Signal-Based Demand Generation Is Quietly Retiring the MQL

The vendor a buying group already favoured before any seller got involved won 82% of deals. That number comes from 6sense's 2026 B2B Buyer Experience Report, which examined nearly 3,600 completed purchases. If it holds in your market, the preference deciding most of your revenue forms during a stretch of the process you can't see, and the form fill is a receipt, not a lead.

The analysis, published on MarTech Cube, argues that signal-based demand generation is replacing the MQL playbook wholesale. The reasoning is worth a CMO's attention.

What was the MQL actually measuring?

To be fair, the MQL solved a real problem. Marketing insisted the leads were good, sales insisted they weren't, and a threshold ended the argument. What it measured, honestly, was the border between two departments rather than a buyer's intentions.

It also matched how buying worked in 2010, when the analyst report sat behind a form and the email address was the toll. Nobody behaves that way now. A buying group asks former colleagues in a group chat, reads the relevant subreddit, skims review sites and watches twenty minutes of a conference talk at one-and-a-half speed. None of it leaves a trace in your systems.

Worse, the people who still fill in forms are the junior, the curious, the student and the consultant researching for somebody else. The executive holding a $600,000 budget and a September board deadline keeps her head down until she's ready to talk.

Then Goodhart's law finished the job. Once MQL volume became the number marketing was judged on quarterly, thresholds drifted downward. The chart rose for years. Pipeline didn't move.

How do signals differ from scores?

An MQL is a running total, and totals carry no timing information. Eighty points gathered across nine months looks identical to eighty earned on one wet Tuesday. A signal describes an event, and events go stale. Three people at one account opening your integration documentation inside a week tells you something specific. So does a competitor's customer reading your migration page twice.

Each means a great deal in the week it happens and almost nothing three weeks later. A signal nobody can act on within days is just an alert you've paid for.

What does intent data actually deliver?

Less than the pitch deck suggests. First-party signals are worth building on: someone reading the pricing page four times in a fortnight, or nine minutes spent inside the API reference. Most organisations already collect this and do nothing with it, because it sits in an analytics tool nobody connected to the CRM. Fixing that connection is the cheapest improvement available.

Third-party intent data covers more ground and deserves to be held more loosely. A surge of interest sometimes reflects six people assembling a business case; sometimes it's a graduate student or an IP range attributed to the wrong parent company. Treat the two as different instruments — first-party justifies picking up a phone, third-party suits advertising. Teams that collapse the distinction spend a quarter cold-calling strangers.

What's the catch?

The benefits are real: timing, account-level tracking for committees of six to twelve, and measurement that's harder to game. Nobody can manufacture three engineers at a target account reading API documentation on a Thursday.

But the MQL threshold was also the line deciding whose problem an account was. Remove it and two teams must agree who responds to which signal, how fast, and what happens when both see the same alert at 4:40 on a Friday. That negotiation is harder than selecting a vendor, and it typically gets postponed until month four.

The verdict: buying intent data takes about a week. Getting two teams to agree on who moves takes far longer. The second piece decides whether any of it produces revenue.