An MQL is a lead that marketing judges worth a salesperson’s time, based on fit and a clear sign of interest. An SQL is a lead that a salesperson has spoken with and confirmed as a real opportunity. Marketing sets the first mark within a day. Sales sets the second within days. Ad platforms should learn from the SQL.
The terms come from B2B sales practice. They spread through the SiriusDecisions Demand Waterfall, first launched in 2006 (Forrester on the Demand Waterfall). No platform or standard defines them for your company. Each company writes its own rules, and most of the trouble starts when marketing and sales write different ones.
What is the difference between an MQL and an SQL?
An MQL, or marketing qualified lead, is a lead that marketing has checked against rules both teams agreed on. The company fits the customers you sell to, and the person showed a clear sign of interest, such as asking for a quote. A lead score or a person can do the check. No one from sales has spoken to the lead yet.
An SQL, or sales qualified lead, is a lead that a salesperson has spoken to and confirmed. There is a real need, a reason to act within a known time, a budget or a way to buy, and a person who can decide. A form, a score or a download can’t make that judgment.
| Question | MQL | SQL |
|---|---|---|
| Who sets it | Marketing, by written rules | A salesperson, after a conversation |
| What it checks | Fit and a sign of interest | Need, timing, budget or buying process, and who decides |
| When | Usually the same day as the inquiry | Usually within days of the inquiry |
| As a signal for ads | Fast, but easy to inflate | A few days slower, much closer to revenue |
What do MQL and SQL look like on one deal?
For example, a company installs pallet racking in warehouses across Texas. Here is one typical inquiry, from click to contract.
Day 0, 10:40 a.m. · Lead
An operations manager at a food distributor searches for pallet racking installation in Dallas, clicks a Google ad and asks for a quote.
Day 0, 2:15 p.m. · MQL
Marketing checks the form against the agreement: a distributor with 120 employees, inside the service area, asking for a quote rather than a brochure. It’s an MQL. A student who downloaded the brochure the same morning isn’t.
Day 2 · SQL
A salesperson calls. The company opens a second warehouse in March, the budget is approved, and the operations manager will bring the CFO to the site visit. The salesperson marks the lead as an SQL in the CRM.
Day 30 · Site visit and quote
The salesperson sends a quote. The CRM now holds an open deal.
Day 75 · Signed
The contract is signed, 75 days after the click.
The MQL took a few hours, the SQL two days and the contract 75 days. Only the SQL is both close to revenue and quick enough for an ad platform to learn from.
Who sets each mark, and when?
Marketing sets the MQL. The rules are written once and applied the same way to every lead. Aim for the same business day, so sales calls while the person still remembers asking.
Sales sets the SQL. Only someone who has talked to the lead can confirm the need and the timing. Every SQL needs a date. Every rejected lead needs a reason from a short fixed list. Without the reason, marketing can’t tell whether the ads or the follow-up went wrong.
Both marks go into the CRM with their dates. The dates show how fast each team works, and ad platforms accept outcomes only within set time limits.
A one-page agreement for marketing and sales
Put the rules on one page that the heads of both teams approve. Here is our template, filled in with typical entries for the racking company.
| Line | What to agree | Typical entry |
|---|---|---|
| Who we sell to | Industries, size, region, roles | Distributors and manufacturers with 50+ employees in Texas and Oklahoma. Operations and facilities roles |
| MQL | Rule, owner, deadline | Fits the profile and asked for a quote, a site visit or a price. Marketing marks it the same business day |
| First call | How fast sales reaches out | Within four business hours of the MQL |
| SQL | Rule, owner, deadline | A project within six months, a budget or a buying process, and the person who decides. The salesperson marks it within five business days |
| Reasons to reject | A short fixed list | Not a fit, no project, no budget, can’t reach, spam or duplicate |
| Not now | Where these leads go | Back to marketing with a date to follow up |
| Ad platforms | Which mark goes back, how often | The SQL, uploaded daily |
| Review | Who meets, when, on what | Both team heads, 30 minutes a month: MQL to SQL rate by campaign and the top reasons to reject |
Why should ad platforms learn from the SQL, not the MQL?
An ad platform optimizes toward whatever you report as a conversion. Report every form, and it finds more people who fill in forms, including students and job seekers. Report the MQL, and it learns your marketing rules. Report the SQL, and it learns from a salesperson’s judgment after a real conversation.
Only the SQL is both close to revenue and quick enough for an ad platform to learn from.
Why not the signed deal? In B2B it comes late and rarely. In the racking example, it came 75 days after the click. Of the three platforms, only Google’s click ID import can still learn from it that late, and Google’s bidding wants 15 conversions in 30 days. Many B2B companies don’t sign that many deals a month.
The switch has a visible cost. LinkedIn’s help page warns that when you optimize for qualified leads, the cost per qualified lead may fall while the overall cost per lead goes up. That is the trade you want. Judge campaigns by cost per SQL, not cost per lead.
How do you send the SQL to Google, Meta and LinkedIn?
Each platform takes the mark from your CRM in its own way. Here is what their help pages asked for on 16 October 2026.
| Platform | How the SQL gets there | Minimum volume | Time limit |
|---|---|---|---|
| Google Ads | A “qualified lead” or “converted lead” goal, uploaded by enhanced conversions for leads (email, phone or address) or by click ID (GCLID), which Google calls legacy | 15 conversions in 30 days across the account, for bidding | 63 days after the click for enhanced conversions, 90 days by click ID. Ideally an event within 7 days |
| Meta | The Conversion Leads goal, with CRM stages sent through the Conversions API. Works with Meta’s instant forms only | At least 200 leads a month, and the chosen stage reached by 1 to 40% of them | The stage you choose happens within 28 days of the lead. Upload at least once a day |
| The qualified leads goal with Lead Gen Forms, with events from the Conversions API or CRM Sync. Pick one source, not both | Five or more qualified leads within two weeks, recommended | Shared within 30 days |
Source: Google lead goals, Google lead bidding, Google import limits, Meta Conversion Leads, LinkedIn qualified leads, checked 16 October 2026.
How long each platform waits
What to keep in the CRM
- The platform’s ID for the lead. Google’s click ID (GCLID) or Meta’s lead ID, plus the email and phone from the form. Platforms match your upload on these.
- The date and time of each mark. Platforms check them against their time limits.
- A daily upload. Google recommends uploading daily, and Meta asks for at least one upload a day.
One more check for Google. Since 15 June 2026, Google takes these uploads through its Data Manager API, as its help page says. Connections built on the older Google Ads API may have stopped, so make sure your SQLs still arrive.
Not enough SQLs yet?
Many B2B accounts start below these numbers. Then let the platform bid on the lead or the MQL, judge campaigns by SQLs in your CRM and move budget by hand toward the ones that bring them. Keep the SQL as a secondary conversion to watch: Google advises against optimizing to several lead stages at once. Check the volume again each quarter.
Which mistakes break MQL and SQL?
- An MQL built on activity, not fit. Ten email opens from a student aren’t a buyer’s interest. Check fit first, then interest.
- Marks without dates, or set weeks late. A mark set after 28 days is too late for Meta, and every delay slows the platforms’ learning.
- Rejections without a reason. “Bad lead” can’t be fixed. “Outside our region” can, in the targeting.
- One lead counted twice. The same person can arrive through your site form and an instant form, or the same SQL through two LinkedIn sources. Count each person once and send each event from one source.
- Sending only the signed deal. It arrives too late and too rarely for the platforms. Send the SQL and keep the deal for your own reports.
- Rules changed quietly. If sales tightens the SQL rule in March, March won’t compare with February. Write the change and its date into the agreement.
Questions about MQL and SQL
What comes first, MQL or SQL?
The MQL. Marketing marks a lead that fits and shows interest, then sales talks to it and marks it an SQL or rejects it with a reason. A demo or quote request can go to sales at once, but it still gets both marks.
What is a sales accepted lead (SAL)?
An optional step between MQL and SQL. Sales agrees to work the lead before qualifying it. It shows whether leads sit in a queue. If your team calls every new lead within a day, you can skip it.
What is a good MQL to SQL conversion rate?
There is no benchmark you can rely on. Published figures vary widely, because every company defines both marks its own way. Track your own rate by campaign and by month.
Can a small B2B company teach ad platforms with SQLs?
Often, yes. LinkedIn asks for the fewest, five or more qualified leads within two weeks. Google wants 15 conversions in 30 days, and Meta at least 200 leads a month, with 1 to 40% of them reaching the stage you choose. Below that, judge campaigns by SQLs in your CRM.
Is BANT the same as an SQL?
No. BANT stands for budget, authority, need and timing. It’s one checklist sales can use to decide on the SQL. Use the questions that predict a deal in your business and write them into the agreement.
Do we need a CRM for this?
You need one place where every lead, both marks and their dates are recorded. A CRM makes the daily upload to ad platforms much easier. A shared lead log is enough to start judging campaigns by SQLs.
Sources and review
- Primary sources
- Help pages of Google Ads, Meta and LinkedIn, listed below.
- Checked
- 16 October 2026. Platform rules change, so each source carries this date.
- The racking deal
- A typical case written for this guide, not a client.
- What this guide does not claim
- That sending SQLs to a platform guarantees lower costs or more deals.
Sezemov, M. (2026). MQL vs SQL: what counts as a qualified lead. Wonderflow. wonderflow.agency/insights/mql-vs-sql/
