Open most CRMs and you will find a pipeline with stages like Contacted, Interested, Proposal Sent and Negotiation. It looks organised. It forecasts terribly, and the reason is in the names.
Every one of those stages describes something the seller did, or something the seller believes. “Interested” is an opinion. “Negotiation” often means one email about price that was never answered. A pipeline built from opinions produces a forecast built from optimism.
Define stages by buyer evidence
The fix is to make every stage depend on something the buyer did, which you can point at. Not “they seem keen” but “they sent us the requirement document”.
Compare these two definitions of the same stage:
- Weak: Qualified – we have had a good conversation and they are interested.
- Strong: Qualified – we know the problem, roughly what solving it is worth, who signs, and the timescale they are working to.
The second cannot be fudged. Either you know who signs or you do not. That is what makes a forecast mean something.
A structure that works for most established businesses
Five stages is usually enough. Fewer and you cannot see where deals stall; more and nobody keeps it up to date.
1. New enquiry
Someone has come in and nobody has spoken to them yet. The only thing that matters here is how long they stay – and the answer should be minutes, not days. Response time is the highest-leverage number in most pipelines.
2. Qualified
You have spoken and established four things: the problem, its approximate value, who makes the decision, and when they want it solved. Missing any of those, it stays here or it is disqualified. A deal sitting in Qualified for three months with no decision-maker named is not a deal.
3. Scoping
The buyer is actively giving you something – access, documents, time with their team. This is the stage most businesses skip, and it is the most predictive. A buyer investing effort is a buyer who intends to buy.
4. Proposal with the decision-maker
Not “proposal sent”. Sent is an action you took. The stage is reached when the person who signs has seen it and you have discussed it. A proposal emailed into silence belongs back in Scoping.
5. Verbal agreement
They have said yes and you are working through paperwork. If a deal sits here longer than a fortnight, something unspoken is wrong – usually an approval you did not know about.
Make disqualifying respectable
The most useful habit a sales team can build is closing deals as lost, early and without shame. Bloated pipelines full of hopeful entries hide the real number and waste the follow-up effort that should go to live opportunities.
Give every stage a maximum age. When a deal exceeds it, someone decides: advance it, or close it with a reason. “No budget this year” and “went with a competitor” are different problems, and you cannot fix either if everything is recorded as “no response”.
The three numbers worth watching
Stage-to-stage conversion. Where do deals actually die? If most enquiries never reach Qualified, your marketing is attracting the wrong people. If Scoping to Proposal is where they go, your proposals are the problem.
Time in stage. Rising time in one stage is the earliest warning you get that something has changed.
Average deal size by source. Often the channel producing the most enquiries is not the one producing the most revenue. Worth knowing before you increase its budget.
Frequently asked questions
Should different services share one pipeline?
If the buying process genuinely differs – a two-week transactional sale versus a six-month consultative one – use separate pipelines. Forcing both through one set of stages makes both forecasts useless.
Who should update the CRM?
Whoever had the conversation, immediately afterwards. CRM updated in a weekly batch is a memory exercise, and memory flatters. Keep the required fields few enough that updating takes under a minute.
What probability should we attach to each stage?
Derive it from your own history rather than copying defaults. Take the last two years of deals, see what proportion closed from each stage, and use those figures. They will be lower than the defaults, and more useful.
How often should we review the pipeline?
Weekly, briefly, focused on movement and ageing rather than reading out every deal. Monthly is too slow to catch a stall while it can still be fixed.
Build a pipeline that tells you the truth
Most of this is a definition problem rather than a software problem. We help established businesses set up pipelines, CRM and reporting that reflect how their buyers actually decide — see our lead generation and sales service, or get in touch.
