What are signals?

An intent signal is a public, observable change at a company that suggests it may be moving toward a purchase.

A company posts nine engineering roles in three weeks. Its competitor's G2 page fills up with one-star reviews. It quietly swaps one analytics tool for another. It raises a Series B. Its CTO turns up on a podcast talking about scaling problems.

Each of those is a clue on its own. Together, they build a case: this company is in motion, and right now is a better time to reach out than three months ago was.

Traditional outbound is built around who. You define a title, an industry, a company size, and you message everyone who matches. The list is static, everyone on it gets contacted whether or not anything is happening, and once it's exhausted you build another one.

Intent signals shift the question to when. The audience definition still matters, but it becomes a filter rather than the whole strategy. Instead of messaging every VP of Engineering at every mid-size SaaS company, you message the ones whose companies just did something that suggests they need what you sell.

The difference shows up in your opening line. "I saw you're hiring across your platform team" lands differently than "I noticed you're a VP of Engineering," because the first one is about them and the second one is about a filter you applied.

Alsona monitors 31 signal sources across nine categories.

You choose which ones an agent listens for. The full breakdown of each source, and what to configure inside it, is in "Choosing your intent signals."

Not every signal carries the same weight, and the strongest ones are the ones causally connected to what you sell.

If you sell developer onboarding tools, a company hiring six engineers is a strong signal, because the problem you solve is created by the thing you detected. If you sell to finance teams, that same hiring spike tells you very little.

Funding rounds are the instructive case. Almost every vendor treats a raise as a buying signal, which means a company gets buried the week it announces one. A raise is strong when your product is specifically what companies buy after raising, and weak when you're competing with two hundred other emails that spotted the same headline.

Multiple signals compound. A company that raised money, changed its tech stack, and started hiring in the same month is showing far more than a company that did one of those. This is exactly what the opportunity score measures: it weighs live intent against how well a company fits your profile, and it ranks prospects accordingly instead of treating every detection as equal.

The practical test for choosing signals: look at your last ten closed deals and ask what those companies were doing in the weeks before they bought. That answer is usually narrower and less obvious than the list you'd assemble from theory.

Everything Alsona detects is something a company published, posted, filed, or made visible. Job listings, websites, public social activity, review sites, press coverage, public code repositories, regulatory filings. There's no private data involved, and nothing here reveals anything a company hasn't chosen to show.

That's worth knowing when you write. The best signal-led messages reference something the prospect knows is public, which is why they read as informed rather than intrusive.

A signal is a reason to reach out and a subject to open with. It isn't a script.

The messages that work connect the observation to a problem you solve, and leave room for the prospect to correct you. "I saw you're hiring across your platform team, which usually means onboarding gets painful around month two" gives them something to agree or disagree with. Asserting that they must be evaluating tools right now gives them nothing to do except decide you're wrong.

You can set rules for this in Additional outreach instructions, including how directly the agent should reference the signal that surfaced someone, and whether to name a source or describe it in plain language.

Open any agent and the Intent Signals tab is the first thing you see. Four counters run across the top: signals detected, high-intent companies, identified prospects, and prospects pushed to outreach. Below that, a breakdown by signal source shows how each of the nine categories is performing, including reply rate.

That breakdown table is the most useful thing on the screen. Nobody can tell you in advance which signal types turn into conversations for your particular offer. After a few weeks, that table tells you plainly, and you can drop the categories producing volume without replies.

One thing that surprises people: the counters don't divide into each other. A thousand signals doesn't mean a thousand prospects, because a signal attaches to a company and a prospect is a person, and the relationship between them runs several ways at once. "Signals and prospects" covers that in full.