Most pipelines are a work of fiction, and they are fiction for one specific reason: the stages describe what the salesperson did rather than what the buyer did.
"Proposal sent" is a stage in thousands of companies. It measures an email leaving your outbox. It says nothing whatsoever about whether the deal will close, which is why forecasts built on it are consistently wrong in the same optimistic direction.
Fix that one thing and everything else about pipeline management gets easier.
The rule: buyer actions only
Every stage transition should be triggered by something the buyer did, observable, that you could point at.
| Seller-defined (weak) | Buyer-defined (strong) |
|---|---|
| Proposal sent | Proposal acknowledged, or questions asked |
| Demo scheduled | Demo attended |
| Contacted | Replied |
| Following up | Asked for a next step |
| Nurturing | Re-engaged after a gap |
The right-hand column is harder to fill in, which is exactly the point. A pipeline that is hard to advance is a pipeline that tells you the truth.
The five stages
This model covers almost every small-team sales process. Rename them to fit your language; do not add to them without a real reason.
1. Contacted
You have reached out. They have not responded.
Entry: first message sent. Exit: any reply, or the sequence ends. Watch: this stage should be large and should drain quickly. If deals sit here for months, they are not deals, they are a list.
2. Engaged
They replied. Something is happening, but you do not yet know whether it is real.
Entry: a reply that is not a rejection. Exit: you have confirmed need, authority and rough timing. Watch: the most commonly inflated stage. A polite "sounds interesting, send me details" is engagement, not qualification.
3. Qualified
There is a real problem, a real budget and a real person who can decide.
Entry: you can answer three questions: what problem, who decides, when. Exit: a proposal or quote is with them. Watch: if you cannot answer all three questions in one sentence each, the deal belongs in Engaged.
4. Proposed
They have your numbers and they are evaluating.
Entry: proposal delivered and acknowledged. Not sent. Acknowledged. Exit: a decision, either direction. Watch: the stage where deals go silent. Set a hard time limit, and use it.
5. Committed
They have said yes. Paperwork, scheduling or onboarding remains.
Entry: verbal or written agreement. Exit: signed, paid or started. Watch: keep it short. A deal that lives here for weeks has not actually been won.
Then closed-won and closed-lost, which are outcomes rather than stages.
The stall rule
Every stage has a natural duration in your business. Measure it for a month and you will have real numbers; until then, estimate.
A deal that has been in a stage for twice its normal duration is stalled, regardless of what the owner says about it. Stalled deals get one direct message asking for a decision either way, and then they get marked lost.
This feels brutal and it is the single healthiest habit in pipeline management. Deals kept alive out of optimism inflate the forecast, absorb attention that new prospects need, and delay the moment you find out you have a problem. A pipeline of twelve real deals is worth more than one of forty hopeful ones, and it is a lot less depressing to look at.
Lost reasons: the cheapest research you will ever do
Every closed-lost deal gets one field: why.
Keep the list short and fixed, because free text will never be analysed:
- Price
- Timing
- Chose a competitor
- Went in-house
- No decision made
- No budget
- Not a fit
Three months of this data answers questions you would otherwise pay a consultant for. Losing consistently on price means a positioning problem. Losing on timing means you are contacting people too early, which is a targeting fix. Losing to "no decision" more than anything else means your offer is not urgent enough, which is a messaging fix.
Where pipeline meets outreach
The stages above only work if deals arrive in them automatically. A pipeline you have to remember to update is a pipeline that is wrong by Wednesday.
Practically, that means:
- A reply moves a lead from Contacted to Engaged on its own, and stops the outreach sequence at the same time.
- Conversation history lives with the deal, so you can read the whole thing without opening a mailbox.
- The stage is visible in the same place you answer messages, not in a separate tool you open weekly.
This is the argument for having outreach and pipeline in one place rather than a sending tool feeding a separate CRM: not elegance, but the fact that every handoff between two systems is a place where updates stop happening.
The weekly review, in ten minutes
- Anything stalled? Twice normal duration. Chase or close.
- Anything in Proposed over the limit? Ask for a decision either way.
- Is Contacted being refilled? Empty top of funnel today is empty revenue in two months.
- What did we lose, and why? One line each.
Four questions. Ten minutes. It will do more for your forecast than any amount of stage engineering, because the problem was never the stage names. It was that nobody was willing to say a deal was dead.