Pipeline stages that mean something

A pipeline stage should describe a verifiable event, not a feeling. "Proposal sent" is a stage because you can check it happened; "Interested" is a mood, and moods make forecasts fiction. Five to seven stages, each with a defined entry event and exit criterion, is enough for almost any 5–50-staff B2B service firm.

Why do most pipelines lie?

Because their stages are adjectives. A pipeline reading Aware → Interested → Engaged → Hot invites every deal to be placed by optimism, and optimism is not evenly distributed across a sales team. The same conversation sits at "Interested" for one rep and "Hot" for another, which means stage-conversion numbers compare personalities, not process.

The damage compounds upstream. Every report an MD reads — pipeline value, weighted forecast, conversion by stage, the whole instrument panel described in The MD Dashboard Blueprint — is computed from stage data. When stages are moods, the dashboard averages moods. Commonly this surfaces as a forecast that has been "about 90 days from a great quarter" for three consecutive years.

What makes a stage real?

Three properties:

1. An entry event. Something observable happened: a discovery call took place, a proposal was sent, verbal agreement was given. If two people could disagree about whether a deal belongs in a stage, it is not a stage.
2. An exit criterion. A defined event moves the deal forward, and a defined condition moves it out. Deals leave the pipeline by winning, losing, or timing out — not by being quietly ignored.
3. A meaningful probability. Over time, each stage acquires an honest conversion rate from your own history. That only happens when entry is consistent, which is property one again.

A workable default for a service firm: Enquiry → Qualified (discovery call held) → Proposal sent → Verbal agreement → Won/Lost. Add a stage only when it earns its keep in reporting.

How do you design the stages?

The mechanism is to work backwards from events that already exist in your sales motion:

1. List the observable milestones every won deal passed through in the last year — first conversation, scoping call, proposal, negotiation, signature. Pull ten real deals and check the list against them.
2. Keep only milestones you can verify from a system record: a calendar entry, a sent document, a signed agreement. When a milestone leaves no trace, then either instrument it or drop it.
3. Write the entry rule as a sentence. "A deal enters Proposal Sent when the proposal email leaves the building." When the event happens, then the stage changes — ideally automatically, and never before.
4. Add a time-out per stage. When a deal exceeds its stage's normal age — say, three weeks in Proposal Sent against a typical ten days — then it flags for action or gets closed as lost. Stale deals are the main source of inflated pipeline value.
5. Publish the rules on one page, and enforce them in the Monday review by evidence, not memory.

When entry rules are events and time-outs are enforced, then the pipeline becomes self-auditing: a deal cannot loiter, and a rep cannot promote it by enthusiasm alone.

What do defined stages fix downstream?

Almost everything the CRM is for. Conversion rates by stage become comparable across reps and quarters, so you can see where deals actually die — typically between proposal and decision, which is a follow-up problem, not a copy problem. Weighted forecasts become believable because the weights come from your own history rather than the CRM vendor's defaults. And adoption improves, because stages tied to events remove the most demoralising kind of CRM admin: the judgement call. The wider adoption question — making the whole system cheap to feed — is covered in nobody updates the CRM.

Stage design also sets up enforcement. Entry rules and time-outs are exactly the sort of checks that should run without a human — the data-hygiene robots that flag stale deals, missing values and impossible stage jumps automatically, so the rules keep working after the launch enthusiasm fades.

Where does speed fit in?

At the front gate. The first stage transition — enquiry to first contact — is the one with the steepest decay: as an industry rule of thumb, contact rates drop roughly 8x after five minutes, which is why the 5-minute rule deserves its own instrumentation. Put "first response time" on the stage itself and report it. A pipeline that measures its own front door tends to keep it open.

Stages are boring infrastructure, deliberately. But they are the difference between a CRM that records what happened and one that records what people wished had happened — and every number you steer the firm by inherits that difference.

---

Next step: the Growth System Audit — £450, seven days, credited against any build — maps where your growth system leaks, including whether your pipeline stages can support an honest forecast, and what to build first.