Most b2b sales pipeline designs have 7-9 stages. The best have 4-5. The difference isn’t cosmetic — it’s whether you can forecast from it. RevHeat’s client results show that 95% of pipeline was unqualified before we simplified the stages. After simplification, 100% became qualified and predictable. The problem isn’t that your reps are bad. Nobody has time to check whether the last call actually moved the deal forward. Most calls just move deals sideways into the next stage.
Key Takeaway: A b2b sales pipeline works when each stage has one exit question. The rep can’t advance the deal until the buyer answers it. RevHeat’s client results show 95% of pipeline was unqualified before simplifying stages to 4-5 with one question each. Afterward, 100% became qualified and predictable. Qualifying shows a 150% performance gap between elite and average salespeople.
TL;DR
- 95% of pipeline was unqualified before RevHeat simplified stages; 100% became qualified after each stage got one exit question
- Most pipelines have 7-9 stages; the best have 4-5, each with a single question the buyer must answer before advancing
- Qualifying shows a 150% performance gap between top and bottom performers — elite reps disqualify bad deals early
- Your forecast accuracy depends on stage design, not rep judgment; one exit question per stage removes guesswork
Prerequisites / What You Need
Before you redesign your b2b sales pipeline, you need three things in place.
First, access to your last 20 closed-won and 20 closed-lost deals. You’ll reverse-engineer the stages from what actually happened. Not from what you wish happened.
Second, a call recording tool like Gong, Chorus, or Fireflies. You can’t write exit questions without hearing what buyers actually said. You need to hear the moment they moved forward.
Third, one hour with your best rep. The rep who closes at 40%+ knows the questions that matter. The playbook just never wrote them down.
Most teams skip the first two. They build stages from a template. That’s why 95% of pipeline is unqualified, according to RevHeat’s client results.
Step-by-Step: Build Your B2B Sales Pipeline
Step 1: Pull Your Last 20 Wins and Listen for the Turn
Open your CRM. Pull the last 20 closed-won deals. For each one, find the call where the buyer stopped being polite. Find where they started being specific. That’s the turn.
The turn is the moment they said a number. Or named a date. Or admitted what wasn’t working. Write down the question your rep asked right before that moment.
You’re not looking for the question that got them to agree to a demo. You’re looking for the question that made them say something real. Something they wouldn’t say to a vendor they were just being nice to.
In most cases, it’s a dollar figure. Or a deadline. Or the name of the person who has to approve it.
Do this for all 20 wins. You’ll see 3-5 questions that show up in every deal. Those are your stage exit questions.
Step 2: Reverse-Engineer Stages from the Questions
Now take those 3-5 questions. Put them in chronological order. The order they happened in the deal, not the order you wish they happened.
Each question becomes a stage. The stage name is just a label. The exit question is what matters.
Here’s what that looks like for most B2B teams:
| Stage | Exit Question | What the Buyer Said |
|---|---|---|
| Discovery | “What happens if you don’t fix this by [date]?” | A consequence with a dollar figure or a named person who’s asking |
| Qualified | “Who besides you has to say yes?” | Names, titles, and whether they’ve been briefed |
| Proposal | “If we solve [specific problem], what does approval look like on your end?” | A process with steps and a timeline |
| Negotiation | “What’s the one thing that would stop this from happening?” | A specific objection, not a stall |
If your pipeline has more than 5 stages, you have stages that don’t move deals. They just move them sideways. Cut them.
According to research by Gradient Works (2024), the average B2B win rate is 20.7% across all sources. If your pipeline has 7-9 stages and your win rate is below that, stage design is the problem.
Step 3: Write the Exit Criteria — One Question Per Stage
For each stage, write the one question the rep must ask. Write the one answer the buyer must give before the deal advances. If the buyer doesn’t answer it, the deal doesn’t move.
This is the part most teams skip. It’s why 95% of pipeline is unqualified.
The exit question has to be something the buyer says out loud on a call. Not something the rep infers. “They seemed interested” is not an exit answer.
“They said the CFO has to approve anything over $50K” is an exit answer. “We’re meeting her next Tuesday” is an exit answer.
Here’s the rule: if you can’t play the call recording and hear the buyer say the answer, the deal hasn’t exited the stage.
RevHeat’s analysis of client calls shows a pattern. When reps advance deals without hearing the exit answer, the deal dies two stages later. By then, it’s in your forecast.
Step 4: Train the Exit Question, Then Check the Next Call
Once you’ve written the exit questions, you have to train them. That means every rep gets the question, word for word. They practice asking it on a role-play call. Then they ask it on a real one.
Then you listen to the next real call. You check whether they asked it.
This is where most sales training dies. Nobody checks whether the rep actually did it.
According to RevHeat’s State of Sales Skills original research, only 6% of all salespeople possess the complete skill set required for elite performance. The research draws on a benchmark of 2.5 million sellers across 33,000 companies.
The gap isn’t that they don’t know the question. The gap is that nobody checked whether they asked it on the next call.
Here’s how we do it at RevHeat. Every rep gets one exit question to practice this week. We listen to their next 3 calls. If they asked it on 2 out of 3, they’re running the playbook.
If they didn’t, we hand them the question again. We check the next 3 calls. Catch the slip before it becomes a habit.
Step 5: Rebuild Your Forecast from Stage Exit Rates
Once every stage has an exit question, you can forecast from exit rates. Not from rep judgment.
Pull your CRM data for the last 90 days. For each stage, calculate what percentage of deals that entered the stage also exited it. They exited by answering the exit question.
That percentage is your stage exit rate. It’s the only number you can forecast from.
If 60% of deals that enter Discovery exit it, you have data. If you have 20 deals in Discovery right now, you can forecast 12 deals moving to Qualified.
Your forecast accuracy just went from “whatever the rep hopes” to “what actually happened in the last 90 days.”
According to our sales forecast accuracy research, teams that forecast from stage exit rates cut their forecast error by half. This happens in the first quarter. The reason is simple: exit rates are based on what buyers said, not what reps think.
Common Mistakes to Avoid
Mistake 1: Adding Stages to Match Your Internal Process
Your internal process doesn’t matter to the buyer. If your legal team needs 2 weeks to review contracts, that’s not a stage. That’s a task.
A stage is a place where the buyer makes a decision. If the buyer isn’t deciding anything, it’s not a stage.
Most pipelines have 7-9 stages because they map to internal handoffs. “Demo Scheduled.” “Proposal Sent.” “Contract Out for Signature.” None of those are decisions the buyer made. They’re things you did.
Cut them.
The test: if the buyer didn’t say something new, it’s not a new stage.
Mistake 2: Writing Exit Criteria the Rep Can Check Without Talking to the Buyer
“Proposal sent” is not exit criteria. “Buyer confirmed they received the proposal and will review it with the CFO by Friday” is exit criteria.
If the rep can check the box without getting the buyer on the phone, it’s not an exit question. It’s an activity metric.
RevHeat’s client results show a pattern. When exit criteria are based on rep activity instead of buyer answers, 95% of pipeline is unqualified.
The fix: every exit question has to produce an answer the buyer says out loud. On a recorded call. That you can play back.
Mistake 3: Skipping the Check on the Next Call
You wrote the exit question. You trained the rep. You assumed they’d ask it. They didn’t.
Now the deal is in your forecast. It’s going to die in two stages.
The gap isn’t that reps don’t know what to ask. According to RevHeat’s State of Sales Skills original research, 94% of sellers have at least one critical gap. Most have 3-5 gaps that compound.
The gap is that nobody checked whether they asked it on the next call. Every team has a sales playbook. Nobody checks if anyone runs it.
Here’s the fix: listen to the next 3 calls after you train the exit question. If the rep asked it on 2 out of 3, they’re running the playbook.
If they didn’t, hand them the question again. Check the next 3 calls.
Mistake 4: Forecasting from Stage Names Instead of Exit Rates
Your CRM says 30 deals are in “Qualified.” Your forecast says 20 of them will close. Your actual close rate is 6.
The problem isn’t that your reps lied. The problem is that “Qualified” doesn’t mean anything. Not unless you define what the buyer had to say to get there.
Stage names are labels. Exit rates are data.
If 60% of deals that enter Discovery exit it, you have data. If you have 20 deals in Discovery, you can forecast 12 moving to Qualified. That’s math. Everything else is a guess.
Our how to calculate win rate research shows results. Teams forecasting from exit rates instead of stage names cut their forecast error by 40-50%. This happens in the first 90 days.
Mistake 5: Designing Stages Before You Listen to the Calls
Most teams design their pipeline in a conference room. They write stages that sound right: Awareness, Interest, Consideration, Decision.
Then they wonder why deals die in “Consideration.” Nobody knows why.
The fix: pull your last 20 wins. Listen to the calls. Write down the questions your rep asked right before the buyer said something specific.
Those questions become your exit criteria. The stages are just labels for the questions.
If you design stages before you listen to the calls, you’re building a pipeline based on what you wish buyers did. Not what they actually do. That’s why 95% of pipeline is unqualified, according to RevHeat’s client results.
Frequently Asked Questions
How many stages should a b2b sales pipeline have?
4-5 stages, each with one exit question. Most pipelines have 7-9 stages because they map to internal handoffs, not buyer decisions.
If the buyer didn’t say something new, it’s not a new stage. RevHeat’s client results show that simplifying from 7-9 stages to 4-5 with one exit question each took pipeline qualification from 95% unqualified to 100% qualified and predictable.
What’s the difference between an exit question and exit criteria?
An exit question is what the rep asks. Exit criteria is what the buyer says in response.
The deal doesn’t advance until the buyer answers the question out loud on a call. “Proposal sent” is not exit criteria — it’s an activity.
“Buyer confirmed they’ll review the proposal with the CFO by Friday” is exit criteria. Because the buyer said it.
How do I know if my pipeline stages are working?
Pull your stage exit rates for the last 90 days. If 60% of deals that enter Discovery exit it, your stages are working. If 40% of deals that enter Qualified exit it, your stages are working.
If your exit rates are below 40% at any stage, that stage has a problem. Either it has no exit question, or the exit question isn’t clear enough for reps to ask consistently.
Our sales pipeline best practices research shows that stage exit rates below 40% indicate unclear exit criteria.
What if my reps say they’re already asking the exit questions?
Listen to the next 3 calls. If they asked it on 2 out of 3, they’re running the playbook.
If they didn’t, hand them the question again. Check the next 3 calls.
According to RevHeat’s State of Sales Skills original research, only 6% of all salespeople possess the complete skill set required for elite performance. The gap isn’t that they don’t know the question. The gap is that nobody checked whether they asked it.
Can I use a standard pipeline template from my CRM?
No. CRM templates are built for every company. Which means they fit no company.
Your pipeline has to be reverse-engineered from your last 20 wins. Not copied from a template.
If you use a template, you’ll have stages that don’t match what your buyers actually do. 95% of your pipeline will be unqualified. Pull your wins, listen to the calls, and write the exit questions from what actually happened.
How long does it take to redesign a b2b sales pipeline?
4-6 hours if you follow the steps. 1 hour to pull your last 20 wins and listen for the turn. 1 hour to reverse-engineer stages from the questions.
1 hour to write the exit criteria. 1-2 hours to train the exit questions and role-play them with your reps.
Then you check the next 3 calls per rep to see if they’re asking the questions. Most teams spend 6 months tweaking stages in their CRM. They never check whether reps are asking the exit questions on calls.
What’s the biggest mistake teams make when designing pipeline stages?
Adding stages to match internal processes instead of buyer decisions. If your legal team needs 2 weeks to review contracts, that’s not a stage. That’s a task.
A stage is a place where the buyer makes a decision. If the buyer isn’t deciding anything, it’s not a stage.
Most pipelines have 7-9 stages because they map to internal handoffs, not buyer answers. Cut them.
How do I forecast from stage exit rates instead of rep judgment?
Pull your CRM data for the last 90 days. For each stage, calculate what percentage of deals that entered the stage also exited it by answering the exit question.
That percentage is your stage exit rate. If 60% of deals that enter Discovery exit it, and you have 20 deals in Discovery right now, you can forecast 12 deals moving to Qualified.
Our how to forecast sales research shows that forecasting from exit rates cuts forecast error by 40-50% in the first 90 days.
What if my buyers don’t follow a linear path through the stages?
They don’t have to. The stages are just labels for the questions.
If a buyer answers the exit question for Stage 3 before they answer the exit question for Stage 2, mark both stages complete. Move them to Stage 4.
The point isn’t to force buyers into a linear path. The point is to make sure they’ve answered every exit question before you put them in your forecast.
How do I get my reps to actually ask the exit questions on calls?
Train the question, then check the next call. Hand them the question word for word. Role-play it on a practice call.
Then listen to their next 3 real calls. Check whether they asked it. If they asked it on 2 out of 3, they’re running the playbook.
If they didn’t, hand them the question again. Check the next 3 calls.
RevHeat’s State of Sales Skills research measured 21 core sales competencies across a weak-to-strong spectrum, benchmarking performance between the bottom 10% and top 10% of sellers. Qualifying shows a 150% performance gap.
Elite salespeople are ruthless about pipeline quality. They disqualify bad deals early. The gap isn’t that they don’t know the question. The gap is that nobody checked whether they asked it.
Bottom Line
A b2b sales pipeline works when each stage has one exit question. The rep can’t advance the deal until the buyer answers it.
RevHeat’s client results show 95% of pipeline was unqualified before we simplified stages to 4-5 with one question each. Afterward, 100% became qualified and predictable.
Same reps. Same leads. Same market. Growth you can plan on.
Ken Lundin is CEO of RevHeat. Over 25 years he has sold, run sales teams and helped 200+ teams, including 5 unicorns, generate $1.5B in client sales, with two exits ($250M and $30M) and four appearances on the Inc. 500. He is a Forbes Business Council member and the author of Strategic Selling Unleashed, Broken Playbook and The SmartScaling Formula.