How to design a sales process with clear stages

What a sales process actually is

A sales process is the repeatable set of stages a deal moves through, from first contact to closed. It is the answer to a simple question asked about every deal in your pipeline: where is this, and what has to happen next? When every deal has an honest answer, selling becomes something you can measure, coach, and forecast. When it does not, your pipeline is a collection of individual gut feelings wearing a spreadsheet.

The word "repeatable" carries the weight. A process is not a description of one great deal your best rep closed. It is the pattern underneath many deals, written down so that a new rep can follow it and a manager can see where it breaks.

Process versus methodology

Keep two terms apart: the sales process is the stages a deal moves through — where a deal is — while the sales methodology is how you sell within those stages, frameworks like MEDDIC or SPIN — what you do to move it. You need both, and confusing them produces stages named after seller tasks, the most common way a process goes wrong. MEDDIC covers the methodology side in full; this page is about the stages.

The rule that keeps forecasts honest

Here is the single most important design rule: a stage is defined by what the buyer has done, not by what the seller hopes.

A rep can always take an action. They can send a proposal, schedule a follow-up, log a call. If your stages advance on seller actions, every deal marches forward regardless of whether the buyer is actually engaged — and your forecast becomes a wish list. Only the buyer can truly move a deal, and only buyer actions belong in your exit criteria.

Compare the two framings:

Seller-defined (inflates pipeline)	Buyer-defined (keeps it honest)
Sent the proposal	Buyer confirmed budget and timeline
Gave the demo	Buyer looped in the economic decision-maker
Followed up	Buyer agreed to a mutual evaluation plan

The right-hand column is harder to check off, which is exactly why it works. A deal that cannot meet a buyer-defined exit criterion is not ready to advance — and knowing that early is the whole value of having a process.

A worked five-stage pipeline

To make this concrete, here is a five-stage pipeline with a single buyer-defined exit criterion on each stage. Read down the right-hand column: every exit is something the buyer does or agrees to, never something the rep produces.

Stage	The buyer must — before the deal advances
Prospect	reply and agree to a first meeting
Discovery	confirm the problem is real and worth solving, and name who else weighs in on the decision
Evaluation	agree to a mutual evaluation plan and bring the person who controls the budget into a conversation
Proposal	confirm budget and timeline, and agree the proposal meets the criteria they set
Closed	sign

The pattern is the point. A deal does not leave Discovery because you sent a recap; it leaves the moment the buyer names the other people who have to say yes. A deal sitting in Evaluation for three weeks is not a rep who has gone quiet — it is a buyer who has not yet agreed to an evaluation plan or introduced the economic buyer, and the stuck stage tells you exactly which. Your stage names will differ from these; the discipline of a buyer-defined exit on each one should not.

Every stage has two gates, not one

Exit criteria get the attention. Entry criteria are the quieter half, and they carry just as much weight. Entry criteria answer a different question: what has to be true for a deal to belong in this stage at all?

The answer is almost always the previous stage's exit, read forward. A deal enters Discovery because it cleared Prospect — the buyer replied and agreed to a first meeting. State that same fact as an entry condition and it reads: a deal in Discovery has an agreed first meeting behind it.

Why name both when they describe the same fact? Because entry criteria are how you catch a deal that was pushed forward without earning it. A deal sitting in Evaluation that never had a real discovery call did not pass Discovery's exit — someone skipped a step. Without entry criteria, that skip is invisible; the deal simply appears in a later stage and everyone treats it as further along than it is. A stage with an exit criterion but no entry criterion lets deals teleport forward. Naming both gates closes the gap, and it is the cheapest way to keep a pipeline from filling with deals that look qualified and are not.

The canonical stages — and why yours will differ

Most B2B pipelines are a variation on one arc. It is worth knowing the common shape, not to copy it, but to have a vocabulary before you build your own.

Stage	What it establishes	The buyer action that ends it
Prospect	There is a person worth a conversation	Replies and agrees to a first meeting
Qualify	The deal is worth the pipeline's time	Confirms the pain is real and they can act on it
Discovery	You understand the problem and who decides	Names the other people who weigh in
Demo / Evaluation	The product fits the problem	Agrees to a mutual evaluation plan and brings in the budget owner
Proposal	The terms match what they asked for	Confirms budget and timeline against the proposal
Negotiate	The last objections are resolved	Agrees to terms, redlines, and procurement steps
Close	The deal is done	Signs

Do not adopt this list as written. It is a starting vocabulary, not a template. Some businesses fold Qualify into Prospect because their first conversation already sorts fit; others split Negotiate out from Proposal because their deals run through a long procurement gauntlet that deserves its own stage. The names matter less than the arc underneath them: each stage sits one notch of commitment above the last, and it ends when the buyer does something only a more-committed buyer would do.

How many stages

Use enough stages to reflect real decision points, and no more. Most B2B processes fit comfortably in four to seven stages. The failure modes sit at both ends:

• Too few stages and the pipeline hides where deals stall. Everything is either "in progress" or "closed," which tells you nothing.
• Too many stages and reps start managing the stages instead of the deals, nudging things forward to look busy. The process becomes theater.

Each stage should mark a genuine change in the buyer's commitment — a moment where the deal became meaningfully more or less likely. If two adjacent stages describe the same level of buyer commitment, they are one stage.

Qualify before you advance

Qualification is the filter that decides whether a deal belongs in the pipeline at all. A prospect can take a meeting and still be a bad deal. Interest is not fit, and qualification is how you tell them apart before you spend a rep's month on the difference.

Two frames do the job. BANT — Budget, Authority, Need, Timeline — asks four plain questions: can they pay, can this person say yes, is the pain real, and is there a reason to act now? Each maps cleanly to a buyer action you can check, which is why it survives. The lighter frame is pain plus fit: does the buyer have the expensive problem you solve, and does your product actually fit their situation? Use whichever your deals reward; the point is to have an explicit bar, not a feeling.

Qualification is not a stage you clear once and forget. It is a bar the deal has to keep meeting as it moves. A deal that qualified on Need but never surfaced Budget is half-qualified, and half-qualified deals are where forecasts go to die — they look advanced and collapse at the end. Re-check the bar at each stage, and treat a criterion that never gets confirmed as a warning, not a detail to chase later.

Tie qualification back to your ideal customer profile. A deal that fails your ICP filter is disqualified no matter how eager the buyer sounds. The ICP decides who is worth qualifying; qualification decides who is worth advancing.

Where the pipeline begins: the handoff into stage one

Every pipeline has a first stage, and something has to feed it. In most B2B teams that something is a sales development rep — the person who turns raw interest into a qualified opportunity an account executive can work. The handoff from SDR to AE is where alignment between marketing and sales either holds or breaks.

Myk Pono's argument is that the break starts with one word: lead. A lead is vague. Nobody agrees what one is, so marketing counts things sales never touches, and sales ignores things marketing counts. The fix is to hand off something concrete instead. A demo — or a qualified meeting — has a definition you can check: the prospect has the authority to decide and a real need behind the conversation. Draw the line there and the two teams finally mean the same thing when they call a deal real.

This is the case for SDRs reporting into marketing rather than sales. When marketing owns acquisition all the way to the qualified demo:

• its spend and ROI are measured against a concrete outcome, not a vague lead count;
• it gets to test the messages and target profile that later feed content and the website, because it stays close to the buyer through the handoff; and
• the incentive to follow up on every lead in time stays with the team that generated it, instead of being dropped by a sales org that has already hit its number and has little reason to squeeze more demos out of the same leads.

Wherever the SDR reports, the design rule does not change: the handoff into stage one must be a buyer action with a concrete definition, exactly like every exit criterion downstream. If your first stage begins with a "lead," you have imported at the top of the pipeline the vagueness you spend the rest of it removing.

Build it from your own deals

The reliable way to design a process is to reverse-engineer it from deals you have already won. Your closed-won history is the evidence; the process is that evidence written down.

Pull your last several won deals and trace the actual path each took. Look for the decision points every winning deal passed through — the moments the buyer's commitment visibly changed. Those recurring moments are your stages. Then name them in the buyer's terms, put them in order, and write the exit criteria for each as a buyer action.

The reason to build from real deals rather than a template is that a borrowed process describes someone else's buyer. Your stages should match how your customers actually decide, and the only record of that is your own history.

Test the stages against live deals

Once you have named the stages and written the exit criteria, test them against a handful of open deals. Try to place each deal in a stage. When a deal does not fit cleanly, the instinct is to force the deal into a box — resist it. A deal that does not fit is telling you the stage definition is wrong, not that the deal is unusual.

Iterate the definitions until every open deal lands somewhere honest. A process that only fits the deals you invented it from is not a process; it is a story.

A stuck deal is a buyer in an earlier stage than your CRM says

Once your stages track buyer actions, a stuck deal stops being a mystery. A deal is stuck when the CRM says it is in one stage but the buyer is behaving like they are in an earlier one.

The tell is always the same: the exit criterion for the current stage was never actually met. A deal marked Proposal where the buyer never confirmed budget is not a Proposal deal that went quiet — it is a Discovery deal wearing a Proposal label. Somewhere upstream a rep advanced it on a seller action (they sent the proposal) instead of a buyer action (the buyer confirmed budget and timeline), and the deal has been carrying a stage it never earned ever since.

This is where the buyer-defined rule pays off a second time. It disciplines the pipeline when you set the stages, and it diagnoses the pipeline every time a deal stalls. To unstick a deal, find the earliest stage whose exit the buyer never truly cleared, and work that stage — not the one the CRM shows. The stuck stage names the missing commitment, which means it also names the next conversation to have.

A deal that has genuinely met its current stage's exit and still isn't moving is a different problem — usually a shifted priority or a competing initiative — and it deserves its own conversation rather than more nudging. But most stuck deals are mislabeled, not stalled, and a buyer-defined process is what lets you tell the two apart on sight.

Move each stage like a consultant, not a salesperson

Your stages tell you which buyer action a deal needs next. Whether the buyer takes it depends on who they think they are talking to. The same request — a meeting, an intro to the budget owner, a signature — advances the deal from a consultant and stalls it from a salesperson.

The only difference between the two is approach. A buyer trusts a consultant and guards against a salesperson because of perceived intent: the consultant seems to care about them, the salesperson seems to care about the quota. You do not have to change your product or your price to move from one to the other. You change the words you use and the behaviors you show.

Most objections do not arrive at the price. They arrive in the first thirty seconds, because that is when the buyer decides which of the two they are dealing with. A handful of words trip the salesperson pattern on contact:

Drop this word	Why it pattern-matches to "salesperson"	Say instead
Available	Nobody uses it outside a sales script	"Is Sarah there?" / "Can I reach…"
Interested	Clinical and unnatural in a real conversation	"Would you be open to…"
Follow-up	Signals one more attempt to sell them	Anchor the next call to a new feature or a price change they asked to hear about
Demo	Tells them they are walking into a pitch	"Would you be open to seeing how two companies you respect got the result you're after?"
Close	Pure sales language	"Move forward together" / "become partners"

The demo reframe is the one that books meetings. An executive agrees to a meeting for a reason — to see how peers solved the problem, to understand pricing, to keep pace with a competitor. Name that reason and the ask stops sounding like a pitch.

Behaviors leak intent as loudly as words. Four to manage on every call:

• Slow down when they say they are busy. Speeding up signals nervousness; a deliberate pace signals confidence and respect. Pace is a tool — use it on purpose.
• Mirror their exact words back; do not paraphrase. Repeating a buyer's own phrasing proves you heard them and builds rapport faster than any pitch.
• Ask a sharp question, then stop talking. Silence is the buyer's processing time. The better the question, the longer the pause — sit in it. Across the whole call the buyer should be talking most of the time; if you are, you are pitching, not consulting.
• When you are tempted to smooth over a gap with a small lie, dig into the need instead. The urge to stretch the truth is the signal to ask more about the why. There is almost always an honest way your product meets the real need, and one discovered lie poisons everything else you have said.

None of this changes what stage a deal is in. It changes whether the buyer takes the action that lets the deal leave it.

Handle an objection to advance the deal, not to win the argument

An objection is a buyer declining, for now, to take the action a stage's exit criterion needs. Treat it as a fight to win and you harden the refusal: a rebuttal tells the buyer they are wrong, and being told they are wrong makes them dig in. The consultative move is to disarm the objection before you try to overcome it. Four steps, in order:

1. Empathy. Acknowledge the objection without arguing it. You are not agreeing it is right; you are refusing to fight it. This breaks the pattern the buyer expects — objection, then rebuttal — and lowers their guard.
2. Show you know them. Reference something real about their world: their role, a recent raise, a shift in their market. People do not take advice from strangers. Proving you understand their situation earns you the standing to question their thinking. This is the step most reps skip, and it is the one that matters most.
3. Confirm the objection. Say the reason back and check that you have it right before you solve anything. This makes the buyer feel heard, and it stops you from solving the wrong problem.
4. Propose a solution built for them. Only now do you problem-solve, and you can be creative — a phased start, trial access now with billing later, a smaller first step. Jump straight here, skipping the first three steps, and it generates friction: the buyer is not ready to be problem-solved until they feel understood.

The order is the point. Lead with the solution and you are arguing; lead with empathy and knowledge and the buyer lets you help. A handled objection ends the same way a cleared stage does — with the buyer doing the specific thing the exit criterion names: agreeing to the meeting, looping in the person who controls the budget, agreeing to terms. The objection was never noise. Like a stuck stage, it named the commitment the buyer was not yet ready to make, which means it also named the conversation that moves the deal.

Forecasting: the honest payoff of buyer-defined stages

The reason to do all of this arrives at forecast time. A forecast is a claim about which deals will close and when, and it is only ever as trustworthy as the stages it rests on.

The mechanism is straightforward. Look at your closed history and measure, for each stage, how often a deal that reached it went on to win. That win rate is the stage's probability. Multiply each open deal's value by the probability of the stage it is genuinely in, sum the results, and you have a weighted forecast built on evidence instead of optimism.

The whole thing stands or falls on the stages being honest. If deals advance on seller actions, every stage's probability is inflated, because the pipeline is full of deals sitting in stages they never earned. The forecast reads high, then misses — and no one can say why, because the numbers looked defensible the whole way. Buyer-defined exit criteria are what make the stage probabilities mean something: a deal in Proposal is only worth Proposal's win rate if it actually cleared Proposal's exit.

Two disciplines keep the number honest. Derive the probabilities from your own closed deals, not from a borrowed template, because a template describes someone else's win rates. And never let a rep set a deal's probability by feel — the stage sets the number, and the rep's only job is to place the deal in the stage its exit criteria actually support. Take probability out of the rep's hands and you remove the single largest source of forecast fiction.

What you end up with

The deliverable is the sales process and its stages, written as one reference: the named stages in order, the exit criteria for each stated as a buyer action, and a short description of what a deal looks like when it sits in each stage.

That document does three things at once. It gives reps a shared path to follow, gives managers a consistent way to inspect deals, and gives your forecast a spine — because a deal in a given stage now means the same thing no matter who owns it. Everything downstream, from coaching to capacity planning, depends on that shared definition holding.