There’s a long tradition in marketing of naming things more aspirationally than accurately. Sometimes that works out; sometimes it sets us back. “Intent data” is one of those terms that did more harm than good. It promised to reveal which buyers were ready to buy, but in reality, it mostly told us who was digitally loafing on our website.
It’s time to retire the phrase and replace it with a better, more actionable model. One that reflects the dynamic, probabilistic nature of buyer behavior. One that distinguishes historical information from fleeting clues. One that helps marketers move from data mining to signal sensing.
From Data to Signal: A Mental Model That Actually Works
Let’s start here: data is what you store; signals are what you act on.
Data is a record. It sits in your CRM or warehouse waiting to be queried. A signal, by contrast, is a momentary indicator. It surfaces in real time and should trigger an action or decision. If it doesn’t, it wasn’t a signal. It was just noise.
Think of it this way: signals are to butterflies as data are to caterpillars — but their lifecycle runs in reverse.
A signal flutters into view, vivid and alive, demanding your attention. Once it’s captured and pinned to the board (or the database or chrysalis), it becomes data — useful, sure, but no longer in motion. If you wait until it’s data to act, you’ve missed the flight.
This difference is more than semantics. Treating signals like data turns marketing into a warehouse operation. Treating data like signals turns noise into false alarms. Both are mistakes. The right approach is to recognize that signals are ephemeral. They must be sensed, sometimes created, and always acted upon. After they’ve been acted upon, then, yes, stored for analysis.
But their primary value is in the moment.
Why “Signal” Is the Right Word
The term “signal” matters because it frames the job differently. Signals are dynamic. They’re probabilistic. They can be strong or weak, clear or noisy. They require interpretation. And they demand action.
Data feels static. It’s something you put in a lake or warehouse and mine. But signals require you to build a sensor array—to engineer your assets, content, and systems to capture and interpret real-time buyer behavior.
Every form, every CTA, every webinar should act as a sensor. Not just collecting clicks, but enabling you to model what’s happening across the buying group.
The Three Signal Types, Two Sources:
Which, when, why? Profile, Readiness, Interest
Forget the tired taxonomy of first-, second-, and third-party data.
It invites unproductive debate about data provenance and distracts from what really matters: What does this signal help me decide? What action does it trigger?
Here’s a better framework, based on function:
- Profile Signals These tell you whether a company is ever likely to be a customer. They include firmographics (size, industry, location), technographics (what’s in their stack), and buying center details (who’s involved, what roles matter). Think of Profile signals as your topographic map. They don’t change often, but you can’t plan without them.
- Readiness Signals These tell you whether now is a good time to engage. Are they growing? Downsizing? Raising money? Hiring a new head of IT? These dynamic conditions matter far more than most marketers realize. They’re the difference between shouting into the void and showing up at exactly the right moment. If Profile is the terrain, Readiness is the weather.
- Interest Signals This is where most so-called “intent data” lives. These are breadcrumbs left behind by buyers as they research: site visits, content consumption, ad engagement, form fills, event attendance. Interest signals aren’t inherently strong, but in patterns and clusters — especially across buying groups — they become much more meaningful.
A single form fill is but a cloud in the distant sky. A pattern of activity across multiple individuals? That’s a storm system.
The Two Signal Acquisition Paths: Created/Received (CR) vs. Acquired
Another way to simplify the discussion is to ignore n-party-based data categories and instead ask: did we create or receive the signal in our own systems, or did we acquire it from elsewhere?
- Created/Received Signals come from your own digital properties and systems. Web traffic, product usage, email engagement, CRM data, and more.
- Acquired Signals come from outside providers. They give you visibility into activity you wouldn’t otherwise see: content syndication, review sites, publisher networks, etc.
This distinction matters because you have more control over CR signals (and how they’re sensed). But both types are essential. Don’t let dogma about data purity keep you from acting on valid, well-modeled information.
And keep in mind that you can create new signals by combining or apply signals that you’ve acquired or received. If you combine the lead scores from two or more leads into a buying group score, you are creating a buying group signal.
One Is the Loneliest, and Least Productive Number
Perhaps the single most important but least understood (tempted to say, respected) aspect of buying signals is that none are reliable on their own. Even when someone fills in a form to say, Call Me!, that signal does not lead to a sale in the majority of cases. It often doesn’t even result in a conversation.
So, it’s critical to understand that we need to create and acquire multiple signals that point to each buying process. The more signals we can apply to the problem, the better off we’ll be.
I still hear marketer after marketer asking for, begging for the silver bullet signal. They simply don’t exist. The silver bullet is created when AI is applied to many signals. Some signals reinforce each other. Some point in different directions. Together, they are a much more reliable signal than any one on its own.
The Industrial vs. Post-Industrial Angle
I talk frequently about how B2B marketing has been industrialized — how it’s become an MQL assembly line, where the buyer is a commodity to be acquired and processed in its final form.
This industrial model is wrong, not least because buyers hate being treated that way. In the industrial marketing model, data is king. It gets collected, stored, and optimized through repeatable, factory-like processes. That seemed ok when sellers controlled access to information and we thought buyer journey was linear.
But, buyers operate more like dynamic weather systems than products that are being assembled. They form and reform buying groups. They change direction mid-cycle. They research anonymously and at odd hours. You can’t just batch-process that. You need to sense it.
Moving from industrial to post-industrial marketing means shifting from MQL assembly lines to real-time sensing and response. From static programs to probabilistic models. From fix funnels to dynamic patterns.
Final Thought: Be a Forecaster, Not a Miner
Buying groups emit literally hundreds of signals on vendor websites over the course of a buying journey. None of the individual signals are reliable indicators of buying journeys. Your MQL isn’t. Your PQL(product qualified lead) isn’t.
But if you have multiple MQLs, PQLs, and acquired buying signals, then you can create (or have created for you), much more robust, reliable signals of buyer intent (now we’re actually approaching that).
And we’ll do a better job of responding and reacting to changing buying behaviors and technologies if we think more clearly about how to detect them. And that means thinking in terms of signals, not data.
