Sales Pipeline: How to Build and Manage One

For sales leaders, bid managers and pre-sales teams who already run a pipeline and keep seeing the same deals sit in the same stage.

A sales pipeline is a staged view of every open deal, from first contact to signed contract, that tells a sales team what each deal needs next. Building one means defining the stages your buyers actually move through, then writing a clear exit criterion for each stage. Managing one means acting on the deals that have stopped moving, which in most B2B teams happens right after a proposal or a formal response goes out. The stage count matters far less than knowing what has to be true before a deal leaves a stage.

Key takeaways

• A sales pipeline tracks open deals by stage. A sales funnel tracks conversion rates across those stages. They answer different questions and are not interchangeable.
• There is no correct number of pipeline stages. HubSpot ships seven by default and Salesforce ships ten. The count is a design choice, and six stages is a workable default for most B2B teams.
• What matters is the exit criterion: one sentence per stage, written as something the buyer has done and observable by someone who is not on the deal. "Proposal sent" is an activity. "The buyer has reacted to the proposal" is evidence.
• Most B2B deals go quiet right after the proposal goes out, when the document is read by people who were never in your meetings. The fix is a better-aimed document sent faster, not more follow-up. At Open, a framework proposal now takes five minutes instead of two to three hours, which leaves the time for adapting it to the client.
• Review three numbers every week: pipeline velocity, stage-to-stage conversion and pipeline coverage. A deal that has not moved in 30 days gets re-qualified, gets the artefact it is missing, or gets marked lost with a reason.

What is a sales pipeline?

A sales pipeline is a staged representation of every deal your team currently has open, organized by how close each one is to closing. Each stage answers one question: what has already happened, and what has to happen next.

The word gets used loosely, and that is where most confusion starts. A pipeline is about deals and the actions your team takes on them. A funnel is about volume and the rate at which prospects drop off between one step and the next. You can run a healthy pipeline and still have a leaky funnel, and the fix is different in each case.

Sales pipeline vs sales funnel

Both views matter. The pipeline is the working document. The funnel is the diagnostic. If your team argues about which one is "the real number," you probably have two teams looking at two different objects and calling them the same thing.

The distinction has a practical consequence. Funnel problems are solved upstream, with better targeting, better messaging or more volume at the top. Pipeline problems are solved deal by deal, by giving a specific opportunity the specific thing it is waiting for. Reaching for a funnel fix when you have a pipeline problem is how teams end up generating more leads to compensate for deals they already had and could not close.

How many stages does a sales pipeline have?

There is no correct number, and the sources that claim otherwise contradict each other. This is worth settling before you design anything, because the number is the part people fight about and the part that matters least.

Go to the primary sources rather than the guides. HubSpot's documentation states that its default Deals pipeline ships with seven stages, from Appointment scheduled through Closed won and Closed lost. Salesforce's documentation lists ten default opportunity stages, running from Prospecting through Negotiation/Review before the two closed states.

Now the part nobody mentions. Salesforce's own editorial guide on sales pipelines describes seven stages. Its product documentation configures ten. Same vendor, two answers, and neither is wrong. They are answering different questions: one is teaching a concept, the other is shipping a default.

That is the whole lesson. Published stage counts across the main vendor guides range from five to eight. The number is a design choice, not a standard.

The six stages most B2B teams actually use

Here is a workable default. What makes it useful is not the list, it is the third column: the condition that has to be met before a deal is allowed to move.

Why the count matters less than the exit criterion

Look again at what both CRM vendors attach to each stage: a probability. HubSpot assigns 90 percent to Contract sent. Salesforce assigns 75 percent to Proposal/Price Quote. Neither says what has to be true before a deal can leave that stage.

A probability is a forecasting input. It tells your CFO what to expect. It tells your rep nothing about what to do on Tuesday morning. That gap is why pipelines look tidy in a dashboard and behave badly in reality: every deal has a percentage, and no deal has a condition.

Write the condition. It should be observable by someone who is not on the deal, and it should be about the buyer, not about you. "Proposal sent" is an activity you controlled. "Buyer has responded to the proposal with questions or objections" is evidence that something moved.

How to build a sales pipeline

Building a pipeline is three decisions, in this order. Most teams make the third one implicitly and pay for it later.

Map the stages your buyers move through

Start from the buying process, not from your internal steps. Take your last ten won deals and your last ten losses, and write down what actually happened in sequence. You will usually find fewer stages than your CRM offers and one or two that nobody uses.

Cut anything that does not change what your team does next. A stage that produces no different behaviour is reporting, not process.

Write an exit criterion for each stage

One sentence per stage, phrased as something the buyer has done. Circulate it. If two reps read the same criterion and place the same deal in different stages, the criterion is not written well enough yet.

This is the single highest-return hour you will spend on your pipeline, and it is the step the guides skip.

Decide what has to exist before a deal advances

Every stage transition in a B2B deal is carried by something you send: a recap, a scoping note, a business case, a proposal, a formal response, a slide deck for an internal champion. Decide now what that artefact is at each stage and who produces it.

Teams that sell into structured processes feel this first. As one Cobl customer at Open describes it, the framework is always the same and the same documents are always required, while only the content changes. If the same documents are required every time, they belong in your stage definition, not in a rep's memory.

Where deals actually stall

Ask a sales leader where deals get stuck and most will say qualification. Look at the data and it is usually later, in the gap between sending an offer and hearing back.

The proposal stage is the blind spot of most pipeline guides

Most pipeline guides are published by CRM vendors, and a CRM records that the proposal stage exists. What happens inside that stage, the formal response, the deadline, the buying committee, the deck, the pre-sales work, is somebody else's job, so it rarely gets written about. Read the guides with that in mind and the vocabulary of a deal with constraints is close to absent.

The cost is real. At consulting firms answering formal requests, teams report five to ten days of work per document before automation. Eric Henon at CERAP Prevention described the pre-automation routine plainly: rereading everything each time, checking that nothing had changed, copying and pasting carefully. His word for it was "error-prone." Ten days of that sits inside one stage transition, and no pipeline dashboard shows it.

What buyers are doing while you wait

Silence after a proposal is the most common complaint in public sales communities, and the threads are consistent about the cause. The buyer is comparing you against other submissions. The buyer is circulating your document internally to people you have never met. The buyer is waiting for a budget decision that has nothing to do with you.

In all three cases, your document is being read by someone who was not in your meetings. That is the moment your pipeline stops being about your conversation and starts being about what you put in writing.

There is a second failure mode, and it is self-inflicted. When pipeline volume becomes the metric a team is judged on, the pipeline inflates. Deals that should have been marked lost stay open. Opportunities get created to satisfy a coverage target rather than because a buyer asked for something. Reps in public sales forums describe being told to build pipeline and then being asked, in the following review, whether that pipeline is real.

An inflated pipeline is worse than a small one, because every forecast, every hiring plan and every territory decision downstream is built on it. The fix is not more discipline in the review meeting. It is an exit criterion strict enough that a deal cannot enter a stage it does not belong in.

What actually moves a deal to the next stage

If the proposal stage is where deals die, the practical question is what unblocks it. Two things, and neither is following up more often.

Speed is not the point, fit is

Faster documents help, but only if the document is right. Damien Hontang, CEO of Cobl, framed it this way when the company raised six million euros in April 2026: AI can produce a document in minutes, and the real challenge is that the document is correct, "the right message, to the right client, at the right time".

Applied to a pipeline, that means a document written for the person who will actually read it. The champion who invited you needs something they can defend internally. The finance reviewer needs numbers and terms. The technical evaluator needs specifics. Sending all three the same PDF is how a deal with real intent still goes to no decision.

Speed then compounds fit rather than replacing it. At Open, producing a proposal from scratch took two to three hours; a framework version now takes five minutes, which leaves the time for adapting it to the client. That is the trade that changes a stage transition: less time assembling, more time aiming.

Documents are not the end of a stage, they are the hinge

It also does not stop at signature. Alberic Mulliez at Free Pro maps the document load across the whole relationship: pre-sales for commercial proposals and company presentations, subscription for contracts and service level agreements, then run mode for service quality reports and operational follow-up.

If you only staff the proposal stage, you have solved one hinge and left the others creaking. Renewals and expansions are pipeline too.

If the artefact at each stage is the bottleneck rather than the follow-up, that is a production problem, and it has a different fix from a process problem. See how a deal workspace handles it.

How to manage a pipeline you can trust

Managing is not inspecting. A weekly review that asks reps to justify each deal produces better forecasting hygiene and no additional revenue. A review that asks what each deal needs next produces movement.

The three numbers worth a weekly review

Track them per segment, not just in aggregate. A blended number will tell you conversion is fine while your largest deals quietly underperform. Our guide to measuring and improving sales win rate breaks down how one blended figure can hide three separate problems.

Velocity is worth calculating rather than estimating, because it makes the cost of a slow stage visible. Take a team with 60 open deals, an average deal value of 40,000 dollars, a win rate of 25 percent and a 90 day cycle. That is 60 times 40,000 times 0.25, divided by 90, or roughly 6,700 dollars of pipeline moving per day.

Now shorten the cycle by 10 days without touching anything else, no extra deals, no better win rate. The same calculation over 80 days gives roughly 7,500 dollars per day, about 12 percent more. Cycle length is the only variable in that formula you can change without hiring, and the largest single block of it usually sits between sending a document and getting a reaction.

The gains from fixing a single stage are larger than they look. At Appel Medical, drafting time went from twenty to thirty minutes down to two to five, an 86 percent reduction, with customer response time seven times faster. At Randstad, consultants moved from an average of twenty-five minutes per output to three. Neither figure is a win rate improvement on paper. Both change how many deals a team can carry through the same stage in the same week.

What to do with a deal that has not moved in 30 days

A stalled deal is not a lost deal, and treating the two the same way is how forecasts drift. The distinction is simple: a lost deal has a reason attached, a stalled deal has an unmet exit criterion. Pick one of three actions and record which one you chose.

• Re-qualify it. Test the exit criterion of the stage it is sitting in. If it was never met, the deal was in the wrong stage, not stuck.
• Give it the artefact it is missing. A one-page internal business case for your champion, a technical annex for the evaluator, a revised scope for finance.
• Mark it lost, with a reason. A parked deal inflates coverage and corrupts every forecast built on it.

One caution on automation. Generated documents still need a human to check the numbers, the claims and the client specifics before anything is sent. The tools shorten the assembly, they do not take responsibility for what goes out under your name.

Start with the stage that is costing you most

You do not need to redesign your pipeline to improve it. Take your stage-to-stage conversion, find the worst transition, and write a real exit criterion for it. If that transition turns out to be the proposal stage, as it is for most B2B teams selling into structured buying processes, the fix is not more follow-up. It is a better document, sent faster, aimed at the person who will actually read it.

Try Cobl's free trial and see what your next proposal looks like when the deal context is already in the room.

Stage configurations verified against HubSpot and Salesforce product documentation on 21 August 2026.