I’ve seen a lot of CRMs. Too many, probably. And in most of them, the pipeline is a fiction.
Not deliberately. Nobody’s trying to mislead anybody. The numbers are just… wrong. Deals that closed six months ago still sitting as “negotiation”. Opportunities that were dead on arrival marked as 40% likely to close. Enterprise prospects who gave a polite no eight weeks ago still in the forecast because no one wants to be the one to move them to “lost”.
The pipeline number your CRM shows you is, in most of the businesses I work with, inflated by somewhere between 30% and 60%. Which means the forecast it produces is worthless. Which means the business decisions made from that forecast are, at best, based on hope.
Why this happens
The CRM lies because people lie to the CRM. Not maliciously, but optimistically. It’s genuinely hard to mark an opportunity as lost when you still think it might come back. It’s uncomfortable to have an empty pipeline, so you keep old deals in there to pad it out. Your rep doesn’t want to lose commission-eligible opportunities, so they leave them open.
The result is a database full of wishful thinking dressed up as data.
There’s also the problem of stage definitions. In most CRMs I inherit, the pipeline stages mean different things to different people. “Proposal sent” to one rep means a formal written proposal. To another it means they mentioned a number on a call. “Qualified” to one person means they’ve confirmed budget, authority, need, and timeline. To another it means they picked up the phone.
When stages mean different things, the pipeline data is incoherent. You can’t forecast from it because you can’t trust what it’s actually telling you.
The two things that fix it
First: stage definitions with exit criteria.
Every stage in the pipeline needs a specific, behavioural definition of what it means to be in that stage, and a specific list of things that have to be true to move out of it.
Not “discovery complete”. What does “discovery complete” mean? The rep had a call? The prospect shared their budget? They’ve confirmed a decision-making process? Pick one, write it down, and make it unambiguous.
Exit criteria only. Not entry criteria. The question is always: what has to have happened for this deal to move to the next stage? If the answer is “the rep feels good about it”, you don’t have exit criteria. You have vibes.
Second: a consistent review cadence with the right questions.
Pipeline reviews that work aren’t updates. They’re challenges. The right questions aren’t “how’s this deal going?” They’re:
- What has actually changed since last week?
- What does the prospect need to do next, and when?
- What are you doing if they don’t?
- Who else is involved in the decision and have you spoken to them?
- What would make you lose this deal?
If the rep can’t answer these questions, the deal doesn’t belong in the forecast. Not as a punishment, but as an accurate reflection of reality.
What a clean pipeline looks like
A clean pipeline is smaller than a dirty one. That’s almost always true, and it’s almost always uncomfortable. When you run proper qualification and honest stage criteria through a bloated pipeline, you typically lose 30–50% of the “opportunities” in it.
That’s not a disaster. It’s clarity. A pipeline of 20 genuine opportunities is worth more than a pipeline of 80 mixed-quality ones, because you can actually work the 20.
It also means your forecast is useful. If you can predict your revenue within 15–20% on a rolling 90-day basis, you have a sales function. If you’re regularly surprised by what closes and what doesn’t, you don’t.
The goal isn’t a full CRM. It’s a true one.
One more thing
If you’re thinking “this is a CRM configuration problem”, it’s not. The tool is fine. You could build what I’m describing in Salesforce, HubSpot, Pipedrive, or a spreadsheet.
The configuration matters, but it’s secondary. The real work is building the culture where people update things accurately and the reviews are honest. That’s a people and process problem. The CRM is just where it shows up.
For the specific mechanics of pipeline stage design and exit criteria, Pipeline Stage Definitions with Exit Criteria has the working template. And if you’re on HubSpot, HubSpot for B2B Founders covers the configuration decisions that actually matter for getting clean data out of it.