Buying signals need context, not automatic triggers.

Intent data has an adoption problem that looks like a technology problem. Teams buy a signal source, wire it to a sequence, and conclude after a quarter that intent data is overhyped or unsuitable for their use case depending on how the quarter went.

One risk is treating a signal as a definitive trigger rather than one piece of evidence. Its quality and permitted use also matter.

Why single signals disappoint

Take funding. A company raises a Series B, so you email them. Reasonable, budget just unlocked. But so did everyone else's reasoning. A funding announcement is public, cheap to monitor, and therefore the most competitive moment in that company's inbox. You're arriving with forty other vendors, all of whom read the same news.

Worse, funding alone says nothing about whether they have your problem. It confirms an announcement, not how resources will be allocated. Those are different things, and conflating them is how you end up with a high-volume sequence that produces polite non-answers.

The same applies to job changes, hiring posts, and tech installs taken individually. Each one is weak evidence. The value appears when they overlap.

What stacking looks like in practice

Instead of "raised funding → sequence," the condition becomes something like: raised in the last two quarters, and posted a role in the function you sell to, and has a competitor's tool installed.

That combination is doing real work. Funding may suggest a change in resources. A role posting may suggest a priority. A detected tool may indicate familiarity with a category, so you don't have to spend the first email explaining why this matters at all.

Each condition cuts the list. Three conditions might take ten thousand accounts to two hundred. Those figures are an illustration, not a benchmark. A manageable segment can make deeper research more practical.

Signal stacking isn't about finding more accounts. It's about earning the right to spend twenty minutes on one.

Not all signals are worth the same

In roughly descending order of what I've found actually predicts a conversation:

• First-party site intent. Someone from the account visited your pricing page. Treat the visit as context, not proof of buying intent or consent. Review the collection and use of visitor information before acting on it.
• Leadership change in the function you sell to. A change in leadership may create a reason to revisit priorities; confirm the person's remit rather than assuming their plans.
• Competitor tool installed. A tool may suggest category familiarity. It does not confirm budget, satisfaction, or a renewal date.
• Relevant role posted recently. Cheap, specific, and a good proxy for "this problem now has an owner."
• Funding. Useful only in combination, for the reasons above.
• Generic "fast-growing company" lists. Not a signal. A description.

The timing question nobody asks

Signals decay, and they decay at very different rates. A visit, leadership change, and tool installation have different useful time horizons. Check freshness and relevance rather than applying a universal expiry date.

Most implementations ignore this entirely and treat every signal as equally fresh, which produces the slightly uncanny experience of being contacted about something you did six weeks ago. If your routing can't distinguish "act now" from "add to the nurture," you're spending the value of the data on latency.

Where this breaks down

Two honest limitations. First, stacking requires enough addressable volume that a narrow condition still yields a workable list, in genuinely small markets, you may only get to two conditions before the list stops being viable.

Second, signals tell you when to reach out, not what to say. I've seen beautifully engineered signal stacks feeding a generic sequence, which is a bit like building a precision timing system to deliver a message nobody wants. The stack earns you attention; the message still has to deserve it.

But when both halves are in place, you have a clearer basis for deciding when a relevant message might be useful. That is a practical use of intent data, just not by wiring one signal to one sequence and hoping.