Every B2B company has the same argument. Marketing says it delivered 400 leads this quarter. Sales says 30 of them were worth a call. Both are right, and the reason is that nobody wrote down what “worth a call” means. This article gives you a working definition of MQL and SQL, a handoff checklist you can adopt this week, and the three numbers that tell you whether the handoff is working.
MQL vs SQL in one paragraph
A Marketing Qualified Lead (MQL) is a contact who fits your target customer and has shown enough interest that marketing believes sales should talk to them. A Sales Qualified Lead (SQL) is a contact a salesperson has spoken to, or at least reviewed, and confirmed has a real problem, a plausible budget and a reason to act. The MQL is marketing’s opinion. The SQL is sales’ opinion. The handoff is the moment one becomes the other, and it fails when the two opinions are formed using different criteria.
| MQL | SQL | |
|---|---|---|
| Who decides | Marketing, usually via a scoring rule | Sales, usually after a conversation |
| Based on | Fit (company size, industry, role) plus behavior (pages viewed, content downloaded, demo requested) | Need, budget, authority, timing, confirmed by a human |
| Typical trigger | Score crosses a threshold or a high-intent form is submitted | Discovery call completed and criteria met |
| What happens next | Routed to an SDR or AE within an agreed time | Opportunity created in the CRM with a value and a close date |
| Owner of the metric | Marketing | Sales |
Why the handoff breaks
Four root causes account for nearly every broken handoff we have seen:
- The MQL definition rewards volume. If marketing is measured on MQL count, the threshold drifts downward until anyone who opens two emails qualifies. Sales stops trusting the queue and works its own list instead.
- No service level on follow-up. An MQL that waits three days for a call is a cold lead by the time someone dials. Sales blames lead quality when the real problem was speed.
- No feedback loop. Sales rejects leads silently. Marketing never learns which rejections were about fit, which were about timing and which were about a bad phone number.
- Different systems, different fields. Marketing scores in the automation platform, sales works in the CRM, and the sync drops half the context. The rep sees a name and a company and nothing else.
None of these are personality problems. They are process gaps, and a checklist closes them.
The handoff checklist
Adopt the whole thing or adopt the first five items and add the rest as you grow. Each item has an owner and a place where it is written down.
Before the lead is handed over (marketing owns)
- Fit criteria are explicit. Company size range, industries, geographies and roles that count as target. Written in a shared document and reflected as required fields in the scoring model. If you do not have a scoring model yet, our guide to the BANT framework is a reasonable place to start.
- Intent criteria are explicit. Which actions count and how much. A pricing page visit is not the same as a blog visit. A demo request is not the same as a webinar registration.
- Negative criteria exist. Students, competitors, existing customers, job seekers and free email domains are excluded automatically. This single rule typically removes 15 to 30 percent of the noise.
- The record is complete. Before routing, the lead has a valid email, a company, a role, the source campaign and the last three meaningful actions. If any of these are missing, it is not an MQL yet, it is a contact to enrich.
- Routing is automatic and logged. The lead is assigned by territory, segment or round robin without a human touching it, and the assignment timestamp is stored.
At the moment of handoff (shared)
- Response time is agreed and measured. Common targets: first touch within 15 minutes for demo requests, within one business day for content-driven MQLs. Measure it from assignment timestamp to first logged activity.
- Context travels with the lead. The rep sees the source, the campaign, the content consumed and any notes from chat or events. If your systems cannot sync this, put it in the lead description field. Ugly beats absent.
- The rep has a script for the first touch. Not a pitch, a reason for the call that references what the lead actually did. “You downloaded our pricing benchmark on Tuesday, I wanted to check whether the numbers matched what you are seeing.”
After the conversation (sales owns)
- Every MQL gets a disposition within five business days. Accepted, Recycled or Rejected. Nothing sits in “new” for a month.
- Rejections carry a reason code. Keep the list short: bad fit, no need, no budget, no timing, unreachable, bad data. Free text is optional, the code is mandatory.
- Accepted leads become SQLs with a value and a next step. An SQL without an opportunity record is a rumor.
- Recycled leads go back to marketing with a date. “Revisit in Q1” is actionable. “Not now” is not.
Ongoing (leadership owns)
- Weekly 20-minute review. Marketing and sales look at last week’s MQLs together: how many, how fast the follow-up, how many accepted, top rejection reasons.
- Quarterly definition review. Adjust thresholds and criteria based on which MQL sources actually became revenue. This is where the definition stops being a debate and starts being data.
The three numbers that matter
You can track twenty metrics around the handoff. Three of them explain almost everything:
| Metric | How to calculate | Healthy range | What a bad number means |
|---|---|---|---|
| MQL to SQL rate | SQLs created / MQLs delivered, same cohort | Roughly 20 to 40 percent for most B2B, higher for demo-request-heavy funnels | Below 15 percent: the MQL bar is too low or routing is broken. Above 60 percent: the bar is probably too high and you are starving the pipeline. |
| Speed to first touch | Median minutes from assignment to first logged activity | Under 30 minutes for hand-raisers, under 24 hours for everything else | Slow follow-up quietly destroys conversion. Fix this before touching the scoring model. |
| Rejection reason mix | Share of rejections per reason code | No single reason above 40 percent | “Bad fit” dominating means scoring is wrong. “Unreachable” dominating means data quality is wrong. “No timing” dominating means recycle, do not reject. |
Track these by source as well as in total. It is common to find that webinar MQLs convert at half the rate of pricing-page MQLs, and that changes where marketing spends. If you already measure cost per qualified lead rather than cost per lead, the MQL to SQL rate per source is the missing multiplier.
A worked example
A 60-person software company delivers 250 MQLs in a quarter. Sales accepts 45, rejects 130 and leaves 75 untouched. On paper the MQL to SQL rate is 18 percent, which looks like a lead quality problem.
Adding reason codes and timestamps changes the picture. Of the 130 rejections, 70 are coded “unreachable,” and the median time to first touch for those was four days. Of the 75 untouched leads, 60 came from a single webinar and were routed to one rep who was on leave. The real problems are a routing gap and slow follow-up, not the scoring model. Fixing both raised the next quarter’s rate to 31 percent without a single change to the MQL definition.
Common questions
Should an SDR or the AE qualify the MQL?
If you have SDRs, they own the MQL to SQL step and the AE owns the opportunity. If you do not, the AE does both and the response time target needs to be realistic. Either way, one named owner per lead.
What about product qualified leads?
For products with a free tier or trial, usage signals such as inviting teammates or hitting a limit are often stronger than any marketing action. Treat a Product Qualified Lead (PQL) as a special MQL with its own scoring and a faster response target. The handoff checklist applies unchanged.
Do we need lead scoring software?
Not to start. A spreadsheet with fit points and intent points, updated by a weekly export, is enough to get the definitions right. Buy software once the definitions are stable and the volume makes manual scoring painful.
Start here
If you do nothing else this week: write the fit and intent criteria on one page, add reason codes to the rejection field in your CRM, and measure speed to first touch. Those three changes end most of the marketing versus sales argument, because they replace two opinions with one set of facts.