MQL vs SQL: the difference, with examples you can use

MQL vs SQL: what’s the difference?

MQL means marketing qualified lead. Marketing has assessed the lead against agreed criteria, usually a combination of customer fit and meaningful interest. It is a decision to pass the lead to sales for follow-up.

SQL means sales qualified lead. Sales has assessed the lead as a potential customer worth pursuing under the company’s qualification rules. A conversation often supplies the evidence, but a completed phone call is not a universal requirement.

For example, HubSpot defines both lifecycle stages through the team that qualifies the lead. It lists both attempted contact and connected contact among lead statuses within SQL. Your company can require a discovery call, but that is your operating rule.

Neither label guarantees revenue. An SQL is also distinct from an opportunity. In HubSpot’s default lifecycle, an opportunity is a contact or company associated with a deal. Other CRMs and teams may use different stage names.

How to qualify a lead: separate fit from intent

Customer fit and buying intent answer different questions. Fit asks whether you can serve this buyer well. Intent asks whether there is a reason to act now.

Your ideal customer profile supplies the first check: industry, location, business size, use case, and meaningful exclusions. A lead from the right company is relevant even if the person is still researching. A pricing request from outside your service area may be urgent but commercially unsuitable.

Useful MQL criteria

Start with information you can actually observe. For a warehouse-racking installer, the lead might need to be in the service area, represent a business with a suitable site, and ask for an installation quote or a site assessment.

A brochure download can be useful evidence of interest. On its own, it says little about whether the reader has a project. Treat the action in context: who took it, what the content addressed, and what else you know.

Lead scoring can help sort these signals. There is no universal score that turns someone into an MQL or SQL. If you award points for repeated browsing, test whether those points identify buyers rather than employees, researchers, or existing customers. HubSpot’s lead-scoring tools distinguish fit and engagement for this reason.

When to use rules, and when to use lead scoring

For this installer, a relevant quote request that passes the service-area and project-fit checks already gives marketing a clear decision. Adding points for the same facts would not improve that route. Send the inquiry for prompt follow-up, then let sales check the SQL criteria.

Scoring can help prioritize known contacts whose interest is less explicit. Here is a small hypothetical model for that separate review queue. A contact must first pass the fit check. Spam, duplicates, student research requests, and unsuitable projects stay outside the queue regardless of activity.

Under these example rules, 15 points puts a contact in the marketing review queue. A guide request plus a pricing-page visit gives 5 + 10 = 15. Twenty more visits still leave the score at 15. Once both signals are older than 30 days, with no fresh qualifying activity, they contribute zero. The weights, window, and review threshold are operating choices for this example, not industry standards.

The reviewer still needs evidence of a relevant quote or assessment request before applying this installer’s MQL rule. Points alone do not create an MQL or SQL. A direct quote request takes the faster rules-based route instead of waiting to collect points.

Useful SQL criteria

Sales needs enough evidence to justify a buying conversation. Agree which questions must be answered at this stage:

• Need: What is the buyer trying to change, and can your offering help?
• Timing: What event or deadline makes this relevant?
• Buying process: Who evaluates, who approves, and how will funding be decided?
• Next step: What has the buyer agreed to do with sales?

For the racking installer, this could mean a suitable warehouse project, a plausible schedule, a known contact in the buying process, and agreement to a site assessment. An approved budget might be required for a quotation stage. It need not be a universal condition for the first sales-qualified stage.

BANT, which stands for budget, authority, need, and timeline, is one qualification framework. Salesforce explains it as a way to assess leads. Use its questions where they help. Requiring a fully approved budget before discussing a project can exclude a buyer who needs your proposal to secure that budget.

Worked example: one inquiry, two CRM snapshots

Consider the warehouse-racking installer. This is a hypothetical buying journey. The record, dates, response times, and decisions below are examples, not a client case.

An operations manager at a food distributor requests an installation quote for a second warehouse in Dallas. Marketing checks the company, site location, and request. Two days later, sales confirms the buying situation. Here is what changes in the same CRM record.

The MQL decision follows the fit check and explicit quote request. No score is needed. The SQL decision follows sales confirmation of a suitable project within the six-month horizon, a credible buying process, and the agreed assessment. An approved budget is not required by this example’s SQL rule. A connected call without that evidence would not be enough.

Download the filled CRM record. It is one lead with both stage dates, not two independent leads. Open it in a spreadsheet and adapt the fields to your CRM. The example uses the UTC−05:00 offset for Dallas on these September dates.

Sales then prepares the assessment and proposal. If a contract is signed on day 75, record that as a separate outcome. For the September conversion example below, WH-1042 contributes once to the MQL count and once to the SQL count because its SQL date is two days after its MQL date.

Three inquiries that need different decisions

In SaaS, product behavior may provide another route. A product qualified lead, or PQL, is identified through meaningful product use and other relevant buying signals. It describes the source of qualification evidence. It does not automatically mean sales has confirmed the buying process. Pocus explains how fit, product usage, and intent inform PQLs.

A marketing-to-sales handoff agreement you can copy

Write down what qualifies, who acts, and how the lead comes back if it is not ready. Both teams should be able to classify the same inquiry and explain their decision.

Below is our agreement format filled in for the hypothetical installer. The response times and six-month horizon are example operating choices, not industry standards.

Download the editable agreement template. It includes these example entries and a column for your own rules.

Before automating the workflow, take a small set of recent inquiries and classify them together. If marketing calls an inquiry qualified and sales rejects it, locate the disputed evidence. A clearer rule is more useful than another status in the CRM.

How to calculate MQL-to-SQL conversion rate

MQL-to-SQL conversion rate = distinct MQLs that became SQLs within the chosen observation window ÷ distinct MQLs in the same starting cohort × 100.

A cohort is simply the group of leads that entered MQL during the same period. Choose an observation window that fits your sales process, then give each cohort the same time to progress.

For example, 120 leads became MQLs in September. Within 30 days of each lead’s MQL date, 30 had become SQLs. The 30-day MQL-to-SQL rate is 30 ÷ 120 × 100 = 25%.

If October’s CRM report shows 50 new SQLs, dividing 50 by September’s 120 MQLs gives a misleading 41.7%. In this illustration, 15 came from older MQL cohorts and five inquiries went directly to sales. Of the other 30, one came from a September MQL but qualified after its own 30-day window, so only 29 count.

WH-1042 became an SQL on September 17. For this illustration, it is the cohort’s only September SQL. The correct numerator includes it: one SQL in September + 29 in October = 30 within the observation window. The month an SQL appears in the CRM is not enough to decide which cohort calculation it belongs to.

For this hypothetical calculation, the final measurement date must be at least October 30 so that September 30 leads have a full 30-day window. An earlier report should show the cohort as still developing.

HubSpot’s journey-reporting documentation also distinguishes skipped stages. Your calculation needs to account for them explicitly, even if your CRM allows those leads to appear in a broader funnel report.

What is a good MQL-to-SQL conversion rate?

There is no universal good rate. The useful comparison is your own earlier cohorts with the same definitions, lead sources and observation window. A demo-request cohort and a webinar-attendee cohort can both be called MQLs while asking very different things of sales. Treat external benchmarks as context until you can establish that they measure the same transition.

A falling rate can mean weaker acquisition. It can also mean a broader MQL definition, slower follow-up, a longer qualification cycle, or a recent cohort that has not matured. Check those alternatives before rewriting campaigns.

Compare cost per MQL and cost per SQL together

Here is a second view of the same hypothetical September cohort. Both campaigns spent $4,000 acquiring their respective leads. All SQLs shown come from those MQL cohorts within the same 30-day window.

Campaign B deserves a closer look despite its higher cost per inquiry. It produced twice as many SQLs from the same spend. Before increasing its budget, check that qualification rules, deal values, sales effort, and eventual win rates are comparable. A cheaper SQL is useful evidence, not a complete business case.

This is also why our B2B PPC approach connects Google and LinkedIn reporting to sales feedback. A platform’s reported conversion needs a clear business meaning before it can guide the budget.

Which lead stage should your ads optimize for?

Pick an outcome that is clearly connected to revenue and that you can report consistently enough for the campaign to learn from. SQL is one candidate. In some businesses, a confirmed appointment, a well-defined qualified lead, or a completed purchase may be a better fit.

Evaluate four things:

1. Meaning. Does reaching this stage help distinguish potential buyers from people who only complete a form?
2. Volume. Does it happen often enough to support useful learning and evaluation?
3. Delay. How long does it take the buyer to reach the stage, and how much longer before the platform receives it?
4. Reliability. Are the definition, identifiers, timestamps, and event counts consistent?

Choose the bidding event for the same business

Now inspect the same fictional installer’s Google Ads reporting. This is a separate rolling 30-day view, not the September MQL cohort used above. The following counts and delays are hypothetical.

CRM stage counts and attributed ad conversions are different views of the same journey. The gap between them can come from attribution rules and timing, not only from broken uploads. So don’t label all 20 MQLs or 15 SQLs in the gap as lost uploads. In this example, the missing identifiers are a separate finding from the export review.

Decision for this example: keep the existing, eligible MQL action as the bidding goal for now. It already represents a suitable quote or assessment request and supplies a more timely, checked signal. Keep SQL outcomes in CRM reporting to judge lead sources and sales quality. The 76-to-3 comparison is not a universal threshold or a claim that MQL always works better.

Before testing SQL as the bidding goal, repair the export, retain matching identifiers with the correct stage timestamp, and reconcile unique accepted, rejected, and attributed events. Then assess a fresh period for this campaign’s eligibility, volume, and delay. Compare lead quality and later sales through the test. Making an event primary is a separate account decision, not a consequence of calling the stage SQL.

Google’s lead-generation guidance recommends considering a conversion action with at least 15 conversions over the previous 30 days at account level. Ideally, the action happens within seven days of the ad interaction, and offline outcomes are uploaded regularly, preferably daily. These are recommendations for choosing and reporting a useful goal. They are not universal technical minimums for automated bidding: Google says Target CPA can start without conversion history.

LinkedIn’s qualified-leads guidance recommends uploading five or more qualified leads within two weeks to help its learning phase. That recommendation applies to the qualified-leads optimization goal using Lead Gen Forms and feedback through Conversions API or CRM Sync. It is not a volume rule for every LinkedIn campaign. Neither platform’s recommendation guarantees profitable results.

My take on small budgets

With a small budget, I would first work out how many SQLs we can realistically expect each month. If it is only a handful, I would test an earlier event that still signals buying intent. That may mean changing the landing-page offer too, such as a project assessment instead of a detailed quote request. The aim is to give the ad system more useful examples to learn from. I would still track which inquiries become SQLs and customers. As the budget grows and SQLs arrive consistently and promptly, I would test using them for bidding. More events are useful only if they help us find buyers.

For a first planning check, divide the planned spend by the observed cost per usable event over a comparable period. Allow for qualification and upload delays. When testing an earlier event or a different offer, keep the customer-fit checks and compare subsequent SQLs and sales. A page view or an easier form completion is not automatically a useful replacement.

Treat a later switch to SQL as a test with time to learn and sales outcomes to compare. Google explains that changing conversion goals or actions can require another one or two conversion cycles for Smart Bidding to adjust. More budget or a higher event count alone does not establish that the switch will improve results.

Avoid counting every step of the same journey as if each were a separate business win. For a straightforward lead-generation setup, choose the intended stage for bidding and retain other stages for observation. Check any value-based or campaign-specific setup against the platform’s current guidance.

The exact goal and conversion settings belong in a Google Ads review based on the account’s real volume and delay. No article can tell you whether your account is ready to switch.

How to send CRM outcomes back to ad platforms

Your CRM stage name and the platform’s conversion category do not have to match word for word. Write down the business event, then map it to the supported integration.

In Google Ads, qualified-lead and converted-lead goals can represent steps in your own conversion process. Google does not require every business to use the same SQL rule.

For LinkedIn’s qualified-leads optimization goal, its documentation explicitly includes MQL and SQL as possible qualified-lead stages. This particular setup uses Lead Gen Forms with Conversions API or CRM Sync. Avoid sending the same outcome through both routes without resolving duplicate counting. Our LinkedIn Ads service covers how targeting, forms, and sales feedback work together.

Keep three different clocks separate

Qualification delay is the time from an inquiry to the chosen CRM stage. Upload delay is how long your system takes to send that event. Attribution lookback is the configured period in which an earlier ad interaction can receive credit. A platform may also impose a maximum age for imported data. These rules answer different questions.

For example, Google’s offline-import guidance describes upload limits of 90 days after the associated last click for offline conversions and 63 days for enhanced conversions for leads. A configured conversion window can be shorter. Meeting an import deadline alone does not prove an event will be attributed or used for a particular bidding goal.

For the qualified-leads goal, LinkedIn requires qualified-lead data to be shared within 30 days for use in ad set optimization. That requirement applies to this Lead Gen Forms setup. It is separate from the attribution lookbacks described in LinkedIn’s conversion-window documentation.

Keep a durable lead ID, source information, relevant matching identifiers, stage timestamps, an event ID, and the upload result. Use the platform’s current data requirements and applicable consent settings. Check accepted and rejected events against the CRM rather than assuming that a running connector means every outcome arrived.

For a new Google integration, check the current offline conversion upload documentation. Restrictions introduced in 2026 affect new users of the older upload API, while some existing integrations retain access. Verify the actual integration before concluding it has stopped.

Common qualification problems and what to check

You can use this review before commissioning more campaigns or a new CRM workflow. If the two teams classify inquiries consistently, respond on time, and can follow the same cohort to sales outcomes, the existing setup may already be sufficient.

Questions about MQL and SQL

Does a booked meeting make a lead an SQL?

Only if it meets your qualification rule. A calendar booking proves an appointment was requested. It does not by itself establish a suitable company, a relevant buying need, or a viable project. Some teams use an accepted meeting as their SQL event, with fit checks built into booking. Others qualify during or after the meeting.

Can a lead skip MQL?

Yes. A referral, an outbound prospect, or a direct sales inquiry may enter sales qualification without a marketing-qualified stage. Define that route and keep it separate when calculating MQL-to-SQL conversion. Do not backfill an MQL date simply to make the funnel look complete.

Can several people from one company be separate SQLs?

They can be separate contacts, but they may belong to one buying project. Choose whether a metric counts people, accounts, or opportunities, and keep that unit consistent. Three contacts discussing one warehouse installation are not automatically three independent sales opportunities.

Is an MQL still useful if we have product-led growth?

It can be, where marketing contributes a meaningful qualification decision. Product usage may supply a PQL route, and direct purchase can bypass a sales-led funnel entirely. Keep stages that change an action or measurement. Adding MQL, SAL, PQL, and SQL to every record is unnecessary if the business does not use those distinctions.

Do we need every qualification field on the first form?

Collect enough to route and respond properly. Ask the questions needed for the next decision at the point you can get a useful answer. A detailed technical project may need a specification. A first product conversation may need much less. Test form changes against downstream lead quality as well as completion rate.

Sources and method

This guide combines primary documentation from HubSpot, Salesforce, Google Ads, LinkedIn, and Pocus with our analysis of qualification and reporting. Sources were checked on October 22, 2026. The warehouse business, CRM record, scoring rules, response times, and numerical examples are hypothetical and were created to explain the decisions.