The definitions, briefly
A marketing qualified lead is a contact who has taken an action marketing considers a buying signal, and who fits the target profile. A downloaded report, a demo request, a webinar registration.
A sales qualified lead is an MQL that sales has accepted: a real company, a real need, a real budget, a real timeline, and a conversation worth having.
Every company defines both differently, and that is fine. What is not fine is defining an MQL by the cheapest action available and then wondering why sales ignores them.
What each stage actually costs
Here is the part most MQL-versus-SQL articles leave out. These stages have prices, and the prices are very far apart.
| Stage | Median cost | What the person actually did |
|---|---|---|
| Landing page click | $14.83 | arrived on your site |
| Lead form open | $18.21 | tapped the form |
| Completed form submission | $140.49 | gave you their details |
| Call booking | $342 | agreed to talk |
| Demo request | $350 | asked to see the product |
The jump from $18.21 to $140.49 is the most expensive step in a B2B funnel, and it is almost pure friction rather than qualification. Note too that a demo costs 65% more than a call, for the same placement and the same audience. That is not a lead quality difference, it is the price of the commitment you asked for.
Every figure here comes from our own panel: 1,000+ advertiser accounts, 22,942 campaigns and $79 million of measured LinkedIn spend across 1.34 billion impressions in the twelve months to 7 September 2026. Medians across advertiser accounts, never pooled totals. LinkedIn Audience Network delivery excluded.
Why the handoff breaks
The classic failure is that marketing is measured on MQL volume and sales is measured on closed revenue. Those two incentives point in opposite directions, and the cost table above explains why.
If marketing is paid per MQL, the rational move is to define an MQL as the cheapest countable action. At $18.21 a form open against $140.49 a submission, loosening the definition cuts reported cost per lead by roughly 7.7 times overnight without a single thing improving.
We see this constantly in reporting. An agency or in-house team quotes a cost per lead that looks remarkable, and it turns out to be measuring oneClickLeadFormOpens rather than oneClickLeads. Both are real fields. Only the second produced a human sales can call.
So before arguing about MQL definitions, check which field the number comes from. A surprising share of marketing-versus-sales conflict is a measurement disagreement wearing a strategy costume.
The 88% nobody counts
Across 350 advertiser accounts running lead gen forms, the median completion rate is 11.6%. Eighty-eight percent of people who open a form abandon it.
Those people were interested. They tapped. They are not in your CRM, they are not an MQL, they are not anything, and you paid for every one of them.
The usual cause is field count. Every field added because somebody in sales wanted better qualification is paid for at the completion rate, not the open rate. Asking for a phone number to improve lead quality can reduce lead quantity by more than it improves anything, and the trade is rarely calculated.
This is the cheapest fix in the entire MQL conversation and it sits in a form builder, not in a definitions workshop.
How to define an MQL that survives contact with sales
Define it by the action that costs something to take. A completed form, a demo request, a reply, a booked meeting. Not an open, not a page view, not a video watch.
Add a fit filter that both teams agreed before the quarter started, not during the argument. Company size, industry, seniority, territory.
Then set the volume expectation from the cost table rather than from a target. A team with $20,000 of monthly media and a $350 demo cost can produce about 57 demo requests, and no definition change creates more of them. Setting an MQL target above what the budget can buy guarantees the definition will be loosened to meet it.
What to report instead
Report influenced pipeline by account rather than lead counts. In a $20,000-plus deal there are five to ten people on the buying committee, and attributing the deal to whichever one filled in a form is close to meaningless.
Report landing page clicks rather than LinkedIn’s default clicks field, which counts social engagement as traffic.
And put a self-reported attribution question on the booking form. It is the only mechanism that captures AI assistants, word of mouth and offline referral, none of which appear in any analytics tool, and it takes an afternoon to add.
Frequently asked questions
What is the difference between an MQL and an SQL?
An MQL is a contact marketing believes is worth a sales conversation based on their behaviour and fit. An SQL is an MQL that sales has accepted after qualifying need, budget and timing. The definitions are company-specific; the useful discipline is defining the MQL by an action that cost the person something.
What is a good MQL to SQL conversion rate?
It varies too widely by definition to publish a single benchmark honestly. The more useful control is the cost of each stage: a lead form open costs $18.21 and a completed submission $140.49 across our panel, so a loose MQL definition can cut reported cost per lead 7.7x without anything improving.
Why does sales reject our MQLs?
Most often because the MQL is defined by a cheap action. If an MQL is a form open rather than a completed submission, 88% of them never finished giving you their details, and sales is right to reject what arrives.
Should we count lead form opens as leads?
No. LinkedIn exposes oneClickLeadFormOpens and oneClickLeads as separate fields. Only the second is a lead. The first is roughly 7.7 times cheaper, which is exactly why it ends up in reports.