A founder pinged me last month. Her marketing team had sent 812 MQLs to sales the previous quarter. Sales closed 4 deals. She wanted to know if that was normal.
It is normal. It is also the reason most B2B SaaS companies under Series B feel like their funnel is broken even when the top of it looks healthy. The number that matters is the one in the middle, MQL to SQL, and it gets less attention than almost any other metric on the dashboard.
I have spent the last decade rebuilding that middle of the funnel for B2B SaaS teams. Same pattern every time. The MQL number looks good in board decks. The closed-won number looks bad in pipeline reviews. Nobody can explain the drop because nobody measures the handoff with any real precision.
This is the post I wish I could send to every founder who asks me what a "good" MQL to SQL conversion rate is. The short answer is that the benchmark you read on a marketing blog is probably wrong for your business. The long answer is below.
What MQL to SQL conversion actually measures
MQL to SQL conversion is the percentage of marketing qualified leads that sales accepts as worth pursuing. It sits between two of the noisiest definitions in B2B, so the number drifts whenever either team changes how they qualify.
A marketing qualified lead is someone who has done enough to look interesting. They downloaded a whitepaper, hit pricing twice, opened five emails, or filled out a demo form. A sales qualified lead is someone a rep has talked to and confirmed is in fit, has a real problem, and has authority or access to it.
The conversion rate is simple math:
MQL to SQL conversion = (SQLs accepted in period) / (MQLs created in period) × 100
The reason it gets tricky is the period. If MQLs created in March become SQLs in April, you have to lag the denominator or your number swings 15 points month to month. Most teams do not.
The benchmark you are probably comparing yourself to is wrong
The number you keep reading is 13%. That is the cross-industry B2B average. It is also useless for a SaaS founder because it includes manufacturing, professional services, and physical product companies whose buying cycles look nothing like yours.
Here is what the data actually shows for B2B SaaS specifically.
If you are at 13% as a B2B SaaS company, you are not average. You are below average for your category, and the gap is almost always the same three things: how you define MQL, how fast you respond, and how dirty your data is.
Why the gap between 13% and 40% is mostly definition
The single biggest reason MQL to SQL conversion looks bad is that "MQL" got defined by marketing to hit a number on a dashboard. If your MQL definition is "filled out any form on the website," your conversion to SQL will look terrible because half those forms are job seekers, students, and competitors.
I worked with a Series A team last year doing exactly this. Marketing was reporting 600 MQLs a quarter. The CMO was happy. Sales was reporting 9 SQLs a quarter from those MQLs. That is a 1.5% conversion rate. The CRO was furious.
When we audited the MQL list, here is what we found:
- 38% had personal email domains
- 22% were students or interns based on title
- 14% worked at companies under 10 employees, well outside ICP
- 9% were existing customers re-downloading content
- 17% were actual ICP fits
We had not lost the funnel. We had lost the definition. After tightening MQL criteria to require company size, role seniority, and ICP industry, the MQL count dropped to 230. The SQL count went up to 41. Conversion went from 1.5% to 17.8% in one quarter. Nobody changed how they responded to leads or what they said in emails. They just stopped calling everything an MQL.
A bad MQL definition will sink your conversion rate before sales ever picks up the phone.
Most "low MQL to SQL conversion" problems are actually MQL inflation problems. Fix the input before you blame the handoff.
The five fixes that actually move the number
After running this audit on roughly 40 B2B SaaS teams, the same five things show up in priority order. Do them in this sequence. Skipping ahead does not work because each one depends on the one before it.
1. Rewrite the MQL definition with sales in the room
The MQL definition should not be a marketing artifact. It should be a contract between marketing and sales that says "if this person hits these criteria, sales will work them within X hours, no questions asked." If sales does not sign that contract, they will reject MQLs as soon as the pipeline gets tight, and your conversion will swing with the season.
A working MQL definition has three layers:
- Fit: company size, industry, geography, tech stack
- Intent: behavior in the last 14 days (pricing visit, demo request, content downloads)
- Role: title seniority and function
Without all three, you are just measuring web traffic. Read more on how to build the underlying ICP that this depends on.
2. Cut response time to under one hour
Speed beats quality up to a point. The data is brutal here. Contacting an MQL within 5 minutes makes you roughly 100x more likely to convert versus contacting them after 30 minutes. Wait an hour, you are still 7x ahead of the 24-hour responder.
Almost no B2B SaaS team I have audited responds in under an hour on average. Most are between 8 and 26 hours. The ones at 26 hours have stopped measuring because the answer is too embarrassing.
The fix is not to staff up. It is to automate the routing so the lead lands in front of a rep, in a Slack channel they actually watch, with a one-click meeting link already in the message. Build it right and a 2-person SDR team can respond in 4 minutes flat. We covered the mechanics in our lead routing rules guide.
3. Fix the contact data before sales touches it
Roughly 30% of B2B contact data is wrong or stale at any given moment. That means almost a third of your "unresponsive" MQLs were never reachable in the first place. Sales marks them as "no response" and they sit in the CRM as proof that marketing leads are bad.
Run waterfall enrichment on every MQL before it gets assigned. Verify the email is deliverable. Update the phone number. Pull a fresh job title. If you are using Clay or a similar tool, this is a 6-step workflow you build once. If you are not, the waterfall enrichment pattern is worth the read.
I have seen MQL to SQL conversion jump 8 points from this fix alone. That is not because the leads got better. It is because sales stopped wasting time on bad data and started spending it on the leads that were actually contactable.
4. Score on behavior, not just demographics
Most lead scoring models I see are demographic plus form fills. Title gets you 10 points. Company size gets you 15. Demo request gets you 25. That model has been around since 2012 and it still does not work because it ignores recency.
A VP at a 500-person company who filled out a form 9 months ago is not a hot lead. A director at a 100-person company who hit pricing 4 times in the last 3 days is. Behavioral scoring weights recent intent above demographic fit. Companies that do this hit 39 to 40% conversion versus 13 to 18% for demographic-only models.
The model does not have to be fancy. A 4-tier system based on recency and engagement frequency, combined with a hard ICP gate, will outperform a 50-criteria HubSpot scoring sheet in most cases. We wrote up the version we actually deploy in our lead scoring model post.
5. Close the loop with sales feedback
Every rejected MQL needs a reason code: wrong title, wrong company size, no budget, bad timing, out of geo, no real intent. Without this, marketing has no signal to adjust the definition with, and the cycle repeats.
The reason codes should feed back into the MQL model monthly. If 25% of last month's rejections were "wrong title," tighten the title filter. If 30% were "no real intent," your behavioral threshold is too loose.
This is the part everyone skips because it feels like process for the sake of process. It is the part that compounds. Teams that run a monthly feedback loop see qualification accuracy improve roughly 20% per quarter for the first two quarters before it plateaus.
How to measure the rate without lying to yourself
The biggest measurement trap is calculating MQL to SQL conversion as a same-month ratio. If your average MQL takes 9 days to become an SQL, a same-month calculation will undercount conversion at month start and overcount it at month end. Your number will look like it is improving when it is just calendar drift.
The honest way to calculate it is by cohort. Take every MQL created in March. Track them for 60 days. Count how many became SQLs in that window. That is your March MQL to SQL conversion. It will lag by 60 days, which is annoying but real.
The dashboard most teams need is two charts side by side:
- Cohort conversion (lagging, real)
- Same-month conversion (current, directional)
If those two numbers diverge by more than 10 points, you have a measurement problem, not a funnel problem.
Also worth tracking:
- MQL to SQL conversion by source (paid, organic, content, partner, referral)
- MQL to SQL conversion by SDR or rep
- Average time from MQL creation to SQL acceptance
- Reason codes for rejected MQLs, sorted by frequency
The source split matters a lot. SEO leads convert at roughly 50%, content downloads at 25 to 30%, paid at 25%, and cold sourced leads under 15%. If you do not segment by source, a shift in marketing mix will look like a conversion problem when it is actually a channel mix problem.
Why this is the wrong fight after Series B
For Series A and SMB B2B SaaS companies, MQL to SQL conversion is the right metric to obsess over. It tells you whether your top of funnel is feeding qualified work to sales. It tells you whether sales is taking that work seriously. It is the cleanest signal of sales and marketing alignment you can put on one chart.
After Series B, the conversation shifts. Pipeline coverage matters more. Account-level signals matter more. The MQL framework starts to feel quaint compared to intent-based, account-based motions where you are not really qualifying individual leads anymore but qualifying accounts and orchestrating multi-thread outreach.
So if you are reading this and you are at $50M ARR with a mature ABM motion, the answer to "should we improve MQL to SQL conversion" is probably "stop tracking MQLs altogether and move to opportunity-based reporting." Most teams reading this are not there yet. Use the metric while it is still useful.
Where a well-run B2B SaaS team should land within two quarters of fixing definition, response time, and data quality. Above 40% means your MQL bar is high and your funnel is tight. Below 15% means the definition is broken.
What I actually recommend if you are starting from scratch
Most founders ask me which fix to start with. The answer depends on which one you can ship this week. The ranking is not what looks most impressive on paper. It is what removes the biggest source of variance from your number.
Week one: audit the last 60 days of MQLs against your stated definition. If more than 25% fail the definition, your input is broken. Tighten the criteria, retrain marketing on what qualifies, and rerun the audit in 30 days.
Week two: pull the average time from MQL creation to first sales touch. If it is over 4 hours, fix routing before anything else. The cheapest improvement in B2B SaaS is making the people you already have respond faster.
Week three: sample 100 MQLs from last quarter that did not convert. Check the email deliverability. If more than 20% bounced, you have a data quality problem, not a conversion problem.
Week four: ask sales to label the last 50 MQLs they rejected with one reason code. Look at the distribution. The biggest bucket is what to fix in the MQL definition next quarter.
That is the order. It works because each step gives you data to make the next decision with, instead of guessing.
Want a second pair of eyes on your MQL handoff?
We run a 30-minute audit on your funnel definitions, response times, and data quality, then tell you the three changes that will move your MQL to SQL conversion the most in 60 days.
Book an audit →FAQ
What is a good MQL to SQL conversion rate for B2B SaaS in 2026?
The B2B SaaS median is around 25%. Top quartile teams hit 40%, and best-in-class with tight ICP and behavioral scoring get above 60%. If you are reading 13% in benchmark reports, that is the all-industry B2B number including manufacturing and professional services. It is not the right comparison for software.
How long should MQL to SQL conversion take?
Average time from MQL creation to SQL acceptance should be under 7 days for inbound, and under 21 days for cold-sourced leads. If yours is longer, you are either waiting too long to make first contact or running too many nurture cycles before sales gets involved.
Should marketing or sales own the MQL to SQL number?
Both, jointly. If marketing owns it alone, they hit the number by lowering the MQL bar. If sales owns it alone, they hit the number by rejecting more leads. The metric belongs in a shared RevOps dashboard with both teams reviewing it monthly and a written definition agreement they both signed.
Is MQL dead in 2026?
Not for companies under $50M ARR. The "MQL is dead" argument is mostly coming from large enterprise SaaS teams running mature ABM programs where individual lead qualification is less useful than account-level orchestration. For most B2B SaaS companies in Series A or B, MQL is still the right unit of work for marketing and the right handoff signal for sales.
How does AI change MQL to SQL conversion?
The biggest shift is in qualification speed and data quality. AI tools like Clay and Apollo can enrich and score an MQL in seconds instead of hours, which collapses the response time gap. AI also makes behavioral scoring much cheaper to build and maintain, so smaller teams can run a top-quartile scoring model that would have needed a data scientist 3 years ago. Read more about how this works in practice in our guides on CRM and RevOps tooling and AI automation.
Closing thought
MQL to SQL conversion is the metric where marketing and sales finally have to agree on what good looks like. Most teams never do, which is why the number stays bad. The fix is not a new tool. It is a written definition both teams sign, a faster handoff, cleaner data underneath, and a monthly feedback loop that lets the model learn from itself.
If you want to talk through what that looks like for your specific funnel, we audit B2B SaaS RevOps setups every week. The first 30 minutes are free. The follow-up depends on what we find.