The sales process audit: how to find where deals are dying
Table of Content
Why pipeline leakage is so hard to find without a structured audit
Most revenue leaders know something's wrong before they can prove it. Close rates are slipping. Deals that looked solid three weeks ago are now "waiting on legal." Your forecast keeps coming in 15-20% light, but nobody on the team can point to a single reason.
This is pipeline leakage, and it's expensive in ways that don't always show up on a dashboard. According to Gartner's B2B sales research, the average B2B deal involves 6-10 stakeholders and spans several months. Each additional touchpoint is another chance for a deal to go cold, get deprioritized, or simply fall through a process gap that nobody's documented.
The frustrating part is that most teams react to symptoms instead of causes. They add more pipeline at the top. They pressure reps to close faster. They restructure comp. None of it works because the problem isn't effort or attitude. It's a specific breakdown at a specific stage that nobody's measured.
A sales process audit fixes that. It's a structured diagnostic that tells you exactly where deals are dying, why, and what to change first. You don't need a consultant to run one, and you don't need six months. A focused team can complete it in a week.
If you've been trying to improve forecast accuracy without fixing the underlying process, the sales maturity model framework is worth reading alongside this audit — it gives you the broader context for where your process gaps fit in your organization's development.
What a sales process audit actually examines
A sales process audit is a structured review of your current pipeline stages, conversion metrics, deal velocity, and the quality of your CRM data. It isn't a rep performance review and it isn't a strategy session. It's diagnostic work.
The audit answers five specific questions:
- What percentage of deals convert from each stage to the next?
- How long do deals spend at each stage, broken down by segment?
- Where do the majority of lost deals die, and why?
- What do won deals have in common that lost deals don't?
- How reliable is your pipeline data as a forecasting input?
Those five questions map to the five steps of the audit process. You work through each one in sequence. The output isn't a presentation. It's a ranked list of constraints with enough evidence to decide which one to fix first.
Who should run it
The audit works best when a RevOps lead or a sales leader drives it, with input from a senior rep and someone from finance or CS who touches the post-sale process. You want people who know where the bodies are buried, not just what the CRM says.
Avoid running it as a committee. Two to three people working with the data for a week will find more than six people talking about the data for a month.
What the audit won't tell you
It won't tell you whether your product is competitive or whether your pricing is wrong. Those are separate questions. A sales process audit assumes your core offer has market fit. If it doesn't, fixing process won't help.
What it will tell you is whether your process is letting good-fit deals slip through cracks that shouldn't exist.
Audit scope: keep it tight
Focus the audit on closed deals and active pipeline from the last 90-180 days. Going further back introduces noise from team changes, product updates, and market shifts that don't reflect current conditions. If you're in enterprise with longer cycles, extend to 12 months but weight the last 6 more heavily.
Step 1: pull the right data before you touch anything else
Most audit attempts fail in the first hour because people pull the wrong data. They export their full CRM, stare at 4,000 rows of opportunities, and don't know where to start.
Pull four specific datasets:
Closed-won and closed-lost deals, last 90-180 days. For each deal you need: entry date per stage, exit date per stage, deal size, segment (SMB/mid-market/enterprise), first-touch source, sales rep, close reason. That's it. Nothing else matters at this step.
Active pipeline by stage. Current deals with their entry dates for each stage they've passed through. This tells you where deals are sitting today vs. historical averages.
CRM field completion rate. What percentage of opportunities have required fields filled in for each stage? This is your data quality baseline.
Rep-level conversion rates. Not to call anyone out, but to find outliers in both directions. A rep converting at 3x the team average at a specific stage knows something the others don't.
Common data problems you'll hit
Deals get manually moved backward in the pipeline when a rep wants to "reset" the timeline. Stages get skipped entirely. Close dates get pushed without any note or reason. These aren't minor annoyances. They're signals that your process governance is weak.
When you find a deal that jumped from demo to proposal without any record of a discovery call, that's not just messy CRM data. That's a rep who either skipped discovery or didn't document it. Either way, you need to know how common it is before you can assess the risk it creates.
Don't clean the data before you audit it. The mess is the information.
Step 2: stage conversion analysis and what normal looks like
Stage conversion analysis is the core of any sales process audit. You're measuring the percentage of deals that advance from each stage to the next, then comparing that against benchmarks and against your own historical patterns.
What to calculate
For each stage transition (e.g., Qualified Lead to Discovery, Discovery to Demo, Demo to Proposal, Proposal to Negotiation, Negotiation to Closed Won), calculate:
- Volume: how many deals entered the stage over the period?
- Conversion rate: what percentage advanced to the next stage?
- Average time in stage: how many days did advancing deals spend there?
- Average time for deals that didn't advance: how long before they stalled or closed-lost?
The gap between advancing deal velocity and stalling deal velocity is your first clue. If deals that convert spend 5 days in discovery but deals that don't spend 22 days, something happens around day 6 that breaks deals.
Benchmark conversion rates by stage
These aren't universal — they vary by deal size, segment, and industry. But they're reasonable starting points for B2B SaaS and professional services. Forrester's B2B sales benchmark research consistently shows that teams with defined stage criteria and documented exit conditions outperform those without by 15-20 percentage points on overall Win Rate.
Use the table below as a starting reference, not a ceiling. If you're consistently beating these, investigate why before assuming your process is fine. Unusually high early-stage conversion often means qualification standards are too loose, which inflates pipeline but kills close rates.
The most important number isn't your overall Win Rate. It's the stage where you lose the most deals relative to the benchmark. That's your constraint.
| Pipeline stage transition | Typical B2B SaaS conversion | Enterprise deal conversion | Warning threshold (below) |
|---|---|---|---|
| MQL to SQL | 20-30% | 10-15% | <15% (loose qualification) |
| SQL to Discovery/Qualified | 60-75% | 50-65% | <50% (sourcing or routing issue) |
| Discovery to Demo/Evaluation | 70-85% | 55-70% | <55% (discovery not connecting need) |
| Demo to Proposal | 50-65% | 40-55% | <40% (ICP mismatch or poor demo) |
| Proposal to Negotiation | 45-60% | 35-50% | <35% (proposal or champion gap) |
| Negotiation to Closed Won | 65-80% | 55-70% | <55% (legal, budget, or urgency failure) |
| Overall pipeline Win Rate | 20-30% | 12-20% | <15% (structural process problem) |
The table is a starting point, not a report card
If your overall Win Rate is 12% but your average contract value is $250K+, that may be perfectly healthy. Context matters. The comparison that matters most is your own conversion trend over 3-6 quarters. If a stage that used to convert at 65% now converts at 42%, that's a signal worth investigating regardless of what any benchmark says.
Step 3: cycle time by segment, not just the average
Average deal cycle time is one of the most misleading metrics in B2B sales. A single enterprise deal that closes after 14 months can make your mid-market average look three times longer than it actually is.
Break cycle time analysis by segment first. Then by deal source. Then by rep. You're looking for three things:
Stage-level velocity gaps. Which stage consistently takes longer than it should for deals that eventually close? If proposal-to-negotiation takes 30+ days for your mid-market segment when the historical average is 12, something changed. Maybe a competitor is getting in late, maybe your proposal quality dropped, maybe legal bottlenecks moved upstream.
Segment-source interaction. Outbound-sourced deals often move faster through early stages but stall at procurement. Inbound deals from your website close faster when they reach the right person, but that's a big "when." Knowing which source-segment combinations produce the fastest and highest-converting deals tells you where to focus pipeline generation efforts.
Rep-level cycle time by stage. If one rep's deals sit in "Demo Scheduled" for 20 days on average while the team average is 8, that's not a volume problem. That's a specific rep behavior at a specific stage. Worth a 30-minute conversation, not a new comp plan.
The stale deal problem
Deals that have been sitting in the same stage for 2x the historical average rarely close at anything approaching normal rates. Most teams let these inflate their pipeline without questioning them.
As part of your cycle time analysis, flag every deal that has exceeded the stage velocity benchmark by 2x or more. That's your zombie pipeline. You'll need to make a decision about each one: pursue with a clear next step, or close-lose it and clean the forecast.
Worth noting: decisions about deal selection and qualification discipline — which directly affect cycle time downstream — are covered in depth in the strategic sales focus framework.
Step 4: win/loss interviews that actually tell you something
Most win/loss programs are theater. You ask the rep why a deal was lost. They say "price" or "went with a competitor." You log it and move on. Nothing changes.
There are two problems with that approach. First, reps rarely know the real reason a deal was lost. Second, even when they do, they have incentives to attribute losses to factors outside their control.
Useful win/loss analysis requires talking to the buyer, not just the rep. Even a 15-minute conversation with a lost buyer tells you more than 10 rep debriefs.
Questions that get real answers
For lost deals, ask buyers:
- What was the problem you were trying to solve when you started this evaluation?
- At what point did you decide we weren't the right fit?
- What would have needed to be different for us to win the deal?
- Who else was involved in the final decision, and what were their main concerns?
For won deals, ask buyers:
- What made you confident enough to move forward?
- What almost stopped you from buying?
- Who in your organization had the strongest concerns, and what resolved them?
The answer to "what almost stopped you" in won deals is as valuable as any lost deal analysis. It tells you what your sales process is successfully handling and what nearly broke it.
What to do with close reasons from the CRM
CRM close reasons are directional at best. They tell you the story the rep chose to tell. Look at them for volume patterns, not diagnostic truth. If "budget" appears in 40% of lost deals, you might have a pricing conversation problem or a champion-building problem, not actually a budget problem. Win/loss interviews will tell you which.
For a deeper look at what happens specifically when deals fall apart at the proposal stage, the sales proposal process optimization guide covers the mechanics of fixing that specific gap.
Step 5: CRM hygiene check and what dirty data hides
CRM hygiene is the least glamorous part of a sales process audit and the part most teams skip. That's a mistake, because bad CRM data doesn't just make your reports wrong. It actively hides the problems you're trying to find.
Five things to check
Required field completion. What percentage of active deals have all stage-required fields filled in? For any field that drives your qualification process (budget confirmed, decision-maker identified, timeline documented), low completion rates mean your qualification is happening in salespeople's heads, not in your system. You can't manage what you can't see.
Stage date accuracy. Were stage transitions logged in real time or backdated? If most of your deals moved through three stages on the same day, someone's batch-updating the CRM at the end of the month. That destroys cycle time data.
Duplicate and stale opportunities. How many active opportunities haven't had any activity logged in 30+ days? This is your zombie pipeline. In most teams it's 20-40% of active pipeline, which means your coverage ratio is much worse than your CRM shows.
Close date discipline. How often do close dates get pushed? If the average deal has its close date extended 3+ times, your forecast is a fiction. Reps learn to set dates that satisfy the CRM without reflecting real buyer intent.
Lost deal documentation. For closed-lost deals, how many have an actual reason logged vs. just a status change? If 60%+ have no documented reason, you can't learn from losses at scale.
What dirty data actually costs you
A team with 35% CRM data completeness on key fields is effectively flying without instruments. Every pipeline review, every forecast call, every board presentation is built on a guess dressed up as a number. When leadership makes headcount or channel investment decisions based on that data, the error compounds.
CRM hygiene isn't a RevOps project. It's a sales leadership accountability issue. If the VP of Sales doesn't demand clean data, it won't happen.
The three most common failure patterns and what they mean
After running sales process audits across B2B teams, three patterns show up repeatedly. Recognizing them early saves weeks of analysis.
Pattern 1: deals die at proposal, which means discovery failed
If your highest drop-off stage is Proposal to Closed, the temptation is to fix the proposal. Better templates, faster turnaround, more competitive pricing. That's usually the wrong fix.
Deals die at proposal when discovery didn't surface what actually matters to the buyer. If the rep doesn't understand the business pain, the economic case, and the internal politics before writing a proposal, no template saves it. The proposal either addresses the wrong problem or goes to the wrong person.
Fix: audit your discovery calls. Are reps asking about business outcomes or just confirming feature fit? Are they identifying the economic buyer or just working with whoever picked up the phone?
Pattern 2: deals stall at legal, which means there's no internal champion
Deals that reach legal and then sit there for 45-90 days usually don't have a champion. They have a contact. A champion is someone who is actively pushing the deal internally, navigating procurement, accelerating legal review, and removing blockers. A contact is someone who answers your emails.
Without a champion, legal review moves at the pace of the organization's default bureaucratic process, which is slow. The rep can't accelerate it from outside. Only the buyer can.
Fix: qualify for champion earlier. Before you get to proposal, you should know who in the buyer's organization has a personal stake in the outcome of this purchase. If you can't name that person, the deal is at risk.
Pattern 3: ghosting after demo, which means ICP mismatch
When buyers go silent after a strong demo, it's tempting to blame the demo itself. Tighten the pitch, add more social proof, improve the follow-up sequence. That's often the wrong lever.
Post-demo ghosting at scale usually means you're running demos for companies that shouldn't be in your pipeline. They were interested enough to take a call but not interested enough to justify the purchase. The product fit isn't there, or the timing isn't there, or the buying process doesn't exist yet.
Check: what's the job title, company size, and growth stage of your post-demo ghosts vs. your post-demo conversions? You'll usually find a clear ICP boundary. Deals inside that boundary convert. Deals outside it go dark. Tighten your qualification criteria before demo, not after.
If this pattern is showing up in your pipeline, the strategic sales focus framework can help you define where you should and shouldn't be competing.
One pattern, one fix at a time
When you run an audit and find two or three failure patterns, it's tempting to fix them all at once. That's how you get a six-month initiative with no measurable outcome. Pick the stage where the most deal value is dying, fix the process there, and measure the impact over 60-90 days before moving to the next constraint.
How to identify your single biggest constraint from audit data
Every pipeline has one constraint that matters more than everything else. Not two constraints, not a set of overlapping issues. One primary bottleneck that, if fixed, improves everything downstream. The goal of your audit is to find it.
Here's a practical method that works consistently:
Calculate deal value lost per stage. Take every deal that entered a stage and didn't advance. Multiply the count by your average deal size. That's the approximate revenue value bleeding out at that transition.
For example: 40 deals entered the Proposal stage. 22 didn't advance. Average ACV is $45K. You lost approximately $990K in annual contract value at that one transition over 90 days. That's your priority number.
Rank all stage transitions by revenue lost. The one at the top of that list is your constraint. Not the one that feels most painful, not the one that's easiest to fix. The one that's costing you the most.
Cross-reference with cycle time. If the highest-revenue-loss stage also has 2x the expected dwell time, that confirms it. Long dwell time plus high drop rate equals your biggest constraint.
Check rep variance. If one or two reps are outperforming at the constraint stage, they have a method that's working. Replicating it is faster than inventing a new one.
This is fundamentally a Theory of Constraints approach applied to pipeline. Fix the bottleneck, and the whole system improves. Skip the bottleneck to fix something else, and you've done a lot of work for minimal revenue impact.
For teams that need an external perspective on identifying and prioritizing these constraints, CRO advisory engagements often start with exactly this kind of diagnostic work.
Need help running a sales process audit?
If your pipeline has consistent conversion problems and you'd rather run this with an experienced operator than figure it out solo, that's what advisory engagements are built for. The audit is usually the first deliverable.
Explore advisory optionsOne-week sales process audit: a day-by-day timeline
Here's a realistic schedule for a team that can dedicate focused time to this work. You won't need external resources and you won't need to pause selling activity.
Day 1 (Monday): data pull and prep
Export closed-won, closed-lost, and active pipeline data for the last 90-180 days. Pull CRM field completion rates. Assign one person to build the stage conversion spreadsheet and one to start the cycle time breakdown. Don't analyze yet. Just get the data into a usable format.
Day 2 (Tuesday): stage conversion analysis
Calculate conversion rates for every stage transition. Compare to your historical baseline (prior quarter or prior year). Flag any stage where current conversion is more than 10 percentage points below historical. That's your watchlist.
Also calculate deal value lost per stage using the method above. Rank all transitions by revenue loss. You'll have your primary constraint candidate by end of day.
Day 3 (Wednesday): cycle time and rep variance analysis
Break cycle time by segment and by rep. Identify stages where dwell time has increased vs. the historical baseline. Flag all deals that are 2x over the stage velocity benchmark. These are your zombie deals. Decide by end of day which ones get a defined next step vs. which ones get closed-lost.
Day 4 (Thursday): win/loss interviews
Schedule and run 4-6 buyer interviews. Two lost deals, two won deals, one deal that stalled and restarted, one deal where the expected close date moved 3+ times. Twenty minutes per call is enough if you use the questions from Step 4. Document the pattern themes, not every answer.
Day 5 (Friday): synthesis and constraint identification
Review all five data sets together. Write down your constraint hypothesis in one sentence: "The single biggest driver of pipeline loss is [stage transition] because [root cause], which costs approximately [$X] per quarter." Share it with your CRO, VP Sales, or RevOps lead. Get one challenge. Update if needed. That's your audit output.
The whole process takes roughly 20-30 hours of focused work spread across a team of two to three people. The output is a specific, evidence-backed diagnosis. Not a list of things to work on. One thing to fix first.
What to fix first after the audit is done
The audit gives you a diagnosis. Now you need a decision about treatment. There are three types of fixes, and they require different approaches.
Process fixes (fastest to implement)
If deals are dying because a step isn't happening consistently — discovery calls being skipped, proposals going out without documented business cases, next steps not being scheduled before demos end — you have a process compliance problem. The fix is clearer criteria, manager reinforcement, and short feedback loops.
Process fixes typically show results within 30-60 days because you're changing rep behavior, not rebuilding systems.
Capability fixes (medium timeline)
If deals are dying because reps don't know how to have a specific conversation — negotiating against procurement, building internal business cases, navigating multi-stakeholder deals — you have a skills gap. No process change fixes a skills gap. You need targeted coaching or enablement on the specific capability that's failing at the constraint stage.
Capability gaps take 60-90 days to address meaningfully. Expect some rep variance before the coaching takes hold.
Structural fixes (longest timeline)
If deals are dying because your ICP is too broad, your qualification criteria don't match your win profile, or your stage definitions encourage gaming instead of accurate reporting — that's a structural problem. These take 90-180 days to address because you're changing how the whole team thinks about the pipeline.
The most common structural problem found in audits is qualification standards that are either too loose (everyone gets a demo) or too rigid (only companies that match last year's wins get through). Both destroy conversion in different ways.
Whatever type of fix you're making, build a 90-day measurement plan before you start implementing. Define what success looks like at the constraint stage, how you'll measure it, and when you'll review. Audits without measurement plans tend to produce short-term behavior changes that revert within two quarters.
If what you find in the audit points to broader organizational readiness questions, the sales maturity model framework helps contextualize which fixes your team is actually ready to sustain.
Why pipeline leakage is so hard to find without a structured audit
Most revenue leaders know something's wrong before they can prove it. Close rates are slipping. Deals that looked solid three weeks ago are now "waiting on legal." Your forecast keeps coming in 15-20% light, but nobody on the team can point to a single reason.
This is pipeline leakage, and it's expensive in ways that don't always show up on a dashboard. According to Gartner's B2B sales research, the average B2B deal involves 6-10 stakeholders and spans several months. Each additional touchpoint is another chance for a deal to go cold, get deprioritized, or simply fall through a process gap that nobody's documented.
The frustrating part is that most teams react to symptoms instead of causes. They add more pipeline at the top. They pressure reps to close faster. They restructure comp. None of it works because the problem isn't effort or attitude. It's a specific breakdown at a specific stage that nobody's measured.
A sales process audit fixes that. It's a structured diagnostic that tells you exactly where deals are dying, why, and what to change first. You don't need a consultant to run one, and you don't need six months. A focused team can complete it in a week.
If you've been trying to improve forecast accuracy without fixing the underlying process, the sales maturity model framework is worth reading alongside this audit — it gives you the broader context for where your process gaps fit in your organization's development.
What a sales process audit actually examines
A sales process audit is a structured review of your current pipeline stages, conversion metrics, deal velocity, and the quality of your CRM data. It isn't a rep performance review and it isn't a strategy session. It's diagnostic work.
The audit answers five specific questions:
- What percentage of deals convert from each stage to the next?
- How long do deals spend at each stage, broken down by segment?
- Where do the majority of lost deals die, and why?
- What do won deals have in common that lost deals don't?
- How reliable is your pipeline data as a forecasting input?
Those five questions map to the five steps of the audit process. You work through each one in sequence. The output isn't a presentation. It's a ranked list of constraints with enough evidence to decide which one to fix first.
Who should run it
The audit works best when a RevOps lead or a sales leader drives it, with input from a senior rep and someone from finance or CS who touches the post-sale process. You want people who know where the bodies are buried, not just what the CRM says.
Avoid running it as a committee. Two to three people working with the data for a week will find more than six people talking about the data for a month.
What the audit won't tell you
It won't tell you whether your product is competitive or whether your pricing is wrong. Those are separate questions. A sales process audit assumes your core offer has market fit. If it doesn't, fixing process won't help.
What it will tell you is whether your process is letting good-fit deals slip through cracks that shouldn't exist.
Audit scope: keep it tight
Focus the audit on closed deals and active pipeline from the last 90-180 days. Going further back introduces noise from team changes, product updates, and market shifts that don't reflect current conditions. If you're in enterprise with longer cycles, extend to 12 months but weight the last 6 more heavily.
Step 1: pull the right data before you touch anything else
Most audit attempts fail in the first hour because people pull the wrong data. They export their full CRM, stare at 4,000 rows of opportunities, and don't know where to start.
Pull four specific datasets:
Closed-won and closed-lost deals, last 90-180 days. For each deal you need: entry date per stage, exit date per stage, deal size, segment (SMB/mid-market/enterprise), first-touch source, sales rep, close reason. That's it. Nothing else matters at this step.
Active pipeline by stage. Current deals with their entry dates for each stage they've passed through. This tells you where deals are sitting today vs. historical averages.
CRM field completion rate. What percentage of opportunities have required fields filled in for each stage? This is your data quality baseline.
Rep-level conversion rates. Not to call anyone out, but to find outliers in both directions. A rep converting at 3x the team average at a specific stage knows something the others don't.
Common data problems you'll hit
Deals get manually moved backward in the pipeline when a rep wants to "reset" the timeline. Stages get skipped entirely. Close dates get pushed without any note or reason. These aren't minor annoyances. They're signals that your process governance is weak.
When you find a deal that jumped from demo to proposal without any record of a discovery call, that's not just messy CRM data. That's a rep who either skipped discovery or didn't document it. Either way, you need to know how common it is before you can assess the risk it creates.
Don't clean the data before you audit it. The mess is the information.
Step 2: stage conversion analysis and what normal looks like
Stage conversion analysis is the core of any sales process audit. You're measuring the percentage of deals that advance from each stage to the next, then comparing that against benchmarks and against your own historical patterns.
What to calculate
For each stage transition (e.g., Qualified Lead to Discovery, Discovery to Demo, Demo to Proposal, Proposal to Negotiation, Negotiation to Closed Won), calculate:
- Volume: how many deals entered the stage over the period?
- Conversion rate: what percentage advanced to the next stage?
- Average time in stage: how many days did advancing deals spend there?
- Average time for deals that didn't advance: how long before they stalled or closed-lost?
The gap between advancing deal velocity and stalling deal velocity is your first clue. If deals that convert spend 5 days in discovery but deals that don't spend 22 days, something happens around day 6 that breaks deals.
Benchmark conversion rates by stage
These aren't universal — they vary by deal size, segment, and industry. But they're reasonable starting points for B2B SaaS and professional services. Forrester's B2B sales benchmark research consistently shows that teams with defined stage criteria and documented exit conditions outperform those without by 15-20 percentage points on overall Win Rate.
Use the table below as a starting reference, not a ceiling. If you're consistently beating these, investigate why before assuming your process is fine. Unusually high early-stage conversion often means qualification standards are too loose, which inflates pipeline but kills close rates.
The most important number isn't your overall Win Rate. It's the stage where you lose the most deals relative to the benchmark. That's your constraint.
| Pipeline stage transition | Typical B2B SaaS conversion | Enterprise deal conversion | Warning threshold (below) |
|---|---|---|---|
| MQL to SQL | 20-30% | 10-15% | <15% (loose qualification) |
| SQL to Discovery/Qualified | 60-75% | 50-65% | <50% (sourcing or routing issue) |
| Discovery to Demo/Evaluation | 70-85% | 55-70% | <55% (discovery not connecting need) |
| Demo to Proposal | 50-65% | 40-55% | <40% (ICP mismatch or poor demo) |
| Proposal to Negotiation | 45-60% | 35-50% | <35% (proposal or champion gap) |
| Negotiation to Closed Won | 65-80% | 55-70% | <55% (legal, budget, or urgency failure) |
| Overall pipeline Win Rate | 20-30% | 12-20% | <15% (structural process problem) |
The table is a starting point, not a report card
If your overall Win Rate is 12% but your average contract value is $250K+, that may be perfectly healthy. Context matters. The comparison that matters most is your own conversion trend over 3-6 quarters. If a stage that used to convert at 65% now converts at 42%, that's a signal worth investigating regardless of what any benchmark says.
Step 3: cycle time by segment, not just the average
Average deal cycle time is one of the most misleading metrics in B2B sales. A single enterprise deal that closes after 14 months can make your mid-market average look three times longer than it actually is.
Break cycle time analysis by segment first. Then by deal source. Then by rep. You're looking for three things:
Stage-level velocity gaps. Which stage consistently takes longer than it should for deals that eventually close? If proposal-to-negotiation takes 30+ days for your mid-market segment when the historical average is 12, something changed. Maybe a competitor is getting in late, maybe your proposal quality dropped, maybe legal bottlenecks moved upstream.
Segment-source interaction. Outbound-sourced deals often move faster through early stages but stall at procurement. Inbound deals from your website close faster when they reach the right person, but that's a big "when." Knowing which source-segment combinations produce the fastest and highest-converting deals tells you where to focus pipeline generation efforts.
Rep-level cycle time by stage. If one rep's deals sit in "Demo Scheduled" for 20 days on average while the team average is 8, that's not a volume problem. That's a specific rep behavior at a specific stage. Worth a 30-minute conversation, not a new comp plan.
The stale deal problem
Deals that have been sitting in the same stage for 2x the historical average rarely close at anything approaching normal rates. Most teams let these inflate their pipeline without questioning them.
As part of your cycle time analysis, flag every deal that has exceeded the stage velocity benchmark by 2x or more. That's your zombie pipeline. You'll need to make a decision about each one: pursue with a clear next step, or close-lose it and clean the forecast.
Worth noting: decisions about deal selection and qualification discipline — which directly affect cycle time downstream — are covered in depth in the strategic sales focus framework.
Step 4: win/loss interviews that actually tell you something
Most win/loss programs are theater. You ask the rep why a deal was lost. They say "price" or "went with a competitor." You log it and move on. Nothing changes.
There are two problems with that approach. First, reps rarely know the real reason a deal was lost. Second, even when they do, they have incentives to attribute losses to factors outside their control.
Useful win/loss analysis requires talking to the buyer, not just the rep. Even a 15-minute conversation with a lost buyer tells you more than 10 rep debriefs.
Questions that get real answers
For lost deals, ask buyers:
- What was the problem you were trying to solve when you started this evaluation?
- At what point did you decide we weren't the right fit?
- What would have needed to be different for us to win the deal?
- Who else was involved in the final decision, and what were their main concerns?
For won deals, ask buyers:
- What made you confident enough to move forward?
- What almost stopped you from buying?
- Who in your organization had the strongest concerns, and what resolved them?
The answer to "what almost stopped you" in won deals is as valuable as any lost deal analysis. It tells you what your sales process is successfully handling and what nearly broke it.
What to do with close reasons from the CRM
CRM close reasons are directional at best. They tell you the story the rep chose to tell. Look at them for volume patterns, not diagnostic truth. If "budget" appears in 40% of lost deals, you might have a pricing conversation problem or a champion-building problem, not actually a budget problem. Win/loss interviews will tell you which.
For a deeper look at what happens specifically when deals fall apart at the proposal stage, the sales proposal process optimization guide covers the mechanics of fixing that specific gap.
Step 5: CRM hygiene check and what dirty data hides
CRM hygiene is the least glamorous part of a sales process audit and the part most teams skip. That's a mistake, because bad CRM data doesn't just make your reports wrong. It actively hides the problems you're trying to find.
Five things to check
Required field completion. What percentage of active deals have all stage-required fields filled in? For any field that drives your qualification process (budget confirmed, decision-maker identified, timeline documented), low completion rates mean your qualification is happening in salespeople's heads, not in your system. You can't manage what you can't see.
Stage date accuracy. Were stage transitions logged in real time or backdated? If most of your deals moved through three stages on the same day, someone's batch-updating the CRM at the end of the month. That destroys cycle time data.
Duplicate and stale opportunities. How many active opportunities haven't had any activity logged in 30+ days? This is your zombie pipeline. In most teams it's 20-40% of active pipeline, which means your coverage ratio is much worse than your CRM shows.
Close date discipline. How often do close dates get pushed? If the average deal has its close date extended 3+ times, your forecast is a fiction. Reps learn to set dates that satisfy the CRM without reflecting real buyer intent.
Lost deal documentation. For closed-lost deals, how many have an actual reason logged vs. just a status change? If 60%+ have no documented reason, you can't learn from losses at scale.
What dirty data actually costs you
A team with 35% CRM data completeness on key fields is effectively flying without instruments. Every pipeline review, every forecast call, every board presentation is built on a guess dressed up as a number. When leadership makes headcount or channel investment decisions based on that data, the error compounds.
CRM hygiene isn't a RevOps project. It's a sales leadership accountability issue. If the VP of Sales doesn't demand clean data, it won't happen.
The three most common failure patterns and what they mean
After running sales process audits across B2B teams, three patterns show up repeatedly. Recognizing them early saves weeks of analysis.
Pattern 1: deals die at proposal, which means discovery failed
If your highest drop-off stage is Proposal to Closed, the temptation is to fix the proposal. Better templates, faster turnaround, more competitive pricing. That's usually the wrong fix.
Deals die at proposal when discovery didn't surface what actually matters to the buyer. If the rep doesn't understand the business pain, the economic case, and the internal politics before writing a proposal, no template saves it. The proposal either addresses the wrong problem or goes to the wrong person.
Fix: audit your discovery calls. Are reps asking about business outcomes or just confirming feature fit? Are they identifying the economic buyer or just working with whoever picked up the phone?
Pattern 2: deals stall at legal, which means there's no internal champion
Deals that reach legal and then sit there for 45-90 days usually don't have a champion. They have a contact. A champion is someone who is actively pushing the deal internally, navigating procurement, accelerating legal review, and removing blockers. A contact is someone who answers your emails.
Without a champion, legal review moves at the pace of the organization's default bureaucratic process, which is slow. The rep can't accelerate it from outside. Only the buyer can.
Fix: qualify for champion earlier. Before you get to proposal, you should know who in the buyer's organization has a personal stake in the outcome of this purchase. If you can't name that person, the deal is at risk.
Pattern 3: ghosting after demo, which means ICP mismatch
When buyers go silent after a strong demo, it's tempting to blame the demo itself. Tighten the pitch, add more social proof, improve the follow-up sequence. That's often the wrong lever.
Post-demo ghosting at scale usually means you're running demos for companies that shouldn't be in your pipeline. They were interested enough to take a call but not interested enough to justify the purchase. The product fit isn't there, or the timing isn't there, or the buying process doesn't exist yet.
Check: what's the job title, company size, and growth stage of your post-demo ghosts vs. your post-demo conversions? You'll usually find a clear ICP boundary. Deals inside that boundary convert. Deals outside it go dark. Tighten your qualification criteria before demo, not after.
If this pattern is showing up in your pipeline, the strategic sales focus framework can help you define where you should and shouldn't be competing.
One pattern, one fix at a time
When you run an audit and find two or three failure patterns, it's tempting to fix them all at once. That's how you get a six-month initiative with no measurable outcome. Pick the stage where the most deal value is dying, fix the process there, and measure the impact over 60-90 days before moving to the next constraint.
How to identify your single biggest constraint from audit data
Every pipeline has one constraint that matters more than everything else. Not two constraints, not a set of overlapping issues. One primary bottleneck that, if fixed, improves everything downstream. The goal of your audit is to find it.
Here's a practical method that works consistently:
Calculate deal value lost per stage. Take every deal that entered a stage and didn't advance. Multiply the count by your average deal size. That's the approximate revenue value bleeding out at that transition.
For example: 40 deals entered the Proposal stage. 22 didn't advance. Average ACV is $45K. You lost approximately $990K in annual contract value at that one transition over 90 days. That's your priority number.
Rank all stage transitions by revenue lost. The one at the top of that list is your constraint. Not the one that feels most painful, not the one that's easiest to fix. The one that's costing you the most.
Cross-reference with cycle time. If the highest-revenue-loss stage also has 2x the expected dwell time, that confirms it. Long dwell time plus high drop rate equals your biggest constraint.
Check rep variance. If one or two reps are outperforming at the constraint stage, they have a method that's working. Replicating it is faster than inventing a new one.
This is fundamentally a Theory of Constraints approach applied to pipeline. Fix the bottleneck, and the whole system improves. Skip the bottleneck to fix something else, and you've done a lot of work for minimal revenue impact.
For teams that need an external perspective on identifying and prioritizing these constraints, CRO advisory engagements often start with exactly this kind of diagnostic work.
Need help running a sales process audit?
If your pipeline has consistent conversion problems and you'd rather run this with an experienced operator than figure it out solo, that's what advisory engagements are built for. The audit is usually the first deliverable.
Explore advisory optionsOne-week sales process audit: a day-by-day timeline
Here's a realistic schedule for a team that can dedicate focused time to this work. You won't need external resources and you won't need to pause selling activity.
Day 1 (Monday): data pull and prep
Export closed-won, closed-lost, and active pipeline data for the last 90-180 days. Pull CRM field completion rates. Assign one person to build the stage conversion spreadsheet and one to start the cycle time breakdown. Don't analyze yet. Just get the data into a usable format.
Day 2 (Tuesday): stage conversion analysis
Calculate conversion rates for every stage transition. Compare to your historical baseline (prior quarter or prior year). Flag any stage where current conversion is more than 10 percentage points below historical. That's your watchlist.
Also calculate deal value lost per stage using the method above. Rank all transitions by revenue loss. You'll have your primary constraint candidate by end of day.
Day 3 (Wednesday): cycle time and rep variance analysis
Break cycle time by segment and by rep. Identify stages where dwell time has increased vs. the historical baseline. Flag all deals that are 2x over the stage velocity benchmark. These are your zombie deals. Decide by end of day which ones get a defined next step vs. which ones get closed-lost.
Day 4 (Thursday): win/loss interviews
Schedule and run 4-6 buyer interviews. Two lost deals, two won deals, one deal that stalled and restarted, one deal where the expected close date moved 3+ times. Twenty minutes per call is enough if you use the questions from Step 4. Document the pattern themes, not every answer.
Day 5 (Friday): synthesis and constraint identification
Review all five data sets together. Write down your constraint hypothesis in one sentence: "The single biggest driver of pipeline loss is [stage transition] because [root cause], which costs approximately [$X] per quarter." Share it with your CRO, VP Sales, or RevOps lead. Get one challenge. Update if needed. That's your audit output.
The whole process takes roughly 20-30 hours of focused work spread across a team of two to three people. The output is a specific, evidence-backed diagnosis. Not a list of things to work on. One thing to fix first.
What to fix first after the audit is done
The audit gives you a diagnosis. Now you need a decision about treatment. There are three types of fixes, and they require different approaches.
Process fixes (fastest to implement)
If deals are dying because a step isn't happening consistently — discovery calls being skipped, proposals going out without documented business cases, next steps not being scheduled before demos end — you have a process compliance problem. The fix is clearer criteria, manager reinforcement, and short feedback loops.
Process fixes typically show results within 30-60 days because you're changing rep behavior, not rebuilding systems.
Capability fixes (medium timeline)
If deals are dying because reps don't know how to have a specific conversation — negotiating against procurement, building internal business cases, navigating multi-stakeholder deals — you have a skills gap. No process change fixes a skills gap. You need targeted coaching or enablement on the specific capability that's failing at the constraint stage.
Capability gaps take 60-90 days to address meaningfully. Expect some rep variance before the coaching takes hold.
Structural fixes (longest timeline)
If deals are dying because your ICP is too broad, your qualification criteria don't match your win profile, or your stage definitions encourage gaming instead of accurate reporting — that's a structural problem. These take 90-180 days to address because you're changing how the whole team thinks about the pipeline.
The most common structural problem found in audits is qualification standards that are either too loose (everyone gets a demo) or too rigid (only companies that match last year's wins get through). Both destroy conversion in different ways.
Whatever type of fix you're making, build a 90-day measurement plan before you start implementing. Define what success looks like at the constraint stage, how you'll measure it, and when you'll review. Audits without measurement plans tend to produce short-term behavior changes that revert within two quarters.
If what you find in the audit points to broader organizational readiness questions, the sales maturity model framework helps contextualize which fixes your team is actually ready to sustain.
Table of Content
