MQL vs SAL vs SQL vs PQL: Which Stages You Need

August 1, 2026

Four qualification stages exist. Most organisations need three, some need two, and adding all four without defining the transitions makes reporting worse rather than better.

Four acronyms describe stages between an anonymous visitor and real pipeline. They are frequently adopted together because a framework recommended them, and the result is four stages nobody can define and conversion rates nobody trusts.

Most organisations need three. Some need one.

What each stage actually claims

The four stages by type of claim

Note that three are opinions held by people and one is measured behaviour.

StageWho asserts itType of claimFailure it reveals
MQLMarketingAssertion — worth attentionDefinition has drifted
SALSalesAcknowledgement — will work itNobody picked it up
SQLSalesVerdict — genuine pipelineQualification was wrong
PQLThe productObservation — demonstrated useNone; it is evidence

Source: Framing proposed here

The difference in kind matters. A PQL is the strongest signal available because the user has already experienced value rather than expressed interest in it, and no scoring model approximates that.

Which set to run

  • Self-serve product: MQL, PQL, SQL. PQL replaces most of what MQL scoring attempts to infer.

  • Sales-led with high inbound volume: MQL, SAL, SQL. The SAL stage earns its place by separating untouched leads from rejected ones.

  • Outbound-led, low inbound volume: SQL only. Marketing qualification stages measure a process that is not happening, and adding them creates empty funnels and arguments about why.

The rule for adding a stage

Add one only when a different team owns each side of the transition, or when it separates two failure modes needing different responses. Otherwise it is reporting overhead.

The structural cause of the argument

Where marketing sets the qualification threshold and is measured on the volume clearing it, the threshold drifts downward. Not through dishonesty — through the ordinary accumulation of edge cases, each of which seemed reasonable to include at the time.

The fix is separation of concerns: whoever is measured on volume should not solely own the definition. Joint ownership with sales, or ownership by revenue operations, removes the mechanism rather than appealing to good faith.

Method: testing whether your definitions mean anything

  1. Take fifty recent records that crossed your MQL threshold.

  2. Have one person from marketing and one from sales classify each independently against the written definition.

  3. Measure the disagreement rate. Anything above a small minority means the definition is ambiguous rather than contested.

  4. For every disagreement, record which criterion was read differently. Those criteria are the ones to rewrite.

  5. Add a rejection reason code if none exists. Without a way to reject and a reason, there is no feedback loop and the definition cannot self-correct.

  6. Re-run quarterly. Definitions drift, and the disagreement rate is the earliest detector.

Where this fails

In account-based motions the lead is the wrong unit entirely. Several people from one company each engaging moderately is a stronger signal than one person engaging heavily, but lead-level qualification reports the first case as several mediocre leads and misses the account. Where that describes your motion, qualify at account level and treat individual lead stages as supporting detail.

COMMON QUESTIONS

What is the difference between an MQL and an SQL?
An MQL is a claim by marketing that a lead warrants attention. An SQL is a verdict by sales that it is genuine pipeline. Collapsing them removes the only checkpoint at which a drifting definition can be detected.
Do we need a sales accepted lead stage?
It earns its place when you need to separate leads nobody worked from leads that were worked and rejected. Those are an operational failure and a definitional failure, and they need different fixes.
What is a product qualified lead?
A user whose product behaviour indicates readiness — reaching a usage threshold, hitting a limit, adopting a key capability. It differs in kind from the others because it is observed rather than asserted.
How do we know if our lead stages are well defined?
Have marketing and sales independently classify the same fifty records. The disagreement rate measures how much of your lead-quality argument is definitional rather than substantive.

KEY TERMS

Go deeper: RevOps 101: Revenue Operations Foundations

The interactive tools behind this writing — process builders, inventories and the audit export — live inside the membership.

See the course