Buying signals: what they are, how to detect them, and how to act

A fit score says whether an account could ever buy: right industry, right size, right tech, right region. It changes slowly. A signal says whether now is a good time. It decays in days or weeks. Teams that add them into one number end up emailing great-fit accounts with no reason to talk, or chasing hot signals at companies that will never buy.

Keep fit and signal in separate fields. Work accounts that are high on both first, nurture high fit with no signal, and mostly ignore signals at poor-fit accounts. This is the core of GTM engineering, and it is how every signal build I run is wired.

Decay windows are rules of thumb. Measure your own: the window closes when replies to that signal drop back to your baseline.

First-party

Events in your own systems: website visits, product usage, form fills, email replies, event check-ins. Highest accuracy, lowest volume, and you already own the data.

Second-party

Public behavior on someone else's platform that you can see directly: LinkedIn engagement, community posts, job postings, review site activity. Good accuracy, and usually where the niche signals live.

Third-party

Aggregated intent from data vendors, built from content consumption across publisher networks. Broad coverage, fuzzy at the account level, and best used as a tiebreaker rather than a trigger.

Every outbound tool watches funding rounds and job changes, so every competitor emails the same accounts the same week. The signals that work best are usually ones nobody packages: a new license in a regulated industry, a project award, a permit, a court filing, a new location.

For Statewise, an EHR for Medicaid care agencies, the buyers are barely on LinkedIn, so I built the account list from public provider registries and the signals from company-level events like new provider numbers, grants, licenses, and EHR changes. For Antimetal, an infrastructure SaaS company, the signal engine watched hires, promotions, job posts, people engaging with competitors, website visitors, and outages. Different markets, completely different signals. More sources are in public records data and developer signal tools.

The signal earns you a reason to write. The message still has to be about them. Name the public context plainly, connect it to a problem they are likely to have because of it, and ask a small question.

Website visits are the exception. Use them to decide who to contact and when, but never say you saw the visit. Write as if the timing were a coincidence.

Speed matters most for fast-decaying signals. A pricing page visit is worth a same-day touch. A funding round can wait a week while you find the right person.

Tag every enrolled lead with the signal that put it there. After a few hundred sends per signal, compare positive reply rate and meetings booked against a no-signal control group from the same fit tier. Signals that do not beat the control go. Signals that do get more sources and more volume.

Expect most of your pipeline to come from two or three signals. That is normal, and it tells you where to spend.

Signal platforms bundle sources and route alerts. Data feeds give you the raw events to build your own. Both are compared, with pricing, in best buying signal tools and intent data providers. If you would rather have the signals picked, sourced, and wired into outbound for you, that is what the signal-based outbound build does.