What is an MQL? A working definition

An MQL (Marketing Qualified Lead) is a contact who shows enough fit and enough interest to be worth a salesperson’s time, and whom sales has not validated yet.

That sentence is the textbook answer. It is also loose enough that two RevOps teams can read it, nod, and run completely different funnels. One team counts every ebook download as an MQL. The other waits for a demo request from a company that matches its ideal profile. Same three words, two different pipelines, and a standing argument between marketing and sales about whether the leads are any good.

This post gives you a definition you can operate: the two signals that define an MQL, the third signal most teams forget, a 10-minute test for whether your definition holds, and the one metric to watch once it does.

Fit and interest, together

A working MQL definition rests on two signals, and a lead needs both.

Fit is whether the lead looks like someone you can sell to: the right role, company size, industry, and region. If most of your revenue comes from US e-commerce companies between 50 and 500 employees, then a VP of Marketing at a 200-person Shopify brand is a strong fit, and a freelance consultant in a market you do not serve is not, no matter how engaged they are.

Interest is whether the lead has done something that signals real buying consideration: a demo request, a trial signup, repeated pricing-page visits from a corporate domain, or a sales question on a form. Reading one blog post does not count.

Neither signal is a yes/no checkbox. Both are scales. A demo request counts for more than an email open. A perfect-fit account counts for more than a borderline one. That weighting is the whole point of lead scoring: a senior buyer at a target-size company who books a call should outrank a student who opened three emails. Our 4-dimension lead scoring framework breaks fit and interest into the specific inputs worth weighting.

What qualifies a lead is the intersection. High interest at a wrong-fit company is a researcher, not a buyer. A perfect-fit account with zero engagement is an ABM target, not an MQL. Only the overlap counts.

The signal most definitions miss: disqualifiers

Fit and interest tell you whom to pull in. They say nothing about whom to keep out, and that gap is where most MQL definitions quietly leak.

Some leads clear the fit and interest bar and still should not reach sales. Common disqualifiers:

• A free or generic email domain (gmail, outlook) on a “request a demo” form, which often means a student, a job seeker, or a competitor rather than a buyer.
• An unsubscribe. A lead who opts out of your emails and looks interested at the same time is a contradiction worth catching.
• A known competitor or existing customer filling out a top-of-funnel form.
• A role with no path to purchase in your motion, such as a candidate researching the company before an interview.

Mature scoring models give these signals negative weight, not zero. An unsubscribe might subtract more points than a pricing-page visit adds, which keeps a contradictory lead out of the MQL bucket instead of letting one strong action carry it over the line. If you have never written down your disqualifiers, that is the fastest place to start. We cover this failure mode in more depth in why HubSpot lead scoring stops working.

So the fuller picture: an MQL has fit, has interest, and carries no disqualifying signal that cancels them out.

The test for a real MQL definition

Here is how to tell whether your definition is real or just words in a doc. Hand it to people and see if they agree.

Take 10 contacts from your CRM. Give them to three people on your team along with your written MQL definition. Ask each person, independently, to mark every contact yes or no. Then compare the three columns.

Agree 80% of the time or better, and you have a definition tight enough to build a pipeline on. Split or contradictory answers mean the definition is too vague, and that vagueness is exactly what reaches sales later as “marketing’s leads are bad.” The argument is rarely about lead quality. It is about a definition nobody pinned down.

A definition passes the test when it answers the edge cases out loud. Does a demo request from a free email address qualify? Does a perfect-fit account that has only opened emails? Does a director at a target account who unsubscribed last week? If your three reviewers have to guess, write the answer down and run the test again.

What an MQL is not

A few contacts get labeled MQLs that are not:

• A subscriber. Someone who joined your newsletter is engaged with your brand, not necessarily considering a purchase.
• A high-engagement researcher. A junior employee who reads three of your guides is learning, not buying.
• A demo request from a wrong-fit company. A demo request is an interest signal, but fit still has to apply.
• A contact who fits the ICP but has never engaged. That is a target, not an MQL.
• A score with no reason attached. If your system says “92” but the rep cannot see why, you do not have an MQL. You have a number.

A tighter bar usually means more revenue, not less

When pipeline looks thin, the instinct is to loosen the MQL bar so more leads pass. It tends to backfire. A looser bar sends sales more leads, a smaller share of them convert, reps stop trusting the queue, and your good leads get the same diluted attention as the weak ones.

The opposite move works better. Picture a team that replaces a loose two-signal filter, email opens and clicks, with a weighted model that scores fit and behavior together and counts negative signals. The MQL count drops sharply, but conversions hold or rise, because sales spends its time on the leads that are actually worth working. Fewer MQLs, more customers.

That is the case for a strict, well-defined MQL. The goal is not a bigger number on a marketing dashboard. It is a shorter list that sales can trust.

What comes before and after

MQL sits in the middle of a longer lifecycle:

Subscriber → Lead → MQL → SAL → SQL → Opportunity → Customer

A SAL (Sales Accepted Lead) is a lead that sales has agreed is worth a first touch. Not every team names this stage, but it is useful for diagnosing handoff health.

An SQL is a lead that sales has validated as a real near-term opportunity, usually through discovery or a framework like BANT or MEDDIC.

The SAL stage between MQL and SQL is what lets you separate two different questions: is sales even looking at marketing’s MQLs (MQL→SAL), and is sales closing the ones it accepts (SAL→SQL). When MQL→SQL drops, the SAL stage tells you whether the leak is in scoring (low SAL acceptance) or in qualification (good SAL, weak SQL conversion). For the full lifecycle, see MQL vs SQL: a RevOps guide.

The metric that matters most

Once your definition is stable, stop watching MQL volume. Watch MQL→SQL conversion rate, with the definition held constant.

A stable definition and a falling conversion rate means something has changed. The usual suspects are ICP drift, a shift in channel mix, sales capacity, or a quietly loosened MQL bar. Check those before you assume the scoring model itself is broken. And resist the urge to fix a slow quarter by widening the definition. As the case above shows, that is the move that lowers conversions, not raises them.

A short summary

An MQL is a contact who has:

• Fit: matches your ICP
• Interest: has taken an action that signals buying consideration
• No disqualifier: no signal (unsubscribe, wrong domain, competitor) that cancels the first two out

All three, together. Test the definition by asking three people to apply it independently to 10 leads. If they do not agree, the definition is not working yet. And when pipeline gets tight, tighten the bar instead of loosening it.