A sales pipeline is only as useful as the agreement behind its stages. If “proposal” means a sent document to one person and a verbal chat to another, the pipeline report is a mix of opinions. This guide gives a practical stage model for agencies and B2B service firms, the exit criteria that make it consistent, and how to weight it without fooling yourself.
What are typical sales pipeline stages for a service firm?
A service firm typically needs five open stages between a new enquiry and a signed contract. Names vary, but the steps are similar because service sales usually involve scoping and a written proposal:
- New lead: an enquiry or referral that nobody has assessed yet.
- Qualified: someone has checked it is a real need, a fit for what you do, and worth a conversation.
- Discovery: you are scoping the work with the client and understand the problem, the budget range and who decides.
- Proposal: a written proposal or estimate has been sent to the decision-maker.
- Negotiation: the client has responded and you are agreeing terms, scope or start date.
Then two closed stages: won and lost. Keep “on hold” as a flag rather than a stage, so paused deals do not quietly inflate the pipeline.
What exit criteria should each stage have?
Exit criteria are facts that must be true before a deal moves forward. They are what make two people’s pipelines comparable. A starting set:
| Stage | Exit criteria (all must be true to move on) |
|---|---|
| New lead | Contact details confirmed; source recorded; someone has read the enquiry |
| Qualified | Need is real and in your service range; rough budget or scale known; timing within the next two quarters; a named owner on your side |
| Discovery | Problem and outcome written down; budget range discussed; decision-maker and process identified; a call held with them or their delegate |
| Proposal | Proposal sent to the decision-maker; a date agreed to discuss it |
| Negotiation | Client has responded in substance; open points listed; target signing date agreed |
| Won | Contract or purchase order signed; first invoice date set |
For the qualification step in more depth, see how to qualify a lead.
What is a weighted pipeline?
A weighted pipeline is the value of each open deal multiplied by the probability that deals at its stage are won, then added up. It turns a list of hopes into a single, probability-adjusted number.
The probabilities should come from your own win rates. If three in ten deals that reach proposal eventually sign, the proposal-stage probability is 30%, whatever a template says.
How do you calculate a weighted pipeline?
Multiply each deal’s value by its stage probability, then sum. A worked example with invented figures. Take a 30-person agency with 15 open deals, using illustrative stage probabilities:
| Stage | Deals | Total value | Probability (illustrative) | Weighted value |
|---|---|---|---|---|
| Qualified | 6 | £240,000 | 10% | £24,000 |
| Discovery | 4 | £160,000 | 25% | £40,000 |
| Proposal | 3 | £135,000 | 50% | £67,500 |
| Negotiation | 2 | £90,000 | 75% | £67,500 |
| Total | 15 | £625,000 | £199,000 |
The headline pipeline is £625,000. The weighted pipeline is £199,000. Neither is a forecast of what will sign next month, because the weighted figure ignores timing. But £199,000 is a far better basis for hiring conversations than £625,000.
Two checks worth making every time:
- Concentration. If one £90,000 deal carries a third of the weighted figure, say so. A weighted number over a handful of deals is fragile.
- Age. Deals that have sat in the same stage for twice the usual time rarely close at the stage probability. Flag them or discount them.
What goes wrong with weighted pipelines?
Most problems come from treating the weighted figure as more precise than it is. The common ones:
- Made-up probabilities. Stage percentages copied from a template, never checked against your own results. Revisit them each quarter.
- Deals skipping criteria. A deal marked “proposal” because a price was mentioned on a call. Exit criteria stop this, but only if someone checks.
- Stale deals. Nothing has happened for two months, but the deal still counts. Set an age limit per stage and review anything beyond it.
- Timing ignored. A deal likely to sign in nine months counts the same as one signing next week. Add an expected close date and report by quarter.
- Weighted pipeline used as cash. It is neither signed work nor money. Keep it apart from planned billables and invoices (see planned, invoiced, collected).
- Personal optimism. Some people move deals forward early, others late. Shared exit criteria and a weekly review even this out.
How often should you review the pipeline?
Review it weekly as a team and monthly as a trend. The weekly review is about individual deals: what moved, what is stuck, what the next step is and who owns it. The monthly review looks at the weighted total over time, win rates by stage and loss reasons.
Keep the weekly meeting short. Go through deals that changed stage, deals past their age limit, and deals with no next step. Everything else can wait.
Once a quarter, compare your stage probabilities with what actually happened. Count how many deals that reached each stage in the past year were won, and update the percentages. If you have fewer than a few dozen closed deals, the new figures will be rough, so move them gradually rather than swinging to whatever last quarter showed.
Where OpsUPLOOP fits, and where it doesn’t
OpsUPLOOP keeps pipeline, planned billables and invoices as separate views, so a weighted pipeline never gets mistaken for signed work or cash. It supports probability-weighted pipeline views, a probabilistic (Monte Carlo) revenue forecast shown alongside the billing already planned on won work, scenario adjustments and dated snapshots so you can compare this month with last, and a weighted deal qualification questionnaire, in stages you define, that your team completes as a deal moves. AI researches and scores leads with visible reasons; a person decides what moves. See OpsUPLOOP Billing and OpsUPLOOP Sales.
It is not a machine-learning prediction, and it cannot make your probabilities accurate for you. The forecast is only as good as the stage definitions and win-rate history you put in, so the discipline of keeping stages honest still sits with your team, whether OpsUPLOOP is your CRM or sits on top of it.
Other questions
How many pipeline stages should an agency have?
Most service firms manage well with five or six open stages plus won and lost. More than that tends to create stages nobody can tell apart.
Where do stage probabilities come from?
Ideally from your own history of how many deals at each stage were eventually won. If you have too few deals for that, start with sensible guesses, label them as guesses and revisit them each quarter.
Should lost deals stay in the pipeline?
Mark them lost with a reason rather than deleting them. Loss reasons are the cheapest source of insight you have, and they keep your win-rate history honest.