It is Monday morning. The pipeline report says you have fifty-one open deals. Your sales lead reads out the total value, which is comfortably above target, and everyone nods. Then you scroll and notice that eleven of those deals have not moved a single stage since the second week of last month, and nobody in the room can tell you what the next action is on any of them.
The report was not wrong. The sales pipeline stages were doing exactly what they were built to do, which is to hold a deal in place until someone moves it. The problem is that nothing in the system defines what “moving it” means.
The stage name is not the problem, the missing exit criterion is
Most teams argue about the wrong thing. They spend a fortnight debating whether the fourth stage should be called Discovery or Needs Analysis, and then never write down the one sentence that matters: what has to be true before a deal is allowed to leave this stage.
Without that sentence, a stage becomes storage. A salesperson drags a deal forward when they feel optimistic and leaves it where it is when they do not. Two people looking at the same deal put it in different stages. The pipeline total stops meaning anything, because half of it is hope and you cannot tell which half.
A nine-stage pipeline, which is what BizvoraOne uses, is long enough to be genuinely useful. It lets you separate a lead you have spoken to from a lead you have qualified, and a proposal you have sent from a proposal that is being negotiated. That granularity only pays off if each stage has a written exit criterion. Otherwise nine stages is just six more places for a deal to hide.
A workable shape for nine sales pipeline stages looks something like this. Your names will differ, and they should.
- New lead
- Contacted
- Qualified
- Discovery or site visit
- Proposal or quotation sent
- Negotiation
- Verbal yes, awaiting PO
- Won
- Lost
Nine boxes. Nine sentences to write. That is the whole exercise.
Writing an exit criterion that survives a Monday review
An exit criterion has to be checkable by someone who was not in the meeting. “Customer is interested” fails that test. “Customer has confirmed a budget range and a decision date” passes it, because you can ask to see where that is recorded.
Here is the pattern. Each criterion names a fact that has been established, and where it lives.
- New lead to Contacted: a two-way conversation has happened, not a missed call or a delivered WhatsApp message. Logged with date and channel.
- Contacted to Qualified: you know what they want, roughly what they can spend, who signs, and when they need it. All four, written in the deal.
- Qualified to Discovery: a specific requirement has been captured, with the technical or site detail you need to price it accurately.
- Discovery to Proposal sent: the document has actually gone out, and you know which named person received it.
- Proposal sent to Negotiation: the customer has responded with a question, an objection or a counter on price or scope. Silence is not negotiation.
- Negotiation to Verbal yes: commercial terms are agreed, including payment terms and GST treatment, and the customer has said so in writing or on a recorded call.
- Verbal yes to Won: the purchase order, signed quotation or advance has arrived.
- Anything to Lost: a reason code and a one-line note. No blank losses.
Notice that most of these hinge on something the customer did, not something your team did. That is deliberate. Deals that advance on internal optimism are the ones that sit in the pipeline for a quarter.
The three sales pipeline stages where deals actually die
Stalled deals are not spread evenly. In practice they pile up in three places, and they pile up for three different reasons.
Contacted, which is often a polite word for touched once
This is the biggest graveyard and the easiest to ignore. A lead comes in from one of your fifteen sources, someone rings once, nobody picks up, and the deal is marked Contacted. It now looks like work in progress on the dashboard. It is not. It is an enquiry nobody called back, wearing a costume.
The tell is a stage full of deals with a first-touch date and no second-touch date. The fix is partly cadence and partly speed, and speed matters more than most teams admit. We have written separately about why first response time is the cheapest thing to fix, so the short version here is this: define Contacted as a conversation, not an attempt, and make three failed attempts either return the lead to New or push it to Lost with a reason code.
An AI voice agent that rings new leads within minutes helps here, because it removes the excuse that nobody was free. It does not remove the need for a rule about what happens on attempt four.
Proposal sent, where waiting is mistaken for progress
The second graveyard is the stage everyone feels good about. The proposal went out. The work is done. The deal value sits in the forecast at a healthy probability. And then two weeks pass.
Proposal sent is dangerous precisely because it feels like an achievement. But a proposal with no follow-up date attached is a document, not a deal. The exit criterion above is deliberately strict: a deal only leaves this stage when the customer responds. If they have not responded in the time your business normally takes, the deal is stalled, whatever the forecast says.
This is also where slow drafting compounds the damage. A proposal that takes four days to leave your office arrives after the customer has already read two others. Whether you draft with AI or from a template library, the point is that the gap between Discovery and Proposal sent should be measured in hours, not working weeks.
Verbal yes, awaiting PO, the most expensive lie in the pipeline
The third graveyard is the smallest by count and the largest by value. The customer has said yes. Everyone believes it. The salesperson has told the operations team to start planning. And the purchase order has not arrived.
Deals sit here for one of three reasons. The person who said yes cannot actually sign. A finance or procurement step nobody mapped is still running. Or the yes was a courtesy and the real answer is no.
This stage is worth watching closely because it sits on the boundary between sales and money. A deal that clears it becomes an invoice, and an invoice that is raised without a clean order reference is the sort of thing that quietly ages past ninety days. That handover is a subject in itself, and it is covered in the piece on keeping your CRM and your books in agreement.
A verbal yes with no purchase order is not a closed deal. It is a deal with a deadline you have not written down.
How to spot the stall in a weekly review
You do not need a new report. You need three numbers per stage, read in the same order every week.
- Count in stage. How many deals are sitting here right now.
- Oldest deal in stage. Days since it last moved. Not days since creation.
- Deals with no next action date. This is the number that predicts next quarter.
Then apply a stage-level ageing limit. Set your own, based on how long your sales actually take, and write it next to the exit criterion. A trading business might allow four days in Proposal sent. A construction or real estate deal might reasonably allow six weeks. What matters is that the limit exists and that breaching it triggers something.
A useful weekly cadence for the sales dashboard:
- Read the three numbers for Contacted, Proposal sent and Awaiting PO first. Skip the rest unless something looks odd.
- For every deal past its stage ageing limit, the owner says one of three things: here is the next action and date, here is why the limit should change, or mark it Lost.
- No deal leaves the review without a named owner and a dated next action.
- Count the losses out loud. A review where nothing is lost is a review where nothing was decided.
Fifteen minutes, done properly, is enough for a team of five to fifteen sellers.
Where this does not apply
If your sale is one conversation at a counter, nine stages is overhead and you should not pretend otherwise. Track the enquiry and the outcome, and leave it there.
Ageing limits also do not travel between industries. A ninety-day silence in an enterprise IT deal may be normal procurement. The same silence on a coaching admission is a dead lead. Borrowing someone else’s thresholds is how you end up chasing the wrong deals politely and the right ones too late.
What to do first
Open your pipeline. Sort by days since last stage change, longest first. Look at the top twenty deals and note which stage each one is sitting in. You will almost certainly find them clustered in two or three stages, and those are your leaks, not the ones you assumed.
Then write nine sentences, one exit criterion per stage, and an ageing limit for each. Put them somewhere the sales team can see while they work, not in a policy document. Run one review against them. The first week will feel like an argument about definitions. That argument is the work.
