Anyone running outbound through one of the common sales tools knows the promise. Your ICP leaves a digital footprint of its operations: a funding round, a particular hire, a technology change. Each of those says something about what the company is dealing with right now, and current problems create urgency to buy.
The logic is sound. The supply chain is not.
Crawling the web continuously is expensive, so signal data gets batched. Batched data ages. And because the same signal has to be sold to thousands of customers, its definition gets widened until it fits everyone. A GTM hire becomes any open role with sales in the title. Recently funded stretches to anything inside the last two quarters.
By the time the signal reaches your sequence, it is doing very little work.
There is a second cost that gets less attention. Filtering on a signal shrinks the list, often sharply. That trade is worth making when the signal is precise. When it is fuzzy, you pay the full cost of a smaller list and earn none of the precision. Most trigger-based campaigns I have run in the past performed no better than campaigns built on a tight ICP definition and no triggers at all. Several performed worse.
That is not an argument against signals. It is an argument against buying someone else's definition of one.
What has changed is the cost of writing your own. I currently use OpenAI Codex to run periodic scrapes on three signals: GTM hires, a seed to Series B raise inside the last four weeks, and website visits. The output is matched against my ICP, enriched, mapped to the right contact persona, and exported as a lead list ready to process. Once it is set up correctly, it runs on autopilot.
A weekly run may produce only a handful of accounts. That used to look like failure. It is the point. The narrow definition is what makes the small number mean anything.
The shift worth noticing is not that AI makes outbound faster. It is that signal precision has stopped being a budget question. You are no longer choosing between a vendor's broad definition and nothing at all. You can hold a narrow definition, refresh it weekly, and accept a short list, because the cost of maintaining it has collapsed.
What outbound signals does your B2B business use?