13/13 Gate✓ IQ Certified10/10?

What's the right conversion rate from SQL to closed-won at our stage?

📖 3,425 words🗓️ Published Jul 18, 2026 · Updated Jul 20, 2026

Direct Answer

There is no single "right" SQL-to-closed-won rate — the correct number is the one that lets you hit your revenue plan with a customer-acquisition-cost (CAC) payback period under ~18 months, and it is dictated far more by your deal size, sales-cycle length, lead-source mix, and growth stage than by any published benchmark. That said, if you need directional guardrails for B2B software, the commonly cited ranges are:

SMB / velocity motion (ACV under ~$25K, 30–45 day cycle, one decision-maker): roughly 18–30% SQL-to-close.
Mid-market (ACV ~$25–100K, 60–90 day cycle, 2–4 stakeholders): roughly 12–20%.
Enterprise (ACV over ~$100K, 4–9 month cycle, 5+ stakeholders): often 6–15%, and single digits is normal, not broken.

A blended B2B SaaS median tends to land in the 15–20% band, but the dispersion around that median is enormous, so a blended figure taken in isolation is almost useless. The actionable move is to stop chasing an external number and instead: (1) lock a hard SQL definition tied to a CRM stage gate, (2) cohort your SQLs by creation date and follow them for two full median sales cycles, (3) segment the resulting rate by ACV tier, lead source, and rep, and (4) judge it against your own trailing six quarters plus your unit economics. If your rate is stable, your CAC payback is under 18 months, and you're hitting plan, your rate is "right" regardless of what a benchmark deck says. If you're a Series A company below ~12% with an inbound-heavy funnel, or a scale-stage company whose rate has been flat for four quarters, you have a problem worth diagnosing — and the sections below show you exactly how.

flowchart TD A[SQLs created this cohort] --> B{Advance to Stage 2 / discovery complete} B -->|advance| C{Advance to Stage 3 / eval} B -->|no fit or no response| X[Closed Lost] C -->|advance| D{Advance to Stage 4 / proposal} C -->|stall| X D -->|advance| E{Negotiation and legal} D -->|deprioritized| X E -->|signed| W[Closed Won] E -->|slip or lost| X W --> R["SQL-to-close rate = Won / cohort SQLs"] X --> R

How to Calculate SQL-to-Closed-Won Correctly

Most teams that think they're "below benchmark" are actually just measuring wrong. Before you touch your sales process, fix your math. There are four decisions that determine whether your number is real.

1. Define the SQL as a completed event, not a hopeful one. An SQL should be the *first completed discovery meeting* where a qualified opportunity was created — not a booked meeting, not a no-show, not an email reply, not a form fill. Anchor it to a specific CRM stage with required qualification fields (a MEDDPICC or BANT-style checklist) that must be populated for the stage to advance. If half your "SQLs" are unqualified meetings that never had a real buying signal, your denominator is inflated and your conversion rate looks artificially terrible. Conversely, if SDRs can freely reject legitimate meetings before they count, your denominator is deflated and your rate is fiction. Write the definition down, put it in your CRM as a gate, and audit it quarterly.

2. Define closed-won as recognized, retained revenue. Count revenue that was actually booked in-period, weighted by ARR rather than logo count so a $200K deal and a $8K deal don't count equally in a blended rate. Strip deals that churn inside 90 days — an early cancel is a customer-success or qualification failure masquerading as a pipeline win. If a meaningful slice of your closed-wons cancel within the first quarter, your "conversion rate" is overstating durable revenue.

3. Cohort by SQL creation date and follow for two median cycles. This is the single most common metric malpractice in SaaS. If you divide *this quarter's* closed-wons by *this quarter's* SQLs, you are dividing deals that started months ago by leads that mostly haven't had time to convert yet — which can deflate the apparent conversion by 30–50%. Instead, take a cohort of SQLs created in a past window, then follow that same set of records forward for at least two times your median cycle length. If your enterprise median is ~130 days, you need roughly an eight-month observation window before the cohort's rate is trustworthy. Early-stage teams routinely underestimate this lag by 40–60% and panic over a number that will look far healthier once the cohort matures.

4. Match your comp model to the real number. If your true SQL-to-close is 8% but your quota-capacity model assumes 14%, your reps do not have enough pipeline to hit target and you will see an attainment crisis within two quarters — followed by regrettable attrition. Sales comp and pipeline-coverage assumptions should be built on the *measured*, cohorted, segmented rate, not on an aspirational benchmark you pasted from a report.

Get those four right and you'll often discover your "problem" rate was a measurement artifact. Only after the math is clean should you compare against benchmarks or start changing the sales motion.

Why Your Growth Stage Changes What "Good" Looks Like

The "right" rate shifts dramatically depending on whether you're pre-product-market-fit, in growth mode, or optimizing for efficiency — and benchmarking across stages is one of the fastest ways to draw the wrong conclusion.

Seed / Series A (ARR under ~$5M). You're still learning who buys and why. Sales is often founder-led, the process is manual, and you're — correctly — disqualifying aggressively to avoid burning limited runway on bad-fit deals. A 10–15% rate can be perfectly healthy here, and founder-led motions that hand-qualify every lead sometimes run much higher (20–40%) precisely because volume is tiny and selection is intense. The catch: at 10 SQLs a month, your sample is too small for statistical significance. A single deal swings the rate by ten points. At this stage, trend direction over 3–6 months matters more than any absolute number, and you should expect noise.

Series B (ARR ~$5–20M). As you introduce structured qualification, sales enablement, and a repeatable demo-to-close motion, the median typically climbs into the 15–20% band. This is also the stage where a low rate is genuinely dangerous rather than just noisy: burning cash on unqualified pipeline at this scale compounds quickly, and investors are watching efficiency. If you're Series B–D and sitting below ~12%, treat it as a red flag worth a full diagnostic — teams that let it stay there tend to consume capital materially faster than efficient peers.

Scale stage (ARR over ~$20M). Top-quartile teams often reach 22–28% in a velocity/SMB motion because they've built playbooks, tightened ICP, and installed CRM hygiene that prevents leaky funnels. But the same 9% that would alarm a Series A company can be *excellent* for a late-stage enterprise vendor with a $150K ACV and nine-month cycles. Your stage dictates your tolerance for lower conversion: earlier stages should expect more noise and lower absolute rates; later stages should demand more precision and steadier trends.

The trap is comparing yourself to a company at a different stage — or worse, to public-company disclosures. IPO'd companies and heavily funded SaaS benchmark samples are survivor-biased: they only capture businesses that made it. If you're bootstrapped or pre-PMF, expect to run 30–50% below those figures until your ICP locks, and don't treat that gap as failure.

Segment Dispersion: Why Blended Numbers Lie

A single blended SQL-to-close rate hides more than it reveals, because the underlying segments behave like completely different businesses.

Enterprise (roughly 6–15%). Long cycles (commonly ~4–9 months), five or more stakeholders, procurement, security review, and legal. A large share of enterprise deals slip at least one quarter before they close or die, so fallout is *baked into* the rate — it's not a sign of a broken process. Enterprise also tends to carry a longer CAC payback (frequently well past 18–24 months), which you must weigh alongside the lower conversion.

Mid-market (roughly 12–20%). Cycles of 60–90 days, two to four stakeholders, some but not overwhelming process. This is often the sweet spot for CAC payback (frequently in the low-to-mid teens of months), which is why a "lower" mid-market conversion rate can be economically superior to a "higher" SMB one.

SMB / velocity (roughly 18–30%). Short cycles (30–45 days), usually a single buyer, minimal process. Conversion looks great — but watch the CAC-payback gotcha: SMB customers often churn faster and carry lower ACV, so a shiny 26% rate can sit on top of a payback period that's *worse* than mid-market's. High conversion on cheap, leaky revenue is a treadmill, not a compounding engine.

Because these segments differ so much, a company drifting upmarket without adjusting its comp plan or stage definitions will watch its blended rate "collapse" and blame the sales team — when the real story is a mix shift. If you compute a 7% blended rate and believe you're a mid-market business, the more likely explanations are that your stage definitions are mislabeled or your ICP has quietly drifted enterprise. Always decompose the blend before you react to it.

The Hidden Variable: Lead-Source Mix

Not all SQLs are created equal, and your blended rate is heavily driven by *where* those SQLs came from. As a rough hierarchy of intent:

Inbound / product-qualified (demo requests, trial conversions, high-intent content): typically the strongest converters, often 20–30%+, because the prospect self-qualified and arrived with a problem in mind.
Partner / referral-sourced: often 25–35%, because trust is pre-established and the referrer has effectively done the qualification for you.
Outbound / cold SDR-sourced: usually the lowest, often 8–15%, because you're interrupting a prospect who wasn't shopping.

This means a blended number is meaningless without the mix behind it. Consider a team at a blended 12%. If the funnel is 70% outbound and 30% inbound, that 12% might reflect outbound running at a healthy ~10% and inbound at a strong ~22% — a good result. Flip the mix to 70% inbound and 30% outbound, and a blended 12% is alarming, because your highest-potential channel is clearly underperforming its ceiling.

Segment SQL-to-close by source for at least three quarters before you react to the blended figure. The most common mistake is lumping all SQLs together, concluding you're below benchmark, and overhauling the entire sales process — when the real issue is that a single source is dragging the average down. Fix that source with better targeting, tighter qualification criteria, or different messaging, rather than rebuilding everything. As a planning heuristic: inbound-heavy mixes should target ~18–22%, while outbound-heavy mixes can be perfectly healthy at ~10–14% as long as the unit economics hold. And when you diagnose a weak channel, look upstream first — a bad outbound rate is usually a targeting or list-quality problem, not a closing problem.

A Worked Example: Decomposing the Funnel Math

Benchmarks feel abstract until you run your own funnel through the stage math, so here's a concrete decomposition. Assume you pull a trailing six-quarter cohort of SQLs and track each stage-to-stage advance rate:

10,000 SQLs (cohort, followed 2x median cycle) -> 5,500 reach Stage 2 / discovery complete (55% advance) -> 2,750 reach Stage 3 / evaluation (50% advance) -> 1,650 reach Stage 4 / proposal (60% advance) -> 660 Closed-Won (40% advance) = 6.6% SQL-to-closed-won

Multiply the stage rates through — 0.55 × 0.50 × 0.60 × 0.40 = 6.6% — and you can see why an enterprise-heavy funnel "looks worse" than people expect. Even a *strong* 40% proposal-to-close rate decomposes to a mid-single-digit SQL-to-close once you chain it behind three earlier stage gates. Enterprise sellers who beat themselves up over a 7% figure are often running a completely healthy funnel; the single number just doesn't communicate that.

Now flip it. Suppose your stage rates are 70% / 65% / 70% / 55% — a tight velocity motion:

0.70 x 0.65 x 0.70 x 0.55 = 17.5% SQL-to-closed-won

Same four stages, radically different result, entirely driven by segment dynamics and cycle friction rather than seller quality.

The practical use of this decomposition is diagnosis by stage, not by aggregate. If your overall rate is disappointing, walk the chain and find the single worst-converting transition. A funnel that hemorrhages at Stage 2→3 (discovery to evaluation) is telling you qualification or discovery is weak — you're advancing meetings that were never real opportunities. A funnel that's healthy until Stage 4→close is telling you the problem is pricing, competition, or negotiation, not lead quality. Fixing the *right* stage yields far more than blanket "close more deals" pressure. And if your worked example lands at 6.6% but you're convinced you're a mid-market business, that's your signal that either your stage definitions are mislabeled or your ICP has drifted upmarket without the comp plan following.

The Bear Case: Five Reasons Your Number Might Be Lying

Before you celebrate a great rate — or panic over a bad one — pressure-test it against these five failure modes.

1. Your great number is fake. An SMB rate of 26% is suspicious if your SDRs can reject first-meetings before they count as SQLs. Reps quickly learn to protect their personal conversion stats by refusing marginal-but-legitimate meetings. Audit the SDR-to-AE handoff rejection rate: if it's above ~15%, your denominator is being gamed and your conversion rate is partly manufactured. Pull the rejection reasons — if "not ICP" exceeds ~40% of rejections, either your ICP documentation is stale or your reps are cherry-picking. Either way, the headline rate isn't real.

2. Stable equals stagnant. A rate that's dead flat across four or more quarters is not a sign of a well-oiled machine — it's a sign of zero funnel learning. Healthy orgs see SQL-to-close *drift upward* over time (on the order of a couple hundred basis points a year) as ICP tightens, messaging sharpens, and reps get coached. Perfect stability usually means nobody is running experiments.

3. Conversion masks a CAC problem. A 22% SMB rate sitting on a ~28-month CAC payback is economically *worse* than a 14% mid-market rate with a ~16-month payback. The first compounds cash-burn; the second compounds value. Never look at conversion in isolation from what the converted revenue actually costs to acquire and how long it retains.

4. Public-company benchmarks are survivor-biased. DEF 14A filings and funded-SaaS benchmark samples only capture companies that made it far enough to be measured. If you're bootstrapped or pre-PMF, expect materially lower rates until your ICP locks, and don't let a survivor-skewed median convince you you're failing.

5. The single ratio is a CFO trap. Finance loves SQL-to-close because it's one clean number — but optimizing it in isolation drives reps toward risk-aversion. To protect their personal conversion rate, sellers will *refuse* harder, higher-value, ICP-fit deals in favor of easy wins. Healthy orgs measure the triplet of conversion × ACV × cycle-time, never conversion alone, so that closing a big, slow, valuable deal isn't penalized against a small, fast, cheap one. The moment a rate becomes a target that people are comped or ranked on, it starts to distort the behavior underneath it (a textbook case of Goodhart's Law).

A Five-Step Diagnostic Playbook

When your rate looks wrong — too low, too high, or just unexplained — don't reach for the aggregate. Walk these five cuts in order, and the real story almost always surfaces.

Trend. Plot the cohorted rate quarter over quarter. A QoQ swing greater than ~10% relative usually signals an ICP shift, a market regime change, a new competitor, or a process change — investigate the cause before assuming it's rep performance.
By rep. A 12% median with a P90 of 22% isn't a conversion problem; it's a coaching-arbitrage opportunity potentially worth millions. If the spread between your best and worst reps exceeds ~20 percentage points, your issue is hiring, ramp, territory design, or enablement — not the funnel.
By source. Inbound and referral typically convert 2–3× cold outbound. If they don't in your data, your inbound qualification is broken and you're wasting your highest-intent leads. This is where mix-shift problems hide.
By segment / ACV tier. Decompose the blend into SMB, mid-market, and enterprise, and cross-reference each against its CAC payback. A "normal-looking" blended rate can conceal a structurally broken enterprise motion or an unprofitable SMB treadmill.
By cohort age. Plot conversion against SQL-creation month. If newer cohorts systematically underperform older ones (after controlling for maturity), your TAM may be saturating or competition may be tightening — a strategic signal, not an execution one.

Run this playbook before you change quotas, restructure the team, or rewrite the sales process. Nine times out of ten, the "bad rate" resolves into one specific, fixable cut — a single weak source, one under-ramped cohort of reps, or a mix shift nobody flagged — rather than a systemic failure requiring a full rebuild.

FAQ

Is a 17% SQL-to-close rate the right target for my company? It's a reasonable directional midpoint for blended B2B SaaS, but it is not a target you should adopt blindly. Enterprise teams (ACV over ~$100K) frequently run in the 6–15% range and are perfectly healthy there, while SMB/velocity teams can reach 18–30%. Set your target from your own cohorted, segmented trailing six quarters and your CAC-payback math — not from a single benchmark figure.

What should I do if our SQL-to-close rate is below 12%? First, confirm it's real: check your SQL definition, your cohort window (are you measuring immature cohorts?), and your SDR rejection rate. If the number survives that scrutiny and you're Series B or later, treat it as a genuine red flag — teams that sit below ~12% tend to consume capital noticeably faster than efficient peers. Focus on tightening qualification, fixing the single worst-converting stage, and correcting any lead-source or comp-plan mismatch, rather than overhauling everything at once.

How do we actually improve SQL-to-close conversion? Diagnose before you prescribe. Walk the five-cut playbook to isolate whether the problem is a stage, a source, a segment, or a rep. Common high-yield fixes: aligning marketing and sales on a shared, gated SQL definition; shortening speed-to-lead on inbound; giving reps a stronger discovery framework so weak opportunities are disqualified early rather than dragged to a loss; and correcting a mislabeled stage model. Focused changes on the right lever often produce a few points of lift over two quarters — blanket "close more" pressure rarely does.

Does conversion rate vary by sales-rep experience? Substantially. Fully ramped, top-quartile reps at your stage may close well above your median while newer or under-ramped reps sit far below it. That's why a blended rate can be misleading: a 15% average might be one rep at 30% and another at 5%. Segment by tenure and ramp status, set fair individual targets accordingly, and treat a wide rep spread as a coaching/hiring/territory issue rather than a funnel issue.

Should we include self-serve or product-led deals in this metric? Only if those deals pass through a formal, sales-touched SQL stage. For pure self-serve motions, a different metric — such as activation-to-paid or free-to-paid conversion — is far more relevant. Mixing sales-assisted and self-serve pipelines in the same SQL-to-close calculation distorts the benchmark and hides what's actually driving revenue in each motion.

How often should we recalculate our target conversion rate? Quarterly at a minimum, using a rolling six-quarter average to smooth out seasonal noise and small-sample swings. Your ideal rate moves as your product, pricing, ICP, and competitive market change, so a target set 18 months ago is probably stale. Compare against external benchmarks only as a directional sanity check — never as a hard rule that overrides your own unit economics.

Sources

The Bridge Group — SaaS AE and inside-sales metrics research: https://www.bridgegroupinc.com/research
HubSpot — sales pipeline, conversion, and benchmark content: https://blog.hubspot.com/sales
Salesforce — State of Sales research reports: https://www.salesforce.com/resources/research-reports/state-of-sales/
Gong Labs — revenue-intelligence and deal-conversion research: https://www.gong.io/resources/labs/
Gartner — B2B sales funnel, win-rate, and buying research: https://www.gartner.com/en/sales
Bessemer Venture Partners — State of the Cloud and SaaS efficiency benchmarks: https://www.bvp.com/atlas/state-of-the-cloud
Harvard Business Review — sales efficiency frameworks and case studies: https://hbr.org/topic/subject/sales

flowchart TD S[Rate looks wrong] --> T["Cut 1: Trend QoQ"] T --> R["Cut 2: By rep dispersion"] R --> SR["Cut 3: By lead source"] SR --> SEG["Cut 4: By segment and CAC"] SEG --> CO["Cut 5: By cohort age"] CO --> D{Root cause located?} D -->|Rep spread over 20 pts| FIX1["Coaching / ramp / territory"] D -->|Inbound underperforms| FIX2[Fix qualification upstream] D -->|Segment mix shifted| FIX3[Re-segment and reset comp] D -->|Newer cohorts weaker| FIX4["TAM / competitive strategy"] D -->|Nothing anomalous| OK[Rate is likely correct for stage]

Related on PULSE

[What is the MEDDPICC approach to inbound qualification, and how does it speed SQL→opportunity conversion?](/knowledge/q582)
[How should a 2027 sales org design the MQL to SQL handoff?](/knowledge/q12623)
[MQL vs SQL — what's the actual difference, and how should you define them?](/knowledge/q10824)
[How do you audit marketing-sourced pipeline quality and spot rotten SQL sources?](/knowledge/q583)
[What role does predictive AI play in forecasting closed-won deals when vendor consolidation collapses legacy data sources?](/knowledge/q16333)
[How Do I Score My AEs on More Than Just Closed-Won?](/knowledge/q16021)

Download:

## Direct Answer

There is no single "right" SQL-to-closed-won rate — the correct number is the one that lets you hit your revenue plan with a customer-acquisition-cost (CAC) payback period under ~18 months, and it is dictated far more by your **deal size, sales-cycle length, lead-source mix, and growth stage** than by any published benchmark. That said, if you need directional guardrails for B2B software, the commonly cited ranges are:



- **SMB / velocity motion** (ACV under ~$25K, 30–45 day cycle, one decision-maker): roughly **18–30%** SQL-to-close.
- **Mid-market** (ACV ~$25–100K, 60–90 day cycle, 2–4 stakeholders): roughly **12–20%**.
- **Enterprise** (ACV over ~$100K, 4–9 month cycle, 5+ stakeholders): often **6–15%**, and single digits is normal, not broken.

A blended B2B SaaS median tends to land in the **15–20%** band, but the dispersion around that median is enormous, so a blended figure taken in isolation is almost useless. The **actionable move** is to stop chasing an external number and instead: (1) lock a hard SQL definition tied to a CRM stage gate, (2) cohort your SQLs by creation date and follow them for two full median sales cycles, (3) segment the resulting rate by ACV tier, lead source, and rep, and (4) judge it against your own trailing six quarters plus your unit economics. If your rate is stable, your CAC payback is under 18 months, and you're hitting plan, your rate is "right" regardless of what a benchmark deck says. If you're a Series A company below ~12% with an inbound-heavy funnel, or a scale-stage company whose rate has been flat for four quarters, you have a problem worth diagnosing — and the sections below show you exactly how.

```mermaid
flowchart TD
    A[SQLs created this cohort] --> B{Advance to Stage 2 / discovery complete}
    B -->|advance| C{Advance to Stage 3 / eval}
    B -->|no fit or no response| X[Closed Lost]
    C -->|advance| D{Advance to Stage 4 / proposal}
    C -->|stall| X
    D -->|advance| E{Negotiation and legal}
    D -->|deprioritized| X
    E -->|signed| W[Closed Won]
    E -->|slip or lost| X
    W --> R[SQL-to-close rate = Won / cohort SQLs]
    X --> R
```

## How to Calculate SQL-to-Closed-Won Correctly

Most teams that think they're "below benchmark" are actually just measuring wrong. Before you touch your sales process, fix your math. There are four decisions that determine whether your number is real.

**1. Define the SQL as a completed event, not a hopeful one.** An SQL should be the *first completed discovery meeting* where a qualified opportunity was created — not a booked meeting, not a no-show, not an email reply, not a form fill. Anchor it to a specific CRM stage with required qualification fields (a MEDDPICC or BANT-style checklist) that must be populated for the stage to advance. If half your "SQLs" are unqualified meetings that never had a real buying signal, your denominator is inflated and your conversion rate looks artificially terrible. Conversely, if SDRs can freely reject legitimate meetings before they count, your denominator is deflated and your rate is fiction. Write the definition down, put it in your CRM as a gate, and audit it quarterly.

**2. Define closed-won as recognized, retained revenue.** Count revenue that was actually booked in-period, weighted by ARR rather than logo count so a $200K deal and a $8K deal don't count equally in a blended rate. Strip deals that churn inside 90 days — an early cancel is a customer-success or qualification failure masquerading as a pipeline win. If a meaningful slice of your closed-wons cancel within the first quarter, your "conversion rate" is overstating durable revenue.

**3. Cohort by SQL creation date and follow for two median cycles.** This is the single most common metric malpractice in SaaS. If you divide *this quarter's* closed-wons by *this quarter's* SQLs, you are dividing deals that started months ago by leads that mostly haven't had time to convert yet — which can deflate the apparent conversion by 30–50%. Instead, take a cohort of SQLs created in a past window, then follow that same set of records forward for **at least two times your median cycle length**. If your enterprise median is ~130 days, you need roughly an eight-month observation window before the cohort's rate is trustworthy. Early-stage teams routinely underestimate this lag by 40–60% and panic over a number that will look far healthier once the cohort matures.

**4. Match your comp model to the real number.** If your true SQL-to-close is 8% but your quota-capacity model assumes 14%, your reps do not have enough pipeline to hit target and you will see an attainment crisis within two quarters — followed by regrettable attrition. Sales comp and pipeline-coverage assumptions should be built on the *measured*, cohorted, segmented rate, not on an aspirational benchmark you pasted from a report.

Get those four right and you'll often discover your "problem" rate was a measurement artifact. Only after the math is clean should you compare against benchmarks or start changing the sales motion.

## Why Your Growth Stage Changes What "Good" Looks Like

The "right" rate shifts dramatically depending on whether you're pre-product-market-fit, in growth mode, or optimizing for efficiency — and benchmarking across stages is one of the fastest ways to draw the wrong conclusion.

**Seed / Series A (ARR under ~$5M).** You're still learning who buys and why. Sales is often founder-led, the process is manual, and you're — correctly — disqualifying aggressively to avoid burning limited runway on bad-fit deals. A 10–15% rate can be perfectly healthy here, and founder-led motions that hand-qualify every lead sometimes run much higher (20–40%) precisely because volume is tiny and selection is intense. The catch: at 10 SQLs a month, your sample is too small for statistical significance. A single deal swings the rate by ten points. **At this stage, trend direction over 3–6 months matters more than any absolute number**, and you should expect noise.

**Series B (ARR ~$5–20M).** As you introduce structured qualification, sales enablement, and a repeatable demo-to-close motion, the median typically climbs into the **15–20%** band. This is also the stage where a low rate is genuinely dangerous rather than just noisy: burning cash on unqualified pipeline at this scale compounds quickly, and investors are watching efficiency. If you're Series B–D and sitting below ~12%, treat it as a red flag worth a full diagnostic — teams that let it stay there tend to consume capital materially faster than efficient peers.

**Scale stage (ARR over ~$20M).** Top-quartile teams often reach **22–28%** in a velocity/SMB motion because they've built playbooks, tightened ICP, and installed CRM hygiene that prevents leaky funnels. But the same 9% that would alarm a Series A company can be *excellent* for a late-stage enterprise vendor with a $150K ACV and nine-month cycles. Your stage dictates your tolerance for lower conversion: earlier stages should expect more noise and lower absolute rates; later stages should demand more precision and steadier trends.

The trap is comparing yourself to a company at a different stage — or worse, to public-company disclosures. IPO'd companies and heavily funded SaaS benchmark samples are survivor-biased: they only capture businesses that made it. If you're bootstrapped or pre-PMF, expect to run 30–50% below those figures until your ICP locks, and don't treat that gap as failure.

## Segment Dispersion: Why Blended Numbers Lie

A single blended SQL-to-close rate hides more than it reveals, because the underlying segments behave like completely different businesses.

**Enterprise (roughly 6–15%).** Long cycles (commonly ~4–9 months), five or more stakeholders, procurement, security review, and legal. A large share of enterprise deals slip at least one quarter before they close or die, so fallout is *baked into* the rate — it's not a sign of a broken process. Enterprise also tends to carry a longer CAC payback (frequently well past 18–24 months), which you must weigh alongside the lower conversion.

**Mid-market (roughly 12–20%).** Cycles of 60–90 days, two to four stakeholders, some but not overwhelming process. This is often the sweet spot for CAC payback (frequently in the low-to-mid teens of months), which is why a "lower" mid-market conversion rate can be economically superior to a "higher" SMB one.

**SMB / velocity (roughly 18–30%).** Short cycles (30–45 days), usually a single buyer, minimal process. Conversion looks great — but watch the CAC-payback gotcha: SMB customers often churn faster and carry lower ACV, so a shiny 26% rate can sit on top of a payback period that's *worse* than mid-market's. High conversion on cheap, leaky revenue is a treadmill, not a compounding engine.

Because these segments differ so much, a company drifting upmarket without adjusting its comp plan or stage definitions will watch its blended rate "collapse" and blame the sales team — when the real story is a mix shift. If you compute a 7% blended rate and believe you're a mid-market business, the more likely explanations are that your stage definitions are mislabeled or your ICP has quietly drifted enterprise. Always decompose the blend before you react to it.

## The Hidden Variable: Lead-Source Mix

Not all SQLs are created equal, and your blended rate is heavily driven by *where* those SQLs came from. As a rough hierarchy of intent:

- **Inbound / product-qualified** (demo requests, trial conversions, high-intent content): typically the strongest converters, often **20–30%+**, because the prospect self-qualified and arrived with a problem in mind.
- **Partner / referral-sourced**: often **25–35%**, because trust is pre-established and the referrer has effectively done the qualification for you.
- **Outbound / cold SDR-sourced**: usually the lowest, often **8–15%**, because you're interrupting a prospect who wasn't shopping.

This means a blended number is meaningless without the mix behind it. Consider a team at a blended 12%. If the funnel is 70% outbound and 30% inbound, that 12% might reflect outbound running at a healthy ~10% and inbound at a strong ~22% — a good result. Flip the mix to 70% inbound and 30% outbound, and a blended 12% is alarming, because your highest-potential channel is clearly underperforming its ceiling.

**Segment SQL-to-close by source for at least three quarters before you react to the blended figure.** The most common mistake is lumping all SQLs together, concluding you're below benchmark, and overhauling the entire sales process — when the real issue is that a single source is dragging the average down. Fix that source with better targeting, tighter qualification criteria, or different messaging, rather than rebuilding everything. As a planning heuristic: **inbound-heavy mixes should target ~18–22%**, while **outbound-heavy mixes can be perfectly healthy at ~10–14%** as long as the unit economics hold. And when you diagnose a weak channel, look upstream first — a bad outbound rate is usually a targeting or list-quality problem, not a closing problem.

## A Worked Example: Decomposing the Funnel Math

Benchmarks feel abstract until you run your own funnel through the stage math, so here's a concrete decomposition. Assume you pull a trailing six-quarter cohort of SQLs and track each stage-to-stage advance rate:

```
10,000 SQLs (cohort, followed 2x median cycle)
  -> 5,500 reach Stage 2 / discovery complete   (55% advance)
  -> 2,750 reach Stage 3 / evaluation           (50% advance)
  -> 1,650 reach Stage 4 / proposal             (60% advance)
  ->   660 Closed-Won                            (40% advance)
  = 6.6% SQL-to-closed-won
```

Multiply the stage rates through — 0.55 × 0.50 × 0.60 × 0.40 = **6.6%** — and you can see why an enterprise-heavy funnel "looks worse" than people expect. Even a *strong* 40% proposal-to-close rate decomposes to a mid-single-digit SQL-to-close once you chain it behind three earlier stage gates. Enterprise sellers who beat themselves up over a 7% figure are often running a completely healthy funnel; the single number just doesn't communicate that.

Now flip it. Suppose your stage rates are 70% / 65% / 70% / 55% — a tight velocity motion:

```
0.70 x 0.65 x 0.70 x 0.55 = 17.5% SQL-to-closed-won
```

Same four stages, radically different result, entirely driven by segment dynamics and cycle friction rather than seller quality.

The practical use of this decomposition is **diagnosis by stage, not by aggregate**. If your overall rate is disappointing, walk the chain and find the single worst-converting transition. A funnel that hemorrhages at Stage 2→3 (discovery to evaluation) is telling you qualification or discovery is weak — you're advancing meetings that were never real opportunities. A funnel that's healthy until Stage 4→close is telling you the problem is pricing, competition, or negotiation, not lead quality. Fixing the *right* stage yields far more than blanket "close more deals" pressure. And if your worked example lands at 6.6% but you're convinced you're a mid-market business, that's your signal that either your stage definitions are mislabeled or your ICP has drifted upmarket without the comp plan following.

## The Bear Case: Five Reasons Your Number Might Be Lying

Before you celebrate a great rate — or panic over a bad one — pressure-test it against these five failure modes.

**1. Your great number is fake.** An SMB rate of 26% is suspicious if your SDRs can reject first-meetings before they count as SQLs. Reps quickly learn to protect their personal conversion stats by refusing marginal-but-legitimate meetings. Audit the SDR-to-AE handoff rejection rate: if it's above ~15%, your denominator is being gamed and your conversion rate is partly manufactured. Pull the rejection reasons — if "not ICP" exceeds ~40% of rejections, either your ICP documentation is stale or your reps are cherry-picking. Either way, the headline rate isn't real.

**2. Stable equals stagnant.** A rate that's dead flat across four or more quarters is not a sign of a well-oiled machine — it's a sign of zero funnel learning. Healthy orgs see SQL-to-close *drift upward* over time (on the order of a couple hundred basis points a year) as ICP tightens, messaging sharpens, and reps get coached. Perfect stability usually means nobody is running experiments.

**3. Conversion masks a CAC problem.** A 22% SMB rate sitting on a ~28-month CAC payback is economically *worse* than a 14% mid-market rate with a ~16-month payback. The first compounds cash-burn; the second compounds value. Never look at conversion in isolation from what the converted revenue actually costs to acquire and how long it retains.

**4. Public-company benchmarks are survivor-biased.** DEF 14A filings and funded-SaaS benchmark samples only capture companies that made it far enough to be measured. If you're bootstrapped or pre-PMF, expect materially lower rates until your ICP locks, and don't let a survivor-skewed median convince you you're failing.

**5. The single ratio is a CFO trap.** Finance loves SQL-to-close because it's one clean number — but optimizing it in isolation drives reps toward risk-aversion. To protect their personal conversion rate, sellers will *refuse* harder, higher-value, ICP-fit deals in favor of easy wins. Healthy orgs measure the **triplet of conversion × ACV × cycle-time**, never conversion alone, so that closing a big, slow, valuable deal isn't penalized against a small, fast, cheap one. The moment a rate becomes a target that people are comped or ranked on, it starts to distort the behavior underneath it (a textbook case of Goodhart's Law).

## A Five-Step Diagnostic Playbook

When your rate looks wrong — too low, too high, or just unexplained — don't reach for the aggregate. Walk these five cuts in order, and the real story almost always surfaces.

1. **Trend.** Plot the cohorted rate quarter over quarter. A QoQ swing greater than ~10% relative usually signals an ICP shift, a market regime change, a new competitor, or a process change — investigate the cause before assuming it's rep performance.
2. **By rep.** A 12% median with a P90 of 22% isn't a conversion problem; it's a coaching-arbitrage opportunity potentially worth millions. If the spread between your best and worst reps exceeds ~20 percentage points, your issue is hiring, ramp, territory design, or enablement — not the funnel.
3. **By source.** Inbound and referral typically convert 2–3× cold outbound. If they don't in your data, your inbound qualification is broken and you're wasting your highest-intent leads. This is where mix-shift problems hide.
4. **By segment / ACV tier.** Decompose the blend into SMB, mid-market, and enterprise, and cross-reference each against its CAC payback. A "normal-looking" blended rate can conceal a structurally broken enterprise motion or an unprofitable SMB treadmill.
5. **By cohort age.** Plot conversion against SQL-creation month. If newer cohorts systematically underperform older ones (after controlling for maturity), your TAM may be saturating or competition may be tightening — a strategic signal, not an execution one.

```mermaid
flowchart TD
    S[Rate looks wrong] --> T[Cut 1: Trend QoQ]
    T --> R[Cut 2: By rep dispersion]
    R --> SR[Cut 3: By lead source]
    SR --> SEG[Cut 4: By segment and CAC]
    SEG --> CO[Cut 5: By cohort age]
    CO --> D{Root cause located?}
    D -->|Rep spread over 20 pts| FIX1[Coaching / ramp / territory]
    D -->|Inbound underperforms| FIX2[Fix qualification upstream]
    D -->|Segment mix shifted| FIX3[Re-segment and reset comp]
    D -->|Newer cohorts weaker| FIX4[TAM / competitive strategy]
    D -->|Nothing anomalous| OK[Rate is likely correct for stage]
```

Run this playbook before you change quotas, restructure the team, or rewrite the sales process. Nine times out of ten, the "bad rate" resolves into one specific, fixable cut — a single weak source, one under-ramped cohort of reps, or a mix shift nobody flagged — rather than a systemic failure requiring a full rebuild.

## FAQ

**Is a 17% SQL-to-close rate the right target for my company?**
It's a reasonable directional midpoint for blended B2B SaaS, but it is not a target you should adopt blindly. Enterprise teams (ACV over ~$100K) frequently run in the 6–15% range and are perfectly healthy there, while SMB/velocity teams can reach 18–30%. Set your target from your own cohorted, segmented trailing six quarters and your CAC-payback math — not from a single benchmark figure.

**What should I do if our SQL-to-close rate is below 12%?**
First, confirm it's real: check your SQL definition, your cohort window (are you measuring immature cohorts?), and your SDR rejection rate. If the number survives that scrutiny and you're Series B or later, treat it as a genuine red flag — teams that sit below ~12% tend to consume capital noticeably faster than efficient peers. Focus on tightening qualification, fixing the single worst-converting stage, and correcting any lead-source or comp-plan mismatch, rather than overhauling everything at once.

**How do we actually improve SQL-to-close conversion?**
Diagnose before you prescribe. Walk the five-cut playbook to isolate whether the problem is a stage, a source, a segment, or a rep. Common high-yield fixes: aligning marketing and sales on a shared, gated SQL definition; shortening speed-to-lead on inbound; giving reps a stronger discovery framework so weak opportunities are disqualified early rather than dragged to a loss; and correcting a mislabeled stage model. Focused changes on the right lever often produce a few points of lift over two quarters — blanket "close more" pressure rarely does.

**Does conversion rate vary by sales-rep experience?**
Substantially. Fully ramped, top-quartile reps at your stage may close well above your median while newer or under-ramped reps sit far below it. That's why a blended rate can be misleading: a 15% average might be one rep at 30% and another at 5%. Segment by tenure and ramp status, set fair individual targets accordingly, and treat a wide rep spread as a coaching/hiring/territory issue rather than a funnel issue.

**Should we include self-serve or product-led deals in this metric?**
Only if those deals pass through a formal, sales-touched SQL stage. For pure self-serve motions, a different metric — such as activation-to-paid or free-to-paid conversion — is far more relevant. Mixing sales-assisted and self-serve pipelines in the same SQL-to-close calculation distorts the benchmark and hides what's actually driving revenue in each motion.

**How often should we recalculate our target conversion rate?**
Quarterly at a minimum, using a rolling six-quarter average to smooth out seasonal noise and small-sample swings. Your ideal rate moves as your product, pricing, ICP, and competitive market change, so a target set 18 months ago is probably stale. Compare against external benchmarks only as a directional sanity check — never as a hard rule that overrides your own unit economics.


## Sources

- The Bridge Group — SaaS AE and inside-sales metrics research: https://www.bridgegroupinc.com/research
- HubSpot — sales pipeline, conversion, and benchmark content: https://blog.hubspot.com/sales
- Salesforce — State of Sales research reports: https://www.salesforce.com/resources/research-reports/state-of-sales/
- Gong Labs — revenue-intelligence and deal-conversion research: https://www.gong.io/resources/labs/
- Gartner — B2B sales funnel, win-rate, and buying research: https://www.gartner.com/en/sales
- Bessemer Venture Partners — State of the Cloud and SaaS efficiency benchmarks: https://www.bvp.com/atlas/state-of-the-cloud
- Harvard Business Review — sales efficiency frameworks and case studies: https://hbr.org/topic/subject/sales

## Related on PULSE

- [What is the MEDDPICC approach to inbound qualification, and how does it speed SQL→opportunity conversion?](/knowledge/q582)
- [How should a 2027 sales org design the MQL to SQL handoff?](/knowledge/q12623)
- [MQL vs SQL — what's the actual difference, and how should you define them?](/knowledge/q10824)
- [How do you audit marketing-sourced pipeline quality and spot rotten SQL sources?](/knowledge/q583)
- [What role does predictive AI play in forecasting closed-won deals when vendor consolidation collapses legacy data sources?](/knowledge/q16333)
- [How Do I Score My AEs on More Than Just Closed-Won?](/knowledge/q16021)

Was this helpful?

Sources cited

joinpavilion.comhttps://www.joinpavilion.com/compensation-report bvp.comhttps://www.bvp.com/atlas/state-of-the-cloud-2026 bridgegroupinc.comhttps://www.bridgegroupinc.com/blog/sales-development-report gartner.comhttps://www.gartner.com/en/sales/research

Kory White