What should happen at each stage of our sales process?

The mistake that makes pipelines lie

Ask most boutique firms what their sales stages are and you will hear a list like: initial call, discovery, proposal, negotiation, closed. Then ask what moves a deal from discovery to proposal, and the answer is usually that discovery is finished.

That is a seller-side definition, and it is why forecasts are unreliable. Every one of those stages can be completed by your own effort. You can hold the call, run the discovery, write the proposal and follow up, and be no closer to a sale than when you started, while the pipeline reports steady progress.

Collective 54 puts the underlying problem plainly. CRMs record sales, they do not manage it. Dashboards report outcomes, they do not enforce anything. Pipelines reflect what sellers chose to update, often late and often optimistically. If the stage definitions themselves are optimistic, the tooling faithfully reports an optimistic picture.

Define stages by buyer evidence and the problem largely disappears. A buyer who has introduced you to the budget holder has done something. A buyer who has agreed a success measure has done something. Those things are observable and cannot be manufactured by your own diligence.

Before the stages: what the process has to reflect

A sales process is not a generic artifact. It has to reflect the way your clients actually buy, which means a buyer journey map, an outline of how a prospective client buys your type of service, comes before the stage design.

Two things about professional services shape that map.

The first is that services are bought and experienced rather than sold and consumed. The service and the person delivering it cannot be separated. This is why a demo-and-close structure imported from software does not transfer: there is nothing to demonstrate, and what the buyer is evaluating is partly you.

The second is that the type of engagement you sell determines the shape of the process. Elephant hunting, a small number of clients each spending a lot, produces long cycles with multiple stakeholders and a formal evaluation. Rabbit hunting, many clients each spending a little, produces short cycles with one or two people. Firms that try to run both through one process end up with stages that fit neither.

The stages, and the one test each

What follows is a structure rather than a script. The stage names matter less than the exit tests, and each test is a piece of evidence the buyer has to supply.

Qualified. The exit test is that you know the problem, the person who owns it, and whether money exists. Not whether they were interested. Interest is free. The buyer has named a problem in their own words and told you who besides themselves cares about it.

Problem agreed. The exit test is that the prospect has stated the cost of the status quo. This is the stage boutique firms skip most often, and skipping it is why so many deals die in silence rather than in a no. If the buyer has not quantified what the problem costs them, they have no reason to prioritize solving it this quarter, and your proposal will lose to doing nothing rather than to a competitor.

Approach agreed. The exit test is that the prospect has agreed how success will be measured. Not that you presented an approach: that they have accepted a measure. This is also the stage where you learn what they will actually buy, as distinct from what they said they wanted.

Access confirmed. The exit test is that you have met everyone who can say no. Deals in professional services are rarely lost by the person you have been talking to. They are lost by someone you never met, usually late, usually on grounds you were never given a chance to answer.

Proposal issued. The exit test is that the proposal contains nothing the buyer is seeing for the first time. A proposal should be a written record of an agreement already reached in conversation. If it is the first presentation of scope, price or measurement, you have converted a conversation into a document and given the buyer something to react to alone.

Commercial terms. The exit test is a signature path with a date and a named signer.

Closed. The exit test is a countersigned agreement. Not a verbal yes.

Seven stages is a design choice, not a rule. What matters is that no stage advances on seller activity alone.

Where these engagements actually come from

One structural point worth stating, because it changes how the early stages should read.

Referrals are the highest-quality revenue motion in professional services, producing shorter cycles, higher close rates, better-fit clients and stronger margins. A referred prospect arrives with borrowed trust and is in a genuinely different state from a cold lead: they ask about fit rather than competence, objections surface earlier and more honestly, and context is shared more freely.

Running that prospect through a full qualification sequence is a category error. It over-qualifies someone who is already qualified, introduces process where trust already exists, and the person who made the introduction notices that their credibility was handled carelessly.

So the process needs two entry paths. A referred prospect should be able to enter at problem agreed, because the qualification the referrer performed is real. A cold prospect starts at the beginning. Same exit tests, different starting point.

What has to exist around the stages

Stages alone are not a sales process. The surrounding elements matter as much: a prospecting process giving business developers a consistent way to find opportunities, a sales methodology for converting leads into clients, a coverage model allocating headcount so the target market is actually covered, an incentive system, and training.

Two of those deserve emphasis in a boutique firm.

Coverage, because in a partner-led firm coverage is not a spreadsheet exercise, it is a real constraint. A hardworking partner has roughly 2,500 productive hours a year, and as the firm scales only about half of that is available for business development. Once each partner is tapped out, sales flatline. Stage design does not fix a coverage problem, and no amount of process discipline creates hours that do not exist.

Training, because the stage exit tests above require conversations most delivery experts have never been taught to run. Asking a prospect to quantify the cost of their status quo is a learned skill, not an instinct.

Managing the process, as opposed to documenting it

A documented process that nobody enforces is a document.

Sales management is a distinct discipline covering call management, opportunity management, account management, territory management, retention and enablement, and it is a full-time job that in most boutique firms has been performed part time by a founder supervising part-time seller-doers. That structural mismatch, rather than any misunderstanding of what good looks like, is why boutique sales processes stay undermanaged.

What changes in Era 3 is capacity rather than understanding. Continuous monitoring of activity and outcomes, enforcement of process discipline, detection of patterns across calls and opportunities, and identification of breakdowns while correction is still possible can now run without human stamina. Interpretation, judgment and intervention stay human.

Applied to stages, that means the useful question stops being whether the deal moved and becomes why deals of this type consistently stall at this stage. That is a pattern question, and patterns are exactly what no part-time sales manager has ever had the bandwidth to see.

When this answer flips

If your firm sells a small number of very large engagements, seven stages is too many and the granularity is false precision. Three or four stages with hard buyer-evidence tests will serve better than a detailed pipeline built on a handful of deals a year.

If you are still genuinely partner-led and happy to be, formalizing stages can add administrative weight without adding revenue. The stage discipline pays off when people other than the owners are selling. Before that, the constraint is partner hours, and process does not create hours.

And if almost all your revenue arrives by referral, build the referral motion before you build the stage structure. A well-designed process applied to insufficient volume is an expensive way to be precise about a problem you do not have.

The short answer

Define every stage by what the buyer has done rather than by what you have done, because a seller can complete every activity in a stage without the buyer moving at all, and that is precisely why pipelines look healthy until they do not. Give each stage one exit test that requires evidence only the prospect can supply: a problem named in their words and the person who owns it, the cost of the status quo stated by them, an agreed measure of success, a meeting with everyone who can say no, a proposal containing nothing they are seeing for the first time, a signature path with a date and a named signer, and a countersigned agreement. Build the sequence from a buyer journey map rather than a generic template, and match it to your engagement type, since elephant hunting and rabbit hunting need different shapes and one process cannot serve both. Give referred prospects a shorter path, entering at problem agreed, because the qualification the referrer performed is real and re-running it costs you the referrer. Then surround the stages with the coverage model, methodology, incentives and training, and remember that a documented process nobody enforces is a document.