Nothing. An MQL is a prediction that sales will act, so a lead nobody calls was never qualified, it was labelled. The label and the action are different things and only one of them is worth anything.
Across the 292 posts on this site before today, the term appeared in exactly zero titles, which is roughly how much attention the handoff gets relative to how much damage it does.
The definition problem, stated plainly
There is no standard for what an MQL is. It is whatever the two teams agreed, and the agreement is usually made by the team that is measured on the count.
That is the whole mechanism. If marketing's target is a number of MQLs, then when the number is short the threshold moves, because the threshold is the only variable marketing controls in the last week of a quarter. Nobody decides to lower the bar. The bar is simply the easiest thing to adjust, and the target is real.
Sales works the list for a quarter, finds the hit rate poor, and stops. Nobody announces this either. The leads continue to be generated, counted and reported, and separately they are ignored, and both of those facts can be true for years.
What an unworked lead actually costs
More than nothing, which is the part that makes it worth fixing rather than tolerating.
| Cost | Why |
|---|---|
| The acquisition spend | Already paid, whether or not anybody calls |
| The decay | Intent is perishable. The same lead next month is a weaker lead |
| The competitor's head start | They were researching several vendors. One of them replied |
| The reporting damage | Channel performance is judged on conversion, so unworked channels look bad and get cut |
| The relationship | Every unworked batch makes the next handoff conversation harder |
The fourth one is the quietly expensive one. If a channel's leads are never called, that channel converts at zero, and the next budget round kills it on the evidence. The decision looks data-driven and is based entirely on an internal process failure.
The two numbers that end the argument
Stop debating the definition. Measure two things and the argument resolves itself.
Time from lead created to first genuine contact attempt. Not an automated email. A human attempt. Measure the median, not the average, because the average is destroyed by the leads nobody ever touched.
Acceptance rate. What share of MQLs did sales work at all. Not convert, work.
Those two numbers tell you which team has the problem, and it is usually both:
- Low acceptance, slow response: sales does not believe the leads. Check the definition.
- High acceptance, slow response: capacity or routing. A process problem, usually fixable in a week.
- High acceptance, fast response, poor conversion: the definition is genuinely too loose, and marketing needs to know rather than being told the leads are bad.
- Low acceptance, fast response on the ones taken: sales is cherry-picking, which may be rational. Find out what they are picking on and put it in the model.
Response time is the one with the longest research record behind it: the odds of a meaningful conversation fall steeply across the first day. Most teams measure it nowhere, which means the most consequential variable in the handoff is invisible to both sides.
The structural fix
The definition should be set by the team that does not benefit from loosening it, and it should carry a shared number.
In practice that means marketing's target includes something downstream: MQLs accepted, or pipeline created. The moment marketing is measured on a number sales controls part of, the incentive to push volume through a soft threshold disappears, and the conversation changes from "these leads are bad" to "which ones were good". That is the subject of should marketing carry a pipeline number.
It also means the two teams need to be looking at one record rather than two systems and a weekly export. When the campaign, the lead, the first call attempt and the deal all sit on the same account in Empiraa Signal, "did anybody call them" is a field rather than a disagreement.
Where the MQL genuinely breaks
Everything above assumes the buyer eventually becomes a lead. Increasingly they do not: with zero-click search at 75.1% and most evaluation happening before contact, the buying committee can shortlist you without a single person filling in a form.
When that is the pattern, fixing the MQL handoff is tuning something that has stopped being the main road. The signal that matters is company-level rather than contact-level, which is a company read three pricing pages and nobody called them.
Both problems are worth fixing. Just know which one you have before you spend a quarter on it.
Related reading: marketing says 400 leads, sales says 12 are real, and should marketing carry a pipeline number.
Marketing qualified lead has a definition on this site. See Signal's marketing side, in the same system as the pipeline.
