A client came to us last year. Forty people, $3M ARR, growing fast enough that their Series A was six weeks out. Their CEO called because close rates had dropped from 28% to 19% over the previous three quarters. He had blamed it on the market. Then on the product. Then on two specific reps.
When we got into HubSpot, the real answer was obvious within about twenty minutes. Sixty percent of open deals hadn't been touched in 30 days. Average deal count per rep was 51. Stage names were things like "In Progress," "Following Up," and one that just said "Hot." The pipeline total was $2.1M. We estimated about $600K of it was real.
I asked their four AEs to each describe their sales process. Got four different answers. One started at outbound. One started at discovery. One didn't really have a step two. The fourth gave me a perfectly articulate six-step answer that matched none of what I saw in the CRM.
That company didn't have a sales process. They had four individual sales habits averaged together and called a pipeline.
Why most 50-person B2B companies have an inherited process
Here's what typically happens. A founder or early VP of Sales closes the first 20-30 deals. They develop a feel for what works. Then they hire a rep, who copies some of what they saw. Then another rep joins and copies that version. By the time you have four AEs, you have four variations of something that worked for one person in a specific market moment, none of which is documented, and none of which match each other.
Nobody designed the process. It accumulated.
This isn't a discipline problem. It's a documentation problem. Most early B2B companies prioritize closing over systematizing. That's the right call at 10 employees. It's the wrong call at 40.
The shift from founder-led to team-led sales almost always requires rebuilding the process from scratch. What worked when the founder was selling isn't inherently repeatable, because founders carry context in their heads that reps don't have.
The difference between a sales process and a sales habit
A sales habit is something you do because it worked once. A sales process is a documented sequence of stages with clear entry criteria, exit criteria, and the buyer actions that define each.
The key word there is "buyer." Most SMB B2B sales processes are built around what the rep does, not what the buyer experiences. "Sent follow-up email" is a rep action. "Prospect confirmed budget and timeline" is a buyer commitment.
If your deal stages are defined by what your reps do, you'll always have a pipeline that looks fuller than it is. Reps will move deals forward by sending emails, not by getting real signals from buyers.
I think about it this way: a deal stage should only advance when the buyer does something. Not when the rep does something. That one mental shift fixes about half the pipeline accuracy problems I see.
What the stages should actually look like
There's no magic number. For most B2B companies with 30-90 day sales cycles, five stages is right. Fewer if the deal is transactional. More if you're selling to enterprise buying committees.
Here's the model I've landed on after using it with a dozen companies:
Stage 1: Qualified (10%) Entry criteria: the contact has confirmed a specific problem worth solving and has the authority to act on it. "Expressed interest" doesn't qualify. The rep must be able to write one sentence about the exact problem.
Stage 2: Discovery done (25%) Entry criteria: a real discovery conversation happened and is documented. The rep can answer: what problem are they solving, by when, who else decides, and is there a budget? All four. Not three.
Stage 3: Solution confirmed (45%) Entry criteria: the prospect has seen a proposal or demo tailored to their situation and given a verbal that it matches what they need. The key word is "confirmed," not "sent." Sending a proposal doesn't advance anything.
Stage 4: Decision process agreed (65%) Entry criteria: the prospect has told you their evaluation process. You know who signs, what internal approval looks like, and what would need to be true for this to close this quarter.
Stage 5: Contract out (85%) Entry criteria: commercial terms sent. That's it. Nothing else belongs here.
The common mistake: activity stages dressed up as pipeline stages
The version of this I see most often is 8-10 stage pipelines with names like "Demo Scheduled," "Demo Completed," "Proposal Sent," and "Proposal Follow Up." They've mistaken activity tracking for process definition.
Eight stages where every advance is a rep action is worse than five stages with real buyer criteria. You end up with reps spending time updating the CRM instead of selling, and a forecast that's still inaccurate because none of the stage moves meant anything to the buyer.
One thing I always push back on: having a "Nurturing" stage inside the sales pipeline. Nurturing is a marketing function. If a deal is "in nurturing" it either belongs in a HubSpot sequence or it's dead. Let it be dead. A pipeline with 30 real deals is worth more than one with 80 zombie opportunities padded in for comfort.
A smaller, accurate pipeline is worth more than a padded one.
Most teams resist cutting ghost deals because the pipeline looks smaller. But forecasting off a padded pipeline just means you're always surprised at quarter end, and your reps are always explaining why those deals "should have closed."
How to build this in HubSpot
HubSpot's deal stages give you a stage name and a close probability. What most teams miss is the descriptions feature, which lets you define exactly what must be true for a deal to sit in each stage.
Go to Settings > CRM > Pipelines and click into each deal stage. Add a description with the specific buyer actions required to enter that stage. This becomes the team's source of truth and makes onboarding a new rep significantly faster.
Then do two things:
First, run a pipeline scrub. For every deal currently in stages 3-5, ask the rep: "What specific buyer action put this here?" If they can't answer, the deal moves back. This will be uncomfortable. The pipeline number will drop, sometimes by 40%. That's the point.
Second, set up a HubSpot workflow that flags any deal with no activity in 14 days and stamps a "Stalled" property. Give your sales manager a weekly view of this list. Manual deal reviews should be a last resort, not the primary inspection mechanism.
For teams that need more depth, our CRM and RevOps setup work typically includes a full pipeline architecture review, stage criteria documentation, and HubSpot implementation. We also build the coaching dashboards that make this visible week over week.
The metrics that tell you whether your process is working
Once you have clear stage criteria, three metrics tell you if the process is actually being followed:
Time in stage. If deals are spending three times longer in Stage 2 than in Stage 3, discovery is your bottleneck. Either it's not happening or reps are skipping it. Look at this by rep and you'll see exactly who's following the process and who's improvising.
Stage-to-stage advancement rate. What percentage of deals entering Stage 2 make it to Stage 3? If the answer is 85%, either your criteria are too loose or you're not qualifying hard enough at Stage 1. If the answer is 20%, that's where you're losing deals and it deserves all your attention.
Win rate by deal source. Most companies track overall win rate. That's a blended average that hides the real story. Inbound deals from referrals often close at 35-40%. Cold outbound deals often close at 12-15%. If you're treating them the same in your pipeline, your forecast will always be off.
These metrics don't mean anything until your stage criteria are consistent. Once they are, they become the most useful data you have. We cover how to surface this in HubSpot as part of our RevOps and AI automation services for growth-stage companies.
When qualification frameworks actually help (and when they don't)
Teams at this stage often get excited about MEDDIC, BANT, or SPICED and want to layer one of them on top of what they're building. My honest advice: don't do this yet.
Qualification frameworks are genuinely useful for enterprise deals above $50K ACV with multi-stakeholder buying committees. For SMB deals with 30-60 day cycles and one or two decision makers, they add complexity without adding clarity.
Build stage criteria that are buyer-defined first. Once you have 60-90 days of data and the team is using the process consistently, look at whether you need a structured qualification layer. By that point, you'll know exactly where deals are dying and whether a framework would actually help.
The goal at 50 people isn't the most sophisticated sales process in the industry. The goal is one that every rep follows the same way, generates forecasts you can trust, and gives your manager clear coaching signals. That's achievable in a few weeks. What makes it fail almost every time is launching it without getting rep buy-in first.
Getting rep buy-in
This is the part that gets skipped. You can design the best stage criteria in the world and fail to implement it because the reps don't believe in it.
The fastest way to get buy-in: involve one or two of your top reps in designing the criteria. Not just reviewing what you built. Actually building it with you. Ask them: "What does a buyer do or say that tells you this deal is real?" Their answers become your criteria. Then when rollout happens, those reps are advocates rather than resistors.
The second thing that helps: commit to using the data to help reps, not just to monitor them. If your first pipeline review after the new process rolls out is "why do you have so few deals in Stage 3," you'll lose the room. If it's "I see you have six deals stalled in Stage 2, let's talk about what happened in those discovery calls," you'll get traction.
A well-designed B2B sales process doesn't constrain good reps. It gives average reps a framework to perform better and gives great reps the data to coach others. That's the version worth building. It takes a few weeks of deliberate work and then pays dividends for years.
If you're building this from scratch or rebuilding after a painful period of pipeline fog, our go-to-market practice can help you design the process and get it into your CRM the right way. We've done this at enough companies to know which shortcuts create problems later.
Is your pipeline telling you the truth?
Book a free 30-minute audit and we will show you the three things we would fix in your sales process first.
Book a free audit →FAQ
What is a B2B sales process?
A B2B sales process is a documented sequence of stages a deal moves through from first contact to close, where each stage advance requires a specific buyer action or commitment. It's different from individual rep habits in that it's consistent across your team and gives you predictable forecasting.
How many stages should a B2B sales pipeline have?
For most B2B companies with 30-90 day sales cycles, five stages works well. Fewer if you're selling a transactional product. The number matters less than whether each stage has clear buyer-defined entry criteria that every rep uses consistently.
How do I document my sales process in HubSpot?
In HubSpot, go to Settings > CRM > Pipelines and add a description to each deal stage. These descriptions should define the specific buyer actions required to enter that stage. Run a pipeline audit against these criteria every quarter to remove deals that don't qualify.
What's the difference between a sales process and a sales cadence?
A sales process defines the stages a deal moves through from qualified to closed. A sales cadence is the sequence of outreach touchpoints you use to get a prospect to engage in the first place. The cadence gets you into the pipeline. The process governs how you move through it.
When should a B2B company redesign its sales process?
The most common triggers are declining close rates, pipeline accuracy problems, or the transition from founder-led to team sales. A practical signal: ask each rep how they move a deal from one stage to the next. If you get different answers, it's time to redesign.