MQL vs SQL: what makes a lead qualified

A marketing qualified lead has done something that suggests interest. A sales qualified lead has done something that suggests intent to buy. That is the whole distinction, and you can find it in a hundred articles.

What those articles skip is the question underneath it: what counts as a lead at all? Get that wrong and the MQL and SQL columns are precise labels on a number that means nothing.

Define the lead before you grade it

On the accounts we run, a lead is a tracked form submission, a tracked phone call, or a tracked WhatsApp conversation. Writing it down that plainly matters for three reasons.

• Channels have different default definitions. A form fill is counted automatically. A phone call is only counted if call tracking is installed. A WhatsApp message is only counted if the click is tagged. Leave one out and that channel looks like it produces nothing.
• Tracking breaks silently. A form that stops recording conversions does not raise an error. It reports zero, which is indistinguishable from a bad month. We had exactly this on a client account for eleven days, and it made the following report understate real performance.
• "Leads" in a proposal and "leads" in your CRM are usually different things. Ask any agency which of the three they are counting before comparing their number to yours.

What actually separates an MQL from an SQL

Not the amount of engagement. Somebody who has read nine blog posts is interested in the topic. Somebody who has read the pricing page twice is interested in buying. The signal is the kind of page, not the count.

The wrong next steps in that table cause most of the damage. Phoning someone who downloaded a checklist teaches them not to download anything again. Putting a person who asked for a price into a nurture sequence loses the deal to whoever picked up the phone.

Lead scoring, and when it is not worth it

Scoring assigns points to actions and attributes, and routes anything over a threshold to sales. It works when you have enough volume that a human cannot read every enquiry, and enough closed deals to know which signals preceded them.

Below that, it is a machine for encoding guesses. If you get thirty enquiries a month, read them. A scoring model built without closed-won data will confidently rank leads by criteria nobody has tested, and it is harder to unlearn than to skip.

The one metric that tells you if this is working

MQL to SQL conversion rate. If it is very low, marketing is counting things sales does not want. If it is very high, the bar is too low and sales is doing marketing’s filtering for it. Neither is fixed by generating more leads.

The related failure is optimising to the wrong end of the funnel. An ad account tuned to maximise lead volume will find you the cheapest possible form fills, and cheap form fills are cheap for a reason. Tuning the same account toward commercial intent produced this, on real months from an account we run:

A short checklist

1. Write down what counts as a lead. Include calls and messaging apps, not just forms.
2. Test each one end to end, and check the conversion actually records.
3. Agree with sales, in one sentence, what makes a lead theirs.
4. Track MQL to SQL conversion monthly. It is the number that catches a definition drifting.