Sales process steps. The stages that actually hold up once you check who's following them.

The short answer. Six stages, written down, is the easy part.

Ask most founders if they have a sales process and they'll say yes, and point to a pipeline with named stages in the CRM: prospecting, qualification, needs analysis, proposal, negotiation, closed. That's a real answer, and it's also an incomplete one. A named sequence of stages describes what a sales process should look like. It says nothing about whether a deal actually meets the standard required to sit in each stage, or whether two reps looking at the same deal would place it in the same place.

Supered's 2026 State of Sales Enablement report puts a number on exactly this gap: 89% of sales organisations report having a defined sales process, but only 36% of reps say they actually follow it as designed. That's not a small implementation lag, it's the majority of defined processes existing mostly on paper. The same report found that teams which consistently inspect deals against the defined process, checking a deal's actual state against its claimed stage in pipeline reviews, hit quota at 6.3 times the rate of teams that rarely or never do that inspection. The process itself isn't what drives the difference. The checking is.

How it works in practice. The six stages, and the exit test for each one.

The pattern worth noticing across every stage is that the exit criteria is never "the rep did the activity for this stage". It's "the buyer did something that proves the stage's purpose has actually been achieved". A rep can send a proposal, which is an activity, without the buyer having formally reviewed it, which is the actual exit condition. Measuring activity instead of the buyer's response is how a pipeline ends up full of deals sitting in a stage they haven't genuinely earned, which is exactly what inflates a forecast until deals start slipping and nobody can say why.

This activity-versus-response distinction is also the simplest way to spot a process that's been copied from a template rather than built around how a specific business's buyers actually behave. A template's exit criteria read generically because they have to work for any business; a real one names the specific document, confirmation or signal that particular buyer type produces at that particular point, which is also why a process genuinely worth having usually takes more than one meeting to write properly.

Qualification is where this distinction matters most, because it's the stage most often rushed under pipeline-generation pressure. A deal qualified on an assumed budget rather than a confirmed one looks identical in the CRM to a properly qualified deal, right up until proposal stage, when the gap between assumed and real budget surfaces as a stalled negotiation that looks like a pricing objection but is actually a qualification failure from several stages earlier.

The stages themselves should map directly onto the CRM pipeline, not sit as a separate document that drifts away from how deals are actually tracked. If the sales process lives in a slide deck or a wiki page and the CRM pipeline uses a different set of stage names, reps default to whichever one is faster to update, almost always the CRM, and the documented process stops reflecting reality within a few months. The process and the pipeline need to be the same six stages, named the same way, with the exit criteria written directly into the CRM stage description where a rep actually sees it at the point of moving a deal, not filed away somewhere they have to go looking for it.

What good looks like. A worked scenario on a process that existed only on paper.

A B2B sales team I worked with, illustrative rather than a named client, had a documented six-stage process, the kind any sales operations wiki would be proud of, with clear stage names and a tidy CRM pipeline view to match. In practice, every experienced rep ran their own personal version of it. One treated "qualification" as complete once a prospect had taken a first call; another wouldn't move a deal past qualification without a signed NDA and three stakeholders on the line. Forecasting was consistently wrong in both directions, some reps' pipelines were inflated with deals that hadn't actually been qualified, others' were understated because their personal bar for "qualified" was stricter than the process required.

The fix wasn't a new process, the six stages were fine. It was writing a specific, checkable exit condition for every stage and then, critically, checking deals against those conditions in the weekly pipeline review rather than taking the stage label on trust. The decision rule that came out of it: a stage isn't real until someone other than the rep running the deal has verified the exit condition at least once per deal. Self-reported stage progress, even from good reps, drifts over time without that outside check, not from dishonesty but from each person's personal calibration quietly diverging from everyone else's.

Within two quarters of adding that check, forecast accuracy improved enough that leadership stopped discounting the pipeline by a standard "sandbagging factor" before presenting it upward, because the stages had started meaning what they claimed to mean again.

The other change worth naming is what happened to average deal cycle time, which actually lengthened slightly in the first quarter after the new exit criteria went in, before shortening below its original baseline the quarter after that. The initial lengthening wasn't a problem, it was deals that had been sitting inflated at a later stage than they'd actually earned finally getting pushed back to where they genuinely belonged, which looked like the process slowing things down when it was really just correcting a backlog of mislabelled deals. Once that one-off correction worked through the pipeline, the shorter cycle time that followed reflected deals moving through real, verified stages rather than stalling unnoticed inside a stage nobody was checking.

Pitfalls to avoid. Where defined sales process steps quietly fail.

The first pitfall is writing the process once and never checking adherence to it. A documented process with no inspection mechanism behind it is a wiki page, not an operating discipline, and Supered's 6.3x quota-attainment gap between teams that inspect and teams that don't is the clearest evidence that the checking, not the documentation, is where the value actually sits.

The second is defining exit criteria around rep activity rather than buyer response. "Sent a proposal" is an activity a rep controls entirely on their own; "buyer has formally reviewed the proposal" is a response that proves the stage actually happened, and only the second kind of criterion catches a deal that's stalled rather than progressing.

The third is building a process with too many or too few stages relative to how the buyer actually decides. A process with ten stages for a straightforward, fast-moving deal adds administrative overhead without adding insight; a three-stage process for a complex enterprise sale hides real friction inside one oversized "in progress" stage, where a deal can sit for months without anyone noticing it's actually stuck.

The fourth is treating the process as fixed once it's agreed, rather than reviewing it against how deals are actually won and lost. A process built around how the business wanted to sell two years ago, with no update since, usually still describes last year's buyer, not this year's, particularly once a new competitor or a changed budget cycle shifts how prospects actually move through a decision. A short, recurring review, looking at closed-won and closed-lost deals against the stage definitions every couple of quarters, catches this before the process quietly becomes fiction.

The fifth, specific to founder-led teams, is the founder's own deals quietly skipping the process because "I know what I'm doing". That's usually true individually, and it still undermines the process for everyone else, since reps notice when the person who wrote the rules doesn't follow them, and the standard erodes fastest from the top down, not the bottom up.

The sixth is applying one process to every deal segment regardless of size or complexity. A straightforward, low-value renewal and a complex, multi-stakeholder enterprise sale genuinely need different depth at the qualification and negotiation stages, and forcing both through an identical six-step checklist usually means the simple deal carries pointless overhead while the complex one skips steps it actually needed. Most teams that get this right keep the same six stage names across every segment, for consistent reporting, but vary how rigorous the exit criteria are within each stage depending on deal size, so a small renewal clears qualification with a lighter check than a six-figure new-logo deal would need to.