Why Are Our MQLs Not Converting to SQLs? How to Fix It

Your marketing team hit its target for marketing-qualified leads (MQLs) again this quarter. Your sales team says the leads are junk, and the number of sales-qualified leads (SQLs), the ones sales has actually confirmed as real opportunities, barely moved. In the board deck, 600 MQLs turned into a few dozen opportunities, and nobody in the room can explain the gap in a way both teams agree with.

That last part is the clue. In most companies the gap isn’t caused by bad leads or lazy reps. It’s caused by two teams counting different things and calling them by the same name. In a Gartner survey of 243 chief sales officers and senior sales leaders published in May 2025, 49% said their sales organization’s definition of a qualified lead differs greatly from marketing’s. If the definitions differ, the conversion rate between them can’t be anything but disappointing.

This guide shows you how to find where your MQLs stall, using data already in your CRM. It covers how to agree one definition with sales, how to rebuild a lead score that rewards buyers instead of clicks, and how to run the handoff so nothing sits untouched. It ends with a worked sample reset and a 30-day plan you can start next week.

Why are our MQLs not converting to SQLs?

Because the two teams are counting different things, and the overlap between them is smaller than either side thinks. Marketing counts leads that showed interest. Sales needs leads that can buy. Only the overlap turns into pipeline.

When marketing says “qualified”, it usually means a contact crossed a score threshold: they downloaded a guide, came to a webinar, opened a few emails and visited the site. When sales says “qualified”, it means something narrower: the right kind of company, a real problem, someone with a role in the buying decision and a reason to act this quarter. Both definitions make sense on their own. Put side by side, they describe different people.

That’s why the usual fixes don’t work. Buying more leads makes the non-overlapping part bigger. Pushing sales to “work every MQL” burns rep time on contacts who were never going to buy, and teaches reps to ignore the next batch. The fix starts with finding where your own MQLs stall. Then you can work on the right step.

Run the checks below on one month of MQLs. Most CRMs, including HubSpot, have the lifecycle stage, owner, lead status and timestamps you need. Find the first row that matches what you see.

The thresholds are practitioner rules of thumb, not industry standards. Use them to decide where to look first, not as pass-or-fail scores.

Definition signs: when “qualified” means two things

The fastest test takes five minutes. Ask your head of marketing and your head of sales, separately, to write one sentence that finishes “A qualified lead is…”. If one sentence talks about engagement and the other talks about budget, timing and authority, you’ve found the root cause. Everything downstream, from the score to the dashboards to the arguments in the pipeline meeting, inherits that mismatch.

Scoring signs: when the score counts attention

Open your scoring model and sort the rules by points. If the top of the list is email opens, content downloads, webinar attendance and blog visits, the score is measuring how much attention someone paid you. That isn’t the same as whether they can buy. The tell in the data is a long tail of MQLs from students, consultants, competitors and tiny companies who engage a lot and never convert.

Handoff signs: when good leads go cold

If the definition and score look reasonable but SQLs still lag, follow the lead after it becomes an MQL. Who owns it? How long until the first touch? Did sales accept it, reject it, or just leave it? A strong lead that waits two days in a shared queue, gets one call and is then forgotten looks exactly like a bad lead in the monthly report.

Is our lead score measuring the wrong thing?

Usually, yes. Most lead scores add points for attention, like opens, downloads and page views, rather than for fit and buying intent. Two leads can reach the same score while only one of them could ever buy.

Lead scoring was built for a world where engagement was a decent proxy for interest. Someone who read five emails and downloaded two guides was probably researching a purchase. That proxy has weakened for two reasons. Buyers now do most of their research without talking to you, so plenty of engaged readers are just learning. And some of the signals scores rely on most no longer measure a person at all.

Look at the two leads in the image. Under an activity score, a student at a two-person agency who opens emails and attends a webinar becomes an MQL. A Head of Finance at a 300-person company who visited the pricing page twice in a week but never opened a newsletter might not. Sales gets the first, misses the second, and concludes marketing sends junk.

Fit: who they are

Fit is everything you can know about a lead before they do anything: company size, industry, region, tech stack, and the person’s role and seniority. Treat fit as a gate, not a bonus. A lead outside your ideal customer profile shouldn’t become an MQL no matter how engaged they are. They can stay in nurture, and if they’re a real exception, they’ll raise their hand.

Intent: what they do that signals buying

Intent comes from actions that point at a purchase rather than at learning. Visits to pricing, comparison, integration and security pages. Repeat visits within 7–14 days. Several people from the same company showing up at once. And above all, a direct request: a demo form, a trial, a “talk to sales” chat. Weight hand-raisers highest. A demo request from a good-fit company is an MQL by default and should go straight to sales.

Decay and the account view

Points should lose value over time. A webinar attended eight months ago says nothing about today. A simple rule, such as halving activity points after 30 days without new activity, keeps old engagement from pushing stale contacts over the line.

Score the account, not just the person. In Forrester’s State of Business Buying, 2026 report, published in January 2026, the typical buying decision includes 13 internal stakeholders and nine external influencers. One engaged contact at a company is interesting. Three people from the same account on the pricing page in a week is a buying signal.

What’s a good MQL-to-SQL conversion rate?

The one you measure honestly, by cohort, and improve quarter over quarter. Published benchmarks vary too much with how each company defines an MQL to be a useful target.

Search for an MQL-to-SQL benchmark and you’ll find the same handful of numbers repeated across dozens of articles, most often “about 13%”. When we went looking for where that figure comes from, different pages credited it to different reports, and we couldn’t trace it to a 2025 or newer primary study. That doesn’t mean it’s wrong. It means it isn’t a number you should run your team against.

There’s a deeper problem too. The rate depends almost entirely on the definition. A company that only calls demo requests from good-fit accounts MQLs will convert a high share of them. A company that counts every ebook download will convert a low share, even if it produces more pipeline overall. Comparing the two rates tells you nothing about which company is doing better.

How to measure your own rate properly

1. Cohort by creation month. Take every MQL created in a given month and count how many became SQLs within 60 to 90 days. Dividing this month’s SQLs by this month’s MQLs mixes different groups of leads and swings with timing.
2. Split by source. Report hand-raisers (demo requests, trials, contact-sales), content leads (downloads, webinars) and event leads separately. Blended rates hide the fact that each source behaves differently.
3. Count the steps in between. For one month, count how many MQLs were routed to an owner, accepted by sales, worked within a few days, and qualified as SQLs. The step with the biggest drop is your first fix.
4. Keep the definition fixed while you measure. If you change the score mid-quarter, start a new baseline.

Do this for one quarter and you’ll have a rate you can defend, a clear view of which source is dragging it down, and a baseline to beat next quarter. That’s more useful than any industry median.

Who should own the definition of a qualified lead?

Both teams, with one owner for each step. Marketing drafts the criteria, sales signs them off, and RevOps builds them into the CRM. Nobody changes the definition alone.

Most definition problems are ownership problems. Marketing owns the score because it owns the automation tool. Sales owns the rejection because it owns the pipeline. RevOps, if it exists, gets asked to “fix the routing” without anyone agreeing what should be routed. Each team optimizes its own step, and the handoff in the middle belongs to no one.

The one-page definition

Write the agreement down. One page, dated, signed by the heads of marketing and sales. It needs four parts:

• MQL: fit criteria (company size, industry, region, role) plus at least one intent trigger (a hand-raise, pricing or comparison visits, or several contacts from one account).
• Sales-accepted: sales agrees the lead meets the MQL definition and takes ownership.
• SQL: sales has confirmed a real problem, a contact with a role in the buying group, and a reason to act within a stated time frame.
• Recycle: what happens to rejected leads, and the reasons sales can pick from.

The workshop that produces it takes about an hour. Bring 20 recent closed-won deals and 20 recent rejected MQLs, and ask both teams what the first group had in common that the second didn’t. The answers become the criteria.

Agree the message, not just the score

The same agreement should cover what buyers hear. In a Gartner survey of 632 B2B buyers published in June 2025, 69% reported inconsistencies between what was on a company’s website and what its sellers told them. A lead who downloaded a guide promising one thing and then gets a sales call about something else is a lead you’re about to lose. Share the top three messages and proof points between both teams.

Keep it alive

Review rejected MQLs every month with both teams in the room: which reasons come up most, and whether the score should change. Review the definition itself every quarter, or whenever you change pricing, launch a product or move into a new segment.

How do we fix MQL-to-SQL conversion?

Agree the definition, rebuild the score on fit and intent, route every MQL to an owner, and make sales accept or reject within 24 hours with a reason. In that order, because each step depends on the one before it.

Step 1: Run the definition workshop

Sixty minutes, both leaders, the 20-and-20 deal review described above. Leave with a signed one-page definition. Don’t touch the score until this exists, or you’ll rebuild it around the old disagreement.

Step 2: Rebuild the score

Turn the definition into rules. Fit criteria become a gate: outside the profile, no MQL. Intent actions become points, with hand-raisers at the top and pricing, comparison and multi-contact activity next. Opens get zero. Add decay. Then back-test: run last quarter’s leads through the new score and check that it would have flagged most of the leads that became SQLs, and few of the ones sales rejected.

Step 3: Route every MQL to a named owner

No shared queues. Use owner rules first: territory, segment, or an existing account owner. Fall back to round-robin. Alert the owner as soon as the lead becomes an MQL, and alert their manager if nothing happens. Demo and trial requests from good-fit companies should reach a rep the same day.

Step 4: Accept or reject within 24 hours, with a reason

This is the step most teams skip, and the one that changes the relationship between the teams fastest. Sales commits to accepting or rejecting every MQL within 24 hours. A rejection requires a reason from a short list: wrong company, wrong person, no current need, duplicate, can’t reach. Now marketing knows why leads fail, and sales has a way to say “not this one” without just ignoring it.

Step 5: Recycle rejects and review monthly

Rejected leads go back to nurture with their reason attached, not into a void. Once a month, look at the reject reasons as a team. If “wrong company” dominates, tighten the fit gate. If “no current need” dominates, raise the intent bar. If “can’t reach” dominates, look at data quality and follow-up.

Worked example: a sample MQL reset

Here’s what this looks like for a fictional Series A company, “Ledgerline,” selling accounts-payable automation to mid-sized finance teams and running marketing and sales in HubSpot. The numbers are illustrative sample data, built to show the method. They compare the quarter before the reset with the quarter after.

In Q1, Ledgerline’s marketing team was hitting its MQL target and sales was openly ignoring the list. The step count told the story. Of 640 MQLs, sales accepted only 38%. Most of the rest weren’t formally rejected; they were simply never touched, with no reason logged. The leads sales did accept waited a median of 26 hours for a first call.

The definition workshop found the cause in half an hour. Marketing’s score gave 25 points for email opens and 20 for an ebook download, so finance students, consultants and two-person firms routinely crossed the MQL line. Sales’ idea of qualified was a finance leader at a company with 100 to 1,000 staff, already using an accounting system Ledgerline integrates with.

The team rebuilt the score with fit as a gate and opens at zero, sent demo requests straight to named reps, and introduced a 24-hour accept-or-reject rule with five reject reasons. MQL volume fell by more than a third, which made for an uncomfortable first week in the marketing meeting. But sales accepted 71% of the new MQLs and worked them the same day, SQLs rose 34%, and pipeline from SQLs rose 40%.

The reject reasons were the unexpected win. “Wrong company” fell from the most common reason to the third. “Can’t reach” rose to the top, which pointed to a phone-number gap on the demo form that marketing fixed in a day.

Should we drop MQLs altogether?

Not until you have something better to hand off. Keep a stricter MQL built on fit and intent, but judge marketing on SQLs and pipeline, and treat MQL volume as a leading indicator only.

The case against MQLs is strong, and it’s getting stronger as buyers change how they buy. In the same June 2025 Gartner survey of 632 buyers, 61% said they prefer an overall rep-free buying experience, and 73% said they actively avoid suppliers who send irrelevant outreach. A content download, in that world, is rarely a request to talk to sales. Treating it as one annoys the buyer and wastes the rep’s time.

At the same time, the decision is often made before anyone talks to sales. In 6sense’s 2025 B2B Buyer Experience Report, a survey of nearly 4,000 buyers, 95% of purchases went to a vendor that was already on the buying group’s Day One shortlist. The signals that matter most are the ones that show a company is evaluating you now. Interest in your content matters much less.

The middle path most teams should take

Dropping the MQL entirely sounds bold, but it leaves sales with no agreed signal of when to step in, and it usually just moves the argument to a new label. A better approach for most Series A and B teams:

1. Hand-raisers go straight to sales. Demo requests, trial sign-ups and contact-sales forms from good-fit companies skip scoring entirely.
2. Content leads stay in nurture until they show intent, such as pricing visits, repeat visits or a second person from the account.
3. Account-level intent creates the MQL, not one contact’s activity.
4. Marketing is measured on SQLs and pipeline, by source. MQL volume stays on the dashboard as an early warning, not as the target.

This keeps a clear handoff signal while removing the incentive to inflate it. When marketing’s goal is SQLs and pipeline, nobody benefits from pushing a student over the MQL line.

Putting it to work: a 30-day MQL reset

You can run the reset in four weeks with the heads of marketing and sales in the room, plus a few hours from whoever runs your CRM. The output is a signed definition, a fit-first score and a monthly SQL report: the core of a lifecycle and handoff system both teams trust.

Week 1: Count the steps and agree the definition

Pull last month’s MQLs and count them at each step: routed, accepted, worked, SQL. Run the 60-minute workshop with 20 closed-won deals and 20 rejected MQLs. Write and sign the one-page definition.

Weeks 2–3: Rebuild the score

Turn the definition into scoring rules in your CRM: fit as a gate, intent as points, opens at zero, decay after 30 days. Back-test it against last quarter’s leads before switching it on.

Week 2: Routing and the SLA

Set owner rules and a round-robin fallback, remove shared queues, and add alerts for the owner and their manager. Agree the 24-hour accept-or-reject rule and the same-day rule for demo requests.

Weeks 3–4: Reject reasons live

Make the reject reason a required field with five or six options. Send rejected leads back to nurture automatically, with the reason attached.

Week 4: Report SQLs and pipeline

Build the monthly report: MQLs by source, accepted rate, MQL-to-SQL rate by creation-month cohort, SQLs and pipeline by source, and the top reject reasons. Book the first monthly review with both teams.

The checklist

• Leads counted at each step for one month: MQL, routed, accepted, worked, SQL
• One-page definition written and signed by both leaders
• Score rebuilt: fit as a gate, intent as points, zero for opens, decay on
• New score back-tested on last quarter’s leads
• Every MQL routed to a named owner; no shared queues
• 24-hour accept-or-reject rule live, with a required reject reason
• Rejected leads recycled to nurture with the reason attached
• Monthly report leads with SQLs and pipeline by source

If you’d rather have someone run this with you, that’s exactly what the free roadmap call is for.

Sources

• Gartner, Survey Reveals Less Than Half of CSOs Report Their Organization Met Several 2024 Strategic Goals (21 May 2025) — 243 CSOs and senior sales leaders, Nov–Dec 2024
• Gartner, Sales Survey Finds 61% of B2B Buyers Prefer a Rep-Free Buying Experience (25 Jun 2025) — 632 B2B buyers, Aug–Sep 2024
• Forrester, The State Of Business Buying, 2026 (21 Jan 2026) — Forrester’s buyer research
• Litmus, Email Client Market Share (September 2026) — 1.5 billion+ opens
• 6sense, 2025 B2B Buyer Experience Report — ~4,000 buyers, Nov 2025