The thesis
Intent is the new demographic. The most valuable commercial opportunities are often visible before the market names them, in what people are doing rather than in who they are. The constraint was never that intent did not exist. It was that intent was hard to observe at scale. That constraint is lifting, and it changes where advantage comes from.
What I am watching
Intent is becoming more valuable than demographics
Who someone is, their age, title, company size, predicts far less about what they will buy than what they are actively doing right now. As intent gets cheaper to observe, spending against identity looks more and more like spending against an average that describes no real buyer.
Buying signals appear before buyers self-identify
By the time someone fills in a form or asks for a demo, the opportunity is already obvious and already crowded. The interesting signal is earlier: the research, the comparison, the quiet behaviour that reveals a problem before the person has named it as a purchase.
Weak signals are now detectable
AI is good at exactly the thing humans are bad at: noticing faint, distributed patterns across a lot of public behaviour. That makes it possible to detect intent and opportunity earlier, across industries and customers, while the signal is still weak enough that almost no one is acting on it.
Study signal, not noise
The discipline is not collecting more data. It is learning which faint patterns actually precede demand and which are just movement. The operators who win are the ones who detect early and act before the signal becomes obvious enough for everyone to bid on.
Practical implications
- Ask what a prospective customer is doing that reveals a problem worth solving, then meet them there with something useful rather than something promotional.
- Treat the moment of intent, not the demographic profile, as the unit you build around.
- Build the habit of looking for opportunities before they are obvious. By the time they are obvious, the margin is gone.
Current questions
- Which early signals reliably precede demand, and which are noise dressed up as insight?
- Where is the line between reading public intent and crossing into surveillance that erodes the trust you are trying to earn?
- How small can an operator be and still detect weak signals usefully?