B2B SaaS pipeline stages: definitions and exit criteria you can enforce

Updated 12 November 2026 · Written by Reyansh Juneja and Naman Dhiman, the co-founders of SOVE7GN, who run inbound, midbound and outbound for founders

A B2B SaaS startup needs six pipeline stages, not twelve: new, discovery held, qualified, evaluation, commercial, and closed. What makes a stage useful is not its name but its exit criterion, a fact that is either true or false, verified by something the buyer did rather than something you felt. A pipeline where stages move because a rep is optimistic gives you a forecast that is wrong in both directions.

What stages should my CRM have?

Start here and resist adding more. Each row is a stage, the fact that lets a deal leave it, and the thing founders mistake for that fact.

Two rules make this work. A deal can only sit in a stage if the previous criterion is still true, and a deal with no next meeting on the calendar does not count as open pipeline no matter which stage it is in. That second rule removes most of the fiction from a startup pipeline in one afternoon.

Why six stages and not twelve?

Longer pipelines get copied from enterprise sales teams where a stage exists so a manager can coach a specific skill. At 5 to 50 people, with the founder selling, extra stages create three problems: deals get parked in a middle stage nobody owns, the forecast gets harder to argue about rather than easier, and nobody updates the CRM because it takes too long.

If you want more detail, add fields, not stages. A field for "budget owner named" or "security review started" tells you the same thing without multiplying the board.

One stage you should not create is a pre-stage for leads you have not spoken to. Those belong in your outreach tool or a list, not in the pipeline. Pipeline starts when a human agrees to a meeting.

How do I set up a sales pipeline that stays honest?

Four rules, enforced on the same day each week:

1. Every open deal has a next meeting with a date, or it moves to closed lost with a reason. "Nurture" is not a stage, it is a list you write to later.
2. Stage moves need the buyer action, written in the deal notes in one line. No action, no move.
3. Close dates are set by the buyer's event, such as their contract renewal, their board meeting or their go-live, not by the end of your quarter.
4. Losses get a reason from the four that matter: no problem, no budget owner, lost to a named alternative, or timing. "Went dark" means you failed to find the budget owner, so file it under that.

Rule 4 is the one that improves your targeting. After 20 losses you will see whether your problem is the list, the message or the product, and the answer usually points back to how you defined your ICP.

A worked pipeline for a made-up startup

Example only. Northmoor is invented, and the deal counts below are inputs chosen to show how the stages read on a Friday, not benchmarks.

Northmoor sells compliance software, 14 people, founder-led sales, average deal $38,000 a year.

Reading it this way takes about 20 minutes. Northmoor's real problem is visible in row 2: two deals are sitting in discovery held without a problem statement, which means those calls did not do their job. That is a call-quality issue, fixed by running the discovery call to a structure, not a pipeline issue.

How much pipeline do I need to hit my number?

Work backwards from revenue to held calls, then forward through your own stage conversion. The full arithmetic, with a worked example at a stated deal size and a table at three close rates, is on how many sales calls a startup needs a month. The short version: measure your own conversion between each pair of stages for a quarter before you trust any published benchmark, because stage definitions differ so much between companies that cross-company conversion rates are mostly noise.

When your stages are defined and your deals still stall at discovery held, the problem is almost always upstream: the wrong people are booking calls. That is a list and message problem. SOVE7GN GTM works on that end, matching fitting companies to the right person, writing one personal message each and handling replies so the calls that land on your calendar are with budget owners. The founder approves posts and takes the calls. Book a call; pricing is shared on the call. The rest of the guides are in the GTM guides hub.

What tooling does this need?

Less than you think. A pipeline this size runs in any CRM, and the setup that matters is three custom fields (budget owner, next step date, loss reason) and a saved view sorted by next step date. The wider question of when a startup needs real operations support is covered in revops for startups. Do not buy a forecasting tool before you have 50 closed deals; there is nothing for it to learn from.

Questions people ask