From MQL to DQL: lead qualification as a state, not a status

The criticism of the marketing qualified lead (MQL) has substance, but abandoning it leaves no useful replacement, while defending it unchanged ignores a real problem. A qualification decision records what the evidence supports at a particular time. If the project changes, a stakeholder leaves or the evidence becomes stale, the decision needs to be reviewed.

LeadScale’s hypothesis is to treat qualification as a continuously updated state across the buying group. It calls this the dynamically qualified lead, or DQL.

An MQL is a lead that has crossed a threshold, such as a score, a set of behaviours or a demo request, and been passed to sales as ready for follow-up. The weakness appears when a programme keeps the qualified status without checking whether its supporting evidence still holds.

What a marketing qualified lead assumes

An MQL threshold helps a team prioritise when it has more inbound demand than sales can handle. Its usefulness depends on the criteria and on whether the team reviews them as circumstances change.

An MQL is a snapshot. It says that at a particular moment, and on the strength of particular signals, one contact looked ready. Each part of that is contingent, from the moment it was granted to the single contact it rests on. The moment passes, the signals age, and the one contact is rarely the whole story. A programme that treats that snapshot as a standing fact can act on outdated evidence. A lead that qualified in March, on a whitepaper download by a mid-level analyst, is filed as qualified in June even if the project has since been shelved, the analyst has moved teams, and the budget has gone elsewhere. Without a review or expiry rule, the qualified status can remain even though its basis has changed.

Why a one-time stamp no longer fits the buyer

Gartner’s research on the B2B buying journey describes a loop rather than a sequence of stages. Buyers work through six buying jobs: problem identification, solution exploration, requirements building, supplier selection, validation and building consensus. They revisit each job at least once rather than completing them in order. A form submission captures one moment in that process.

The limited time with sellers and the involvement of buying groups deepen the mismatch. Gartner reports that buyers spend only around 17% of their total purchasing time meeting with potential vendors, and that time is split across all the vendors they are considering, so any single supplier sees a sliver of the process. Forrester’s May 2021 release reports its B2B Buying Study’s finding that more than 80% of purchases involve complex buying scenarios. These include consensus scenarios, where 95% involve three or more people across two or more departments, and committee scenarios, where strategic purchases involve multiple people and departments across the organisation. Both involve groups. These findings support the B2B Revenue Waterfall’s shift from the individual lead to the buying group. Forrester sells advisory services around that framework, so read its framing with that interest in mind; the buying-group finding comes from its B2B Buying Study rather than a sales claim.

A qualification granted once, on one person, at one moment fits a looping, largely invisible, group-run process poorly. Static lifecycle markers were not wrong when buying was more linear. They fit less well now: a one-time stamp discards the intent signals that would let a model keep up with a buyer whose situation changes.

Why “qualified” means different things

Teams also use different definitions of “qualified”, which makes comparison difficult.

The marketing-analytics firm marqeu, looking at MQL-to-opportunity conversion, found that the reported rate for the same funnel stage ranges anywhere from 13% to 45%, and that the spread is not measurement noise. It comes from the fact that “MQL” is defined differently across scoring models. A team that calls any high-intent form-fill an MQL ends up with a small, high-quality pool that converts north of 35%. A team that calls any lead crossing a behavioural score an MQL ends up with a larger pool converting closer to 13 to 15%. The industry median sits somewhere around 15 to 21%, but that average hides the fact that the same three letters describe several incompatible things. marqeu sells scoring and waterfall consulting, so treat its specific numbers as vendor-published rather than audited; the useful part is the range and the reason for it.

Different standards make the same label describe different pools of leads. A 2026 Demand Gen Report opinion argues that the MQL is dead because the industry has optimised for volume and qualifies poorly. That describes a symptom of fragmented definitions, rather than a new problem. Whatever label the programme uses, sales and marketing need a shared definition before they can compare performance.

The DQL hypothesis: qualification as a changing state

In a May 2026 piece, LeadScale’s Robin Caller named this concept the dynamically qualified lead.

The proposal is to reassess qualification as account and stakeholder signals change. If an account’s intent surges on a Friday, its qualification would rise with it; if a stakeholder’s behaviour shifts on Monday, the read would recalibrate. This is the changing state that DQL is intended to describe.

The idea leans on a composable architecture, where data moves freely between the data warehouse and the activation layer, so that the qualification could in principle be recalculated at the edge of each interaction rather than fixed at a form-fill. Robin pairs it with what he calls differential activation: designing the response to a buying group in real time from the signals as they arrive, rather than following a journey mapped six months earlier. Robin also pairs the idea with a learning loop, using AI less to write copy than to work out which actions actually moved an account and to adjust accordingly. The retrospective nature of lead scoring, accumulating activity that has already happened, is part of what the dynamic frame reacts against.

DQL is a hypothesis, not an available product or a deployed mechanism shown to outperform an MQL. The hypothesis follows logically from the buying evidence: a changing read of a group decision is at least the right shape for the problem. It changes the mental model before it changes the tooling.

Qualification needs to fall as well as rise

A programme that only adds points or escalates follow-up can miss evidence that an opportunity is weakening. In Robin’s framing, the best next action is often to decline, detarget, or suppress: a falling state should carry a decision to stop pursuing demand that no longer justifies the effort.

Robin’s proposal puts that decision to stop at the centre of qualification. In that framing, a group going quiet or a record becoming invalid is a signal to stop spending, rather than a gap to fill with more follow-up. Knowing when to stop matters as much as knowing whom to pursue. This connects with LeadScale’s Zero Waste principle: avoid continued spending on demand that is no longer worth pursuing.

Accurate data and re-validation would be necessary inputs. Whether qualification can be connected to data truth and validation at source in a working system remains an open question.

What an operator can apply now

The practical discipline does not depend on waiting for a new product:

• Re-check a lead that qualified a month ago before assuming its status still holds.
• Review evidence that weakens qualification as well as evidence that strengthens it.
• Define when to pause, exclude or suppress an account rather than escalating automatically.
• Keep qualification criteria current and agreed between sales and marketing: the fragmentation marqeu describes starts at home, and a shared, current definition is worth more than a sophisticated model reading an outdated one.

This also gives sales and marketing a basis for resolving disagreements over lead quality: a quality judgement that is re-read as evidence changes is harder to dispute than one asserted once.

At capture, the LeadScale Engine applies the compliance, truth and value checks purchased for the programme. For qualification to stay current, the dynamic hypothesis would require those conditions to be re-checked as circumstances change.

Whether the full DQL arrives as Robin describes it or in some other form, the direction is worth watching, and the change in how you think about qualification is worth making now.