A buying signal is an observable event at an account suggesting it has entered a buying window: a funding round, a hiring surge, a new tool in the stack, a leadership change. Signals answer a question firmographics cannot, which is not who looks like a customer but who changed recently enough to be worth contacting this week.
Why signal-led outbound works when list-led outbound stopped
A list is a claim that an account resembles your customers. That claim was worth something when few teams could make it. It is worth much less now that every data vendor sells the same firmographics to everyone, which means your competitors are contacting exactly the accounts you are, with exactly the same justification, in exactly the same week.
A signal is a claim about timing, and timing is the part that is genuinely scarce. It also solves the opening-line problem that consumes so much of a rep's day. "You raised a Series A eleven days ago and have three sales roles open" is a reason to be in someone's inbox. "You are a 50-person SaaS company" is a filter half your market shares.
The three dimensions of a signal, which are not one number
Most scoring models collapse a signal into a single strength score. That is the fastest way to rank a perfect-fit account with a nine-month-old event above a fresh event at a company you cannot sell to.
Keep them separate:
- Recency. When did the event happen, and when did you find out? Those are two different dates and the gap between them is your real competitive position.
- Fit. Does the account match your ICP on the dimensions your closed-won deals actually cluster on, rather than the ones in your pitch deck?
- Implication. What does the event genuinely say about budget? A funding round says capacity exists. A job post says headcount was approved. A blog mention says someone in marketing wrote a blog post.
Only the third dimension is specific to the signal type, which is why the pages below spend most of their length on it.
What every signal page here has to say out loud
There is no shortage of "40 buying signals that predict pipeline" content. What almost none of it does is state what a signal fails to predict, which is the half a reader actually needs before building a play on it.
So every page in this cluster carries a "what it does not predict" section, and it is enforced by the build rather than left to review. If we cannot write that section honestly for a signal, we do not publish the page.
Freshness is a pair of dates, not an adjective
"Real-time" is the most abused word in this category. Almost nothing in B2B data is real-time; most of it is periodic collection described optimistically.
The honest version is two timestamps. When did the event happen in the world, and when did the system first see it? Signl stores both on every signal, as observed_at and detected_at, so the gap is visible rather than asserted. Current-state matches, where a company simply fits your criteria today, carry no event date at all, because nothing happened. They are a filter result, and treating them as fresh events is how scoring models end up ranking noise first.
The signals worth building a play on
Each page below covers one type: what it is, what it does and does not predict, where the data comes from and how stale it can be, how to score it, which ICPs it matters for, the outreach angle it justifies, and the query that returns it.