives the fact that has to be true to enter, the fields that have to be filled for the stage to mean anything, and a starting conversion rate you should replace with your own after one full cycle.

Two rules make sales pipeline stages work rather than decorate a slide. The first is that a deal cannot sit between two stages: if the fact is not true, the deal stays where it is, however encouraging the last call felt. The second is that the fields are not admin, they are the stage: a proposal stage with no amount cannot be forecast, so the deal is not really in it.

Six sales pipeline stages is a deliberate number. Five is common and works. Seven appears when teams split negotiation into legal and procurement, which is worth doing only if those two genuinely have different owners and different durations. Above eight, the weekly review stops being possible in an hour, and stages start to be chosen by feel.

Stage 0: the verified record that earns a place in the pipeline

Stage 0 is the one almost no pipeline diagram draws, and it is where most forecast error is born: a record enters the pipeline before anyone has checked that the company fits, the person is still there, and the address works. Everything downstream inherits that.

The entry test for stage 0 is three facts, and they are cheap to check. The company matches your ICP on the criteria you actually wrote down, which is the subject of our guide to the ideal customer profile in sales. The person holds a role that can influence the decision, not just a role that will answer politely. And the contact details resolve, because a deal you cannot reach is not a deal, it is a row.

This is the one place on this page where Derrick belongs, and it is worth being precise about what it costs. Pulling the people from a list of target accounts is Import Leads from Target Companies at 1 credit per lead. Filling in the company side is Enrich Companies at 1 credit per company, and the person side is Enrich Leads at 1 credit per profile. All three are available on the free plan of 100 credits per month, at zero euros and with no card, from the Derrick sidebar in Google Sheets, and all three need a connected LinkedIn session to run. Find Duplicates costs no credits at all and is what stops the same account entering your pipeline twice under two spellings. If you would rather the enrichment ran inside the CRM than in a spreadsheet, the same operations are on the REST API and on an MCP server for an AI assistant, both from the Standard plan at 20 euros a month.

The arithmetic matters more than the list of names. Taking 100 accounts through company enrichment and then profile enrichment is two steps, so 200 credits, which is more than the free plan gives you in a month: the free tier covers about 50 records through both steps, not 100. And checking the address is Email Verification, at 1 credit per email, which sits on the paid plans from 9 euros a month and is billed only when a result comes back. No free credit can be spent on it. Count the steps, not the features, before you plan a batch.

Where the records come from is a separate job. Our guide to building a prospect list covers sourcing end to end, and if your deals arrive inbound rather than outbound, the five stages a form fill goes through before it deserves stage 1 are in inbound lead management.

Stages 1 to 3: qualified, met, proposed

These three stages are where the fact-based test earns its keep, because they are the three most often entered on optimism.

Stage 1, qualified. The fact is that a need or a trigger has been stated by the prospect, in their words, and that the person you are talking to can influence the decision. "Seemed interested" is not a fact. A renewal date, a new hire on the team that owns the problem, a stated deadline: those are. The field that proves it is the one holding the trigger, written as the prospect said it. If that field is empty across half your stage 1, your qualification standard lives in people's heads.

Stage 2, meeting held. Note the word held. A booked meeting is not a stage, it is a hope with a calendar entry, and pipelines that count bookings overstate themselves by exactly the no-show rate. The entry fact is that the meeting happened; the fields are who attended and what the next step is, with a date on it. A stage 2 deal with no dated next step is the single most reliable predictor of a deal that will quietly stop.

Stage 3, proposal sent. The fact is that a document containing a price is with the customer. Not drafted, not discussed verbally, sent. The fields are the amount, the expected close date and what was actually proposed, because three months later nobody remembers which version won. This is the first stage where a deal carries a number you can forecast on, which is why the temptation to enter it early is strongest here.

Stages 4 and 5: negotiation, and closing with a coded reason

Stage 4 is entered when a budget holder is identified and terms are being discussed, and stage 5 is entered only with an outcome and a reason code.

Stage 4, negotiation. The fact is two-part: somebody who controls the money is named in the record, and the conversation has moved to terms instead of value. Deals sit here longest and hide the most. The fields that matter are the confirmed budget, the decision process as the customer described it, and the current blocker. A negotiation stage with no named blocker usually means nobody asked.

Stage 5, closed. Won or lost, and lost needs a reason code from a short closed list rather than free text. Four or five codes are enough: price, timing, no decision, competitor, disqualified. The value of the code is not the individual deal, it is the distribution six months later, which tells you whether you have a pricing problem or a qualification problem. Free text produces a field nobody can count.

One stage worth adding only if somebody owns it: post-sale. Some pipelines add an onboarding stage after the signature, and it is useful when the same team is accountable for the first ninety days. If nobody is, it becomes a stage deals enter and never leave.

How to build a sales pipeline from your real cycle

Build it backwards from deals you have already closed, in five steps, instead of forwards from a template. A sales pipeline copied from a blog post describes somebody else's sales cycle, and the stages nobody can enter are the ones that were never true of your deals.

• Take your last ten closed deals and write what actually happened. Not what should have happened. The positions that appear in all ten are your stages; the ones that appear in three are somebody's habit.
• Write the entry fact for each stage as a sentence someone else could check. If two people would disagree about whether a given deal qualifies, the sentence is not finished.
• Decide the required fields per stage, and keep the list short. Three fields that get filled beat eight that get skipped. Required means the deal cannot move without them.
• Set starting conversion rates and durations, and mark them as guesses. They are there so the first weekly review has something to compare against. Replace them with your own after a full cycle.
• Decide who is allowed to move a deal forward, and put the weekly review in the calendar. A sales pipeline with no scheduled review degrades into a list of open tabs within a quarter.

Two things to settle before any of that: how long your real cycle is from first contact to signature, and whether outbound and inbound deals follow the same path. If they do not, you have two pipelines, and pretending otherwise is what makes the pipeline conversion rates meaningless. Our outbound sales guide covers the sequence that feeds the outbound one.

Sizing the top: how many contacts a target of N deals needs

Work backwards from signed deals through your own rates and the top of the pipeline sizes itself. The calculator below runs that chain in reverse, stage by stage, and ends on the number of raw records you have to source to keep stage 0 filled each month. The same arithmetic, written from the outbound side rather than the pipeline side, is in our guide to B2B lead generation.

Run it once with last quarter's real rates, then again with the unusable share at 2 percent. The gap between the two raw-contact figures is what the quality of stage 0 is worth every month, in the only unit that survives a budget conversation.

The metrics that make sales pipeline stages worth having

Five numbers come out of a sales pipeline and none of them exists without the stages: stage-to-stage conversion, time in stage, weighted value, velocity, and coverage.

• Stage-to-stage conversion. Deals leaving a stage forwards, divided by deals that entered it. Read it per stage, not as one win rate: a healthy total can hide one stage where three quarters of deals die.
• Time in stage. The median, not the average, because one deal stuck for nine months moves an average and tells you nothing. Compare each stage against its own typical duration, and treat double that as a flag.
• Weighted value. Each deal's amount multiplied by the historical conversion of the stage it sits in. The method and its failure modes are in our weighted sales pipeline entry.
• Velocity. How much value moves through the pipeline per unit of time, which is the number that improves when you shorten a stage rather than when you add deals. Definition and formula: sales pipeline velocity.
• Coverage. Open pipeline value divided by the target for the period. Teams commonly aim for three to four times, and the right multiple for you is simply the inverse of your own conversion from the stage your deals sit in.

What to do with them week to week, and which of them belong in a report a manager reads, is covered in sales pipeline management and sales pipeline reporting.

The weekly review: five questions per stage

A sales pipeline is only worth the hour a week somebody spends walking it, and the walk is five questions per deal, not a status update.

Ask, for every deal in a stage: what is the fact that put it here, what is the dated next step, what has changed since last week, what is the one thing that would kill it, and does the close date still hold. Five answers, thirty seconds a deal. A deal that fails two of the five goes back a stage or out.

Zombies. Every sales pipeline accumulates deals that nobody will ever close and nobody will ever remove, because removing one feels like admitting something. Set a rule instead of relying on courage: no dated next step and no contact in thirty days means the deal is closed as no decision, with a reason code. It can always be reopened. A pipeline that only grows is not a forecast, it is a scrapbook.

The forecast. Take the weighted value, subtract anything with a close date inside the period but no dated next step, and compare against coverage. If the committed number relies on deals that have not moved in three weeks, say so out loud in the review rather than in the following month.

A worked example: 100 records through six stages

Here is the shape of a small B2B SaaS sales pipeline with 100 verified records entering stage 0 in a month, using the starting rates from the table and a four week average cycle per stage. The numbers are illustrative arithmetic, not a benchmark: their purpose is to show which stage carries the loss.

Read the table for its shape, not its numbers. Of the 97 records that never sign, 85 are gone before a meeting is even held, which is the part most pipeline reviews spend the least time on because it is not where the revenue is visible. And the cycle length matters as much as the rates: at four weeks per stage, a record entering stage 0 in January signs in May, so a quarter that starts with an empty stage 0 was decided a quarter earlier.

Common mistakes with sales pipeline stages

Five mistakes account for most broken sales pipelines, and four of them are the same mistake: a stage defined by something other than a verifiable fact.

• Stages named after an action. "Email sent", "following up", "nurturing". Nobody can say when a deal leaves them, so nothing ever does.
• Too many stages. Nine stages means nine judgement calls per deal per week. The review becomes impossible and people stop updating the record, which removes the only data the stages existed to produce.
• A pipeline inflated with unqualified deals. Coverage looks healthy, conversion looks terrible, and the diagnosis is wrong: it is not that sellers close badly, it is that stage 1 lets anything in.
• No exit rule. Deals enter and never leave, and the pipeline grows every quarter regardless of performance. See the zombie rule above.
• Rates copied from an article. Including the ones in the table on this page. They exist to give the first review something to argue with, and they should be replaced by your own after one full cycle.

If the part that is missing is the records at the top rather than the stages themselves, that is the cheapest thing to fix on this page. Derrick fills a company and a contact record at 1 credit each from the Google Sheets sidebar, and the free plan gives you 100 credits per month, enough to take about fifty accounts through both steps. Start with the free plan, or read the rest of our B2B marketing guides first.

We also publish one email every 2 weeks with what we are measuring on pipeline and prospecting data. Subscribe from the homepage.