THE SHORT ANSWER An MQL is a behavioral signal that someone looks interested; an SQL is a lead sales has verified as worth pursuing — right fit, real intent, and an owner who accepts it. The distinction only works when both teams share a written definition, a fast handoff, and quality judged by revenue rather than funnel stages.
Plain English first. An MQL — marketing qualified lead — is someone whose behavior says "interested": they downloaded the guide, attended the webinar, visited the pricing page twice. A SQL — sales qualified lead — is someone sales has verified is worth pursuing: right fit, real intent, and a genuine possibility of buying. MQL is a signal. SQL is a verdict.
Simple enough — and yet the space between those two acronyms is where more B2B revenue arguments happen than anywhere else in business. Here's why, and how to fix it.
Where the model breaks
- The vanity MQL. When marketing gets measured on MQL volume, MQLs get cheaper to earn — one whitepaper download and a pulse. The stage inflates, sales stops trusting it, and "qualified" becomes a word nobody believes. Volume metrics corrupt whatever they measure.
- No written definition. Ask sales and marketing separately what qualifies a lead and you'll get two different answers — which means every disputed lead is settled by whoever complains loudest. The alignment war is mostly a definition war in a trench coat.
- The handoff void. The MQL-to-SQL transition is a physical event — a lead moving from one team's queue to another's — and leads rot in that void for hours or days. A "qualified" lead dialed late performs like an unqualified one, and then gets blamed for it.
- Qualification as procrastination. Some floors "qualify" so hard they're really just cherry-picking — disqualifying anyone who'd require actual selling. Rigor is good; cowardice with a scorecard is not.
Write YOUR definition — the three-part test
Generic frameworks don't survive contact with a real pipeline. Build your own qualified-lead definition on three axes, in writing, with both teams signing it:
- Fit: do they match who you actually sell to? Industry, size, geography, budget authority — the demographic floor. (Steal from your closed-won list, not your wish list.)
- Intent: did they do something a buyer does? Requesting contact, asking about pricing, and booking a call are intent. Downloading one PDF is curiosity.
- Behavior threshold: the specific, observable actions that trip the wire — spelled out so a robot could apply the rule. "Visited pricing twice AND fits the profile" is a definition. "Seems warm" is a mood.
Then add the part everyone skips: the recycle rule. A lead that fails qualification today isn't garbage — it goes back into a nurture cadence with a defined re-entry trigger. Disqualified is a timestamp, not a tombstone.
The handoff SLA: where definitions become revenue
A definition without a service-level agreement is a philosophy. The working version has numbers on both sides: marketing routes qualifying leads to sales instantly — not in a daily batch; sales makes first contact in minutes and commits every accepted lead to the full cadence — not two dials and a shrug; and every disputed lead gets dispositioned with a reason code, because "bad lead" without a why teaches nobody anything. Review the SLA numbers weekly in the one meeting both teams attend. The definition tells you what a qualified lead is; the SLA is what makes the definition worth the paper.
Measure quality in revenue, not stages
Here's the trap even disciplined teams fall into: optimizing stage conversion instead of dollars. The only true test of lead quality is downstream — cost per closed customer by source, the same math that governs everything else. A source producing "worse" MQL-to-SQL rates but cheaper actual customers is a better source, full stop. Stages are for managing workflow; revenue is for judging quality. Confusing the two is how companies kill their best channels while celebrating their prettiest funnels.
The operator's shortcut: if all this feels heavy for your size, compress it. Small teams need exactly two things from this whole framework — a one-paragraph written definition of a lead worth calling, and a speed-plus-cadence commitment for every lead that meets it. The acronyms are optional. The discipline isn't.
Lead scoring: useful tool, easy trap
Most CRMs will happily score leads for you — points for the pricing-page visit, points for the title, points for opening three emails. Used well, scoring automates your written definition: the criteria you agreed on become the model, and the handoff fires itself. Used lazily, it's the vanity MQL problem with extra math — a lead "scores 85" because it opened newsletters, sales wastes an hour, and trust in the whole system erodes. Two rules keep scoring honest: the score must be explainable (any rep can see exactly why a lead qualified — black-box scores breed black-box excuses), and the model gets audited quarterly against closed-won — pull last quarter's customers, check what they scored at handoff, and tune the criteria toward what your actual buyers did rather than what a template assumed. Scoring is a servant of the definition, never a substitute for one.
Straight answers
Fit plus verified intent: they match your real customer profile and they've taken a buyer's action — and a human on the sales side has accepted responsibility for pursuing them. Acceptance matters; a lead nobody owns is qualified for nothing.
Whatever your history says it is — benchmarks vary wildly by industry and by how honest your MQL definition is. Track your own rate, watch its trend, and treat sudden jumps with suspicion: they usually mean a definition got looser, not a funnel got better.
No — they're just not sales' job yet. An MQL is a nurture asset: someone who raised a hand halfway. The waste isn't in generating them; it's in either dumping them on sales prematurely or letting them rot with no cadence at all.
