What's a good win rate by deal stage in B2B SaaS?
Direct Answer
Healthy B2B SaaS funnels convert roughly 70% from stage 1 to 2, 55% from 2 to 3, 45% from 3 to 4, 65% from 4 to 5, and 75% from late stage to closed-won. That compounds to about 25-35% inbound lead-to-close and 15-25% outbound, per Gong Labs 2024 and ICONIQ Operating Metrics. SMB deals close faster but win less (~22%), mid-market lands around 27%, and enterprise hovers at 18-20% with longer cycles.
The number you should care about is not the total — it is which gate is silently bleeding.
TL;DR
- Stage-by-stage benchmarks for healthy B2B SaaS: 70% / 55% / 45% / 65% / 75% transition rates, ending at 25-35% inbound win.
- ACV bands matter more than the average. SMB ($10K), MM ($50K), and ENT ($250K+) have structurally different win profiles.
- A stage-1-to-2 below 50% means your MQL definition is too loose — fix qualification, not selling.
- A stage-5-to-close below 60% means AEs are commiting deals too early — fix the late-stage exit criteria.
- The most common fraud is the stage-gate trap: deleting losses or kicking them backward to inflate the number. Detect it by tracking total opp count, not just win rate.
Benchmarks by Stage and ACV Tier
A win rate without an ACV band attached is roughly useless. A 30% win rate is a hero number for enterprise and a red flag for transactional SMB. The shape of the funnel changes too — SMB tends to lose at the top (volume in, fast disqualify) while enterprise loses in the middle (long evaluations that die in legal or procurement).
The Bridge Group SDR Metrics Report and ICONIQ Growth's Operating Metrics deck both show this pattern holding across three consecutive years.
| Stage transition | SMB ~$10K ACV | Mid-Market ~$50K ACV | Enterprise ~$250K ACV |
|---|---|---|---|
| Stage 1 to 2 (MQL to SQL) | 65% | 70% | 60% |
| Stage 2 to 3 (SQL to Demo) | 50% | 55% | 55% |
| Stage 3 to 4 (Demo to Proposal) | 40% | 45% | 50% |
| Stage 4 to 5 (Proposal to Negotiation) | 70% | 65% | 55% |
| Stage 5 to Closed Won | 80% | 75% | 60% |
| Compounded lead-to-close | ~22% | ~27% | ~18% |
| Median sales cycle | 14 days | 84 days | 184 days |
Inbound runs hotter than outbound everywhere on the table. Gong Labs' 2024 win rate study put inbound at 25-35% and outbound at 15-25% across roughly 600K opportunities. The gap is widest at Stage 1 to 2 (because inbound MQLs self-select for problem awareness) and narrowest at Stage 5 to Close (because once a deal is in negotiation, lead source no longer matters — pricing and procurement do).
If your own numbers are far off these bands and you cannot point to a specific structural reason — vertical, motion, sales cycle, deal size — the gap is almost certainly process, not market.
One more nuance worth surfacing: pipeline coverage and win rate are inversely correlated in healthy orgs. Teams that need 4x or 5x coverage to hit number are usually running win rates in the high teens. Teams sitting at 30-35% can plan with 2.5-3x coverage.
When you see a team carrying 6x+ coverage and still missing, the win rate is being inflated somewhere upstream — usually by leaving dead opps open instead of marking them lost. Pavilion's 2024 RevOps benchmark report flagged this exact pattern as the single most common pipeline hygiene failure.
How to Diagnose a Broken Stage
Each stage has a signature failure mode and a specific fix. Treating "low win rate" as one problem and throwing generic training at it is how revenue teams burn quarters.
If Stage 1 to 2 conversion is below 50%, your MQL definition is too loose. Marketing is passing content downloaders and tire-kickers as qualified leads. Fix: tighten MQL to require explicit buying intent — meeting requested, pricing page hit twice in 30 days, or a fit-score floor.
Salesforce State of Sales 2024 found median MQL-to-SQL at 13% across all B2B and 31% for teams with a documented MQL definition; the delta is the cost of vague qualification.
If Stage 2 to 3 conversion is below 45%, your discovery is failing. Reps are advancing deals without a confirmed pain or budget signal. Fix: require a written "why now" and an identified economic buyer to advance from SQL to demo. MEDDIC users hit ~60% at this gate; everyone else lives in the 40s.
If Stage 3 to 4 conversion is below 35%, your demo is not landing on what the buyer actually cares about. The signal: high demo attendance, no follow-up engagement. Fix: switch from feature demos to outcome demos, and require a confirmed next step before leaving the call.
If Stage 4 to 5 conversion is below 55%, you have a champion problem. Proposals go out, then silence. The deal needed a stronger internal seller and never had one. Fix: gate proposals behind a verified champion call with at least one other stakeholder present.
If Stage 5 to Close is below 60%, your AEs are commiting deals too early. They are pulling things into "negotiation" because the forecast cycle pressures them to, not because the buyer is actually negotiating. Fix: redefine Stage 5 to require either a redlined order form or a verbal verbal — and audit any deal that sits in Stage 5 longer than 1.5x the historical Stage 5 dwell time.
The Stage-Gate Trap
The fastest way to "improve" a win rate is to lie about it, and most teams do this without realizing they are doing it. Three patterns to watch for.
Pattern 1: deleting losses. A rep marks a deal as "no decision" or "duplicate" instead of "closed-lost." The opp disappears from the win rate denominator. Detection: track total opportunities created per period and the disposition mix. If closed-lost as a share of all-closed is trending down while win rate is trending up, somebody is hiding bodies.
Gong's 2024 study found teams with healthy hygiene close-lost roughly 1.8 opps for every 1 close-won; teams gaming the metric show ratios under 1.0.
Pattern 2: backward stage movement. A stage-4 deal stalls, the rep moves it back to stage 2 to "reset," and now the stage-4-to-5 conversion looks artificially clean. Detection: a "regression report" — count of opps that moved to a lower stage within a quarter. Healthy regression rate is under 8% of total stage transitions.
Above 15% means the funnel is being laundered.
Pattern 3: never-aging open pipeline. Deals sit in "negotiation" forever because closing them as lost would tank the AE's number. Detection: dwell time per stage versus the 75th percentile of historical dwell. Anything past 2x dwell is dead — force it to a disposition.
Bessemer's 2024 State of the Cloud Report called this "phantom pipeline" and estimated it inflates reported win rates by 4-7 percentage points across the median SaaS org.
The real fix is cultural, not technical: reward accurate forecasting more than optimistic forecasting. Teams that grade reps on forecast accuracy and inspect stage-gate criteria weekly run cleaner funnels and — almost paradoxically — actually win more deals, because the time saved on dead pipeline gets reinvested in live opportunities.
Frequently Asked Questions
Q: Is a 50% win rate good? A: Only if you are counting from a clean qualified stage (Stage 3+) and your ACV is enterprise. From top-of-funnel, 50% is either a small sample size or a hygiene problem. Top-of-funnel win rates above 40% almost always mean opps are being created only after the deal is already qualified — which makes the metric meaningless.
Q: What is the single best leading indicator of win rate? A: Multithreading. ICONIQ's data shows deals with 4+ stakeholder contacts close at ~2.3x the rate of single-threaded deals. It outperforms demo count, email volume, and even discovery quality scores.
Q: How often should I recalibrate stage-conversion benchmarks? A: Quarterly at minimum, with a full audit every fiscal year. Markets shift, ICP shifts, and a benchmark from 18 months ago will quietly mislead your forecast.
Sources
- Gong Labs, "Win Rate Benchmarks: Analysis of 600K+ B2B Opportunities," 2024.
- ICONIQ Growth, "Operating Metrics for Cloud Companies: 2024 Edition."
- Bessemer Venture Partners, "State of the Cloud 2024."
- Pavilion, "2024 RevOps Benchmarks Report."
- Salesforce, "State of Sales, 6th Edition," 2024.
- Bridge Group, "SaaS AE Metrics and Compensation Report," 2024.
- Bridge Group, "SDR Metrics and Compensation Report," 2024.
- SaaStr, "B2B SaaS Sales Benchmarks by ACV Tier," 2024.