What MQL vs SQL (vs AQL) meansand when to hand off to sales
What an MQL, SAL, SQL and AQL are, how to score leads into each stage, and the handoff rules that stop good leads going cold.
Marketing says it sent sales 400 qualified leads last month. Sales says it got maybe 20 worth calling. Both are telling the truth, because they mean different things by "qualified".
That gap is what the MQL and SQL labels are supposed to close. Here is what each stage means, how to score leads into them, when to hand off, and how AI agents are adding a new stage in between.
- What is the difference between an MQL and an SQL? An MQL (marketing qualified lead) has shown interest that marketing thinks is worth following up. An SQL (sales qualified lead) has been checked by sales, or against sales criteria, and is ready for a sales conversation.
- Where does SAL fit? A SAL (sales accepted lead) sits between them. Sales has looked at the MQL and agreed to work it, but has not qualified it yet.
- What is an AQL? An AQL (agent qualified lead) is a lead an AI agent has qualified in a conversation, by asking your questions and checking the answers against your criteria.
- When should marketing hand off? Hand off when a lead fits your target customer and has shown buying intent, and agree a response time with sales, ideally minutes, not days.
MQL, SAL, SQL and AQL defined
What is an MQL (marketing qualified lead)
A marketing qualified lead is someone who has engaged enough with your marketing that you think they might buy.
Downloading a guide, attending a webinar, visiting the pricing page three times or opening every email can all push a contact over your MQL line. The key word is *might*. An MQL has shown interest. Nobody has talked to them yet.
What is a SAL (sales accepted lead)
A sales accepted lead is an MQL that sales has reviewed and agreed to follow up. It is a checkpoint, not a judgement on whether the person will buy.
Its job is to stop MQLs piling up in a queue nobody owns. Many teams skip it, and many of those teams also argue about lead quality every month.
What is an SQL (sales qualified lead)
A sales qualified lead has been qualified against sales criteria, usually in a conversation.
You know they have a real need, they fit your target customer, the timing is close enough and the person you're talking to can buy or can bring in the person who does. An SQL is ready for a sales meeting, a demo or a proposal.
What is an AQL (agent qualified lead)
An agent qualified lead is a lead an AI agent has qualified in a live conversation with the buyer. The term is new.
Docket started using it in May 2026 and sells its website agent on turning demand into "Agent Qualified Leads", and a few other vendors have picked it up since.
In practice an AQL is an SQL that software qualified: the agent asks your questions, checks the answers against your criteria and books the meeting.
It isn't an industry standard yet, so treat it like your other stages. Write down exactly what an AI agent has to confirm before a lead counts as an AQL, and get sales to agree to it.
MQL vs SQL vs SAL vs AQL side by side
| Compare | MQL | SAL | SQL | AQL |
|---|---|---|---|---|
| Who decides | Marketing, usually by score | Sales, by reviewing the MQL | Sales, after qualifying | An AI agent, using your criteria |
| Based on | Engagement and fit data | A quick review | A conversation about need, fit and timing | A conversation about need, fit and timing |
| Has anyone talked to them | ||||
| Typical next step | Nurture or pass to sales | First outreach from sales | Meeting, demo or proposal | A booked meeting with sales |
| Main risk | Interest without intent | Sits in a queue | Slow follow-up after qualifying | Criteria set too loosely |
The real difference between MQL and SQL is a conversation. An MQL is a guess based on what someone did. An SQL is a decision based on what they said.
How to score leads into MQL and SQL
Most teams use two kinds of signals. Fit tells you whether this is the kind of customer you want. Intent tells you whether they want something now. You need both. A perfect-fit company that downloaded one ebook a year ago isn't ready, and a keen student isn't a customer.
| Signal | Type | Example points |
|---|---|---|
| Company size or industry matches your target customer | Fit | +15 |
| Job title can buy or influence the purchase | Fit | +10 |
| Visited the pricing page | Intent | +15 |
| Requested a demo or a quote | Intent | +30 |
| Attended a webinar | Intent | +5 |
| Downloaded a guide | Intent | +5 |
| Personal email, student or competitor | Fit | -20 |
| No activity for 30 days | Intent | -10 |
The point values are an example, not a standard. Pick a threshold, say 40 points, for MQL. Then check it against reality: pull last quarter's closed deals and see what score they had when they first became MQLs. If most of your customers scored 25, your threshold is wrong.
SQL is not a higher score. It's a qualification step. A popular way to run it is with a short set of questions. BANT (budget, authority, need, timing) is the classic one. Many teams now lead with need and timing and treat budget as a later conversation.
When to hand off from marketing to sales
Most MQL vs SQL arguments are really about the handoff. Write these rules down and get both teams to agree:
- Agree the MQL definition
The fit and intent signals, and the score that makes a lead an MQL. One page, signed off by both teams.
- Set a response time
Agree how fast sales replies to an MQL, and how fast hand-raisers hear back.
In a Harvard Business Review study of 1.25 million leads, firms that tried to reach a lead within an hour were nearly seven times as likely to qualify it as firms that waited even an hour longer.
- Make acceptance explicit
Sales accepts or rejects each MQL, and gives a reason for each rejection. Rejected leads go back to nurture, not to the bin.
- Agree the SQL definition
The questions that have to be answered before a lead counts as sales qualified, and who answers them.
- Review it every month
Look at how many MQLs became SQLs and how many SQLs became customers, by source. Move the thresholds when the numbers say so.
Common MQL and SQL mistakes
- Counting MQLs as the goal. Marketing gets rewarded for volume, so the threshold drifts down. Forrester's benchmarks put the typical inquiry-to-close rate in an MQL-led process at under 1%. Track MQL to SQL and SQL to customer instead.
- Scoring only on activity. Lots of clicks from the wrong kind of company still isn't a lead.
- Letting hand-raisers wait. A demo request treated like any other MQL sits in the same queue as an ebook download. It shouldn't.
- No feedback loop. If sales never says why it rejected a lead, marketing can't fix anything.
- Qualifying by email ping-pong. Three emails to find out someone's timeline takes a week. One conversation takes three minutes.
How an AI agent turns an MQL into a booked SQL
The slowest part of the funnel is usually the gap between "interested" and "qualified". An MQL waits for an SDR, the SDR emails, the lead replies two days later and the meeting lands next week, if it lands at all.
An AI agent closes that gap by having the qualifying conversation the moment the lead shows up. On your website, it answers their questions, asks the questions your SDR would ask and, if the answers fit, books the meeting on the right rep's calendar.
The lead goes from MQL to SQL, and onto the calendar, in one conversation. Leads who aren't ready get a follow-up later instead of a sales call they didn't want.
Flora, the agent on this page, runs on Flowxperia. Ask her something about pricing, and notice how she answers and then asks a question back.
On your site, the same agent asks your qualifying questions and books the meeting.
Flowxperia's agent does this on your website by voice or text, and on WhatsApp and Instagram for leads from ads and social.
Every answer goes to HubSpot, Salesforce, Zoho, Pipedrive, Dynamics 365 or Zendesk with the full conversation, so sales can see why the lead was qualified.
See how it works as an AI SDR for inbound leads, or read our guide to the best AI agents for lead qualification.
See Flowxperia ask your qualifying questions on your own site and book the meeting with the right rep. Try it free: 20 minutes on the website and 7 days on WhatsApp and Instagram.
Harsh Shah
MQL vs SQL questions
The difference between MQL and SQL is qualification. An MQL has shown interest through your marketing, while an SQL has been qualified against sales criteria, usually in a conversation, and is ready to talk to sales.
MQL stands for marketing qualified lead. It is a contact whose fit and engagement suggest they could become a customer.
SQL stands for sales qualified lead. It is a lead that sales, or an agent working to sales criteria, has confirmed has a real need, fits your target customer and is ready for a sales conversation.
MQL vs SQL vs SAL describes three stages in order. An MQL is flagged by marketing, a SAL is an MQL that sales has accepted to work, and an SQL is a lead sales has qualified and is ready for a meeting.
An agent qualified lead, or AQL, is a lead that an AI agent has qualified in a conversation by asking your questions and checking the answers against your criteria, usually before booking a meeting.
Sales should follow up on an MQL within the hour, and hand-raisers like demo requests within minutes. The odds of qualifying a lead fall quickly once it has waited.
Yes, an AI agent can qualify leads. It asks your qualifying questions in a conversation, checks the answers against your criteria and books qualified leads straight into a sales meeting.
Turn MQLs into booked meetings
An AI agent that answers every lead in seconds, asks your qualifying questions and books the meeting, day or night.