Here’s a number worth knowing about your own funnel: what percentage of your MQLs does sales actually accept and work? Not convert — just accept.
In healthy funnels, that number sits above 80%. In most funnels, it’s somewhere between 30% and 60%, and everyone has learned to live with it. Marketing keeps generating “qualified” leads, sales keeps ignoring a large share of them, and the gap gets explained away as lead quality — a budget problem, an audience problem, a content problem.
It’s usually none of those. It’s a definition problem, and it’s the single most common leak in the MQL to SQL handoff.
How definitions drift
No one sets out to have marketing and sales disagree about what a lead is. The definition was agreed — once. Then reality moved:
- The ICP tightened after a repositioning, but the scoring model still rewards the old profile.
- A form or event source was added that inflates scores without indicating intent.
- An admin adjusted thresholds to hit an MQL target, then left the company.
- Sales started informally requalifying leads with criteria that live only in reps’ heads.
Each change is small. Compounded over two or three years, marketing and sales end up describing genuinely different populations with the same three letters. Every dashboard that says “MQL” is now an argument waiting to happen.
The diagnostic: measure the disagreement
You don’t need a workshop to find out if you have this problem. You need three numbers from the last 90 days:
1. Sales acceptance rate — MQLs accepted (or worked within SLA) ÷ MQLs delivered. Under ~60%: definition drift, almost certainly.
1. Rejection reasons — if you capture them, cluster them. If “not a fit” dominates, the definition is wrong. If “no response / bad timing” dominates, the definition might be fine and the problem is speed and routing.
1. The interview gap — ask one marketer and one rep to describe a good lead. Write both answers down. The distance between them is your leak, in prose form. (In audits, this interview pair is often the finding.)
A typical case runs like this. Sales acceptance sits somewhere in the forties, and both sides have quietly decided the other is the problem. Write down each team’s definition of a good lead and the gap is usually a single criterion: marketing counts anyone who fits the profile and downloaded something; sales only wants the ones showing buying intent. Neither is wrong — they’ve just never re-agreed the line, so every MQL number describes two different things at once. Naming that one disputed criterion out loud is often the whole fix.
Rebuilding a definition both teams will sign
A definition that works has three properties most don’t:
It’s written in observable criteria. “Senior enough” is not observable; “director-level or above at a company with 50+ employees in these industries” is. Every criterion must map to a field or behaviour your systems can actually see — otherwise it can’t be automated, measured, or enforced.
It’s signed by both teams — literally. Put the definition in a one-page doc with the date and both leaders’ names. This sounds ceremonial. It isn’t. The signature is what lets you distinguish “the definition is wrong” from “the definition isn’t being honoured,” which are different problems with different fixes.
It has a review date. Definitions don’t drift because people are careless; they drift because nothing forces a re-look. A quarterly 30-minute review — acceptance rate, rejection reasons, proposed changes — keeps the definition tracking reality. Most quarters it takes ten minutes and changes nothing. The one quarter it matters, it saves you a year of quiet leakage.
Pair the definition with an SLA
A definition says what a qualified lead is. An SLA says what happens next: how fast sales touches it, how many attempts before recycling, what marketing owes back on rejected leads. The definition without the SLA just moves the argument downstream — I’ve written up the SLA structure that survives contact with reality separately.
What this fixes downstream
Get the definition right and several “unrelated” problems shrink at once: lead scoring starts predicting acceptance instead of gaming it, routing rules stop fighting over ambiguous records, and — most valuably — the MQL line on the board deck becomes a number both leaders will defend in the same meeting. That’s what the handoff is supposed to produce: not more leads, but one shared version of the truth about them.
—
Want the definition, scoring, and SLA looked at together, with evidence? That’s lenses two and three of the Pipeline Leak Audit. Or start with a free 30-minute health check.