The distinction from a marketing qualified lead is ownership, not quality. An MQL is marketing's judgement that someone looks promising based on behaviour and firmographics; an SQL is sales agreeing after human review. The conversion rate between the two is the single most revealing number in a B2B funnel.
When that rate is low — and below roughly a quarter it usually indicates a real problem — the cause is almost always a definition mismatch rather than lazy sales follow-up. Marketing is optimising for a threshold that does not predict revenue, typically because the scoring model rewards content engagement rather than buying intent.
The definition has to be jointly owned and periodically revisited. In practice most companies write it once during a tooling implementation, never revalidate it against closed-won data, and continue reporting against a threshold that stopped predicting anything years ago.
Why SQL matters
It is the point where marketing's numbers become sales' numbers, so it is where most B2B reporting disputes originate. It is also the highest-leverage place to fix acquisition economics: raising the bar for an SQL usually reduces lead volume and increases revenue, because sales capacity stops being spent on prospects that were never going to buy.
Fewer leads, more revenue
A company generates 400 MQLs a month, of which 60 are accepted as SQLs — a 15% acceptance rate — and sales complains constantly about lead quality. Reviewing twelve months of closed-won deals shows almost none started from the ebook downloads driving most of the MQL volume. Redefining the threshold around pricing-page visits and demo-request behaviour cuts MQLs to 130 and raises SQLs to 71. Marketing's headline number fell by two thirds; pipeline went up, and so did the amount of selling time available per opportunity.
Benchmarks
- MQL to SQL acceptance — investigate the definition below this
- Under 25%
- Healthy range for most B2B funnels
- 30–50%
- Unusually high — the bar may be set too low
- Above 70%
Ranges drawn from Digital Squad client accounts and published industry data. Treat them as orientation, not targets — your category may differ substantially.
Common mistakes
Marketing defining the criteria alone
A threshold sales did not agree to will not be honoured, and the acceptance rate becomes a running argument rather than a measurement. The definition has to be co-signed or it is fiction.
Scoring engagement rather than intent
Opening emails and downloading guides indicates interest in content. Visiting pricing, comparing you against a named competitor, or returning repeatedly in a week indicates interest in buying. Only the second predicts revenue.
Never revalidating against closed-won data
The only honest test is whether SQLs become customers at a materially higher rate than non-SQLs. Most companies have never run it, and a meaningful share would fail.
Treating rejected leads as waste
A lead sales declines is usually early rather than bad. Without a nurture path back, the company pays to acquire the same prospect again in six months when they are actually ready.
Where we work on this