A buying signal is an observable change inside an account that makes the problem you solve urgent for a specific person, right now. The useful ones are events you can see from outside — a leadership hire, a funding round, a job posting, a public complaint, a tech-stack change — and each one points at both a person and a reason to contact them this week rather than next quarter.
The signals worth building a workflow around:
- Leadership change in the function that owns your problem
- Hiring activity that describes the pain in the job description
- Funding or a new mandate that creates pressure to deliver
- Tech-stack change — adopting a complement, or dropping a competitor
- Public commentary where someone states the problem in their own words
The rest of this guide covers what separates a real signal from noise, where each type is found, and how to convert one into a message that gets answered.
What counts as a buying signal?
A buying signal has three properties. It is observable from outside the company, it is recent enough that the situation is still unresolved, and it attaches to a person who now has a reason to care.
Drop any one of those and it stops being useful. A company that fits your ICP perfectly is not a signal, because nothing changed. A funding round from fourteen months ago is not a signal, because the money is already allocated. A signal you cannot tie to a named human is market research, not a prospecting trigger.
That third property is where most teams lose the plot. “Account is hiring three SDRs” is a fact about a company. “The new VP of Sales who started nine weeks ago is hiring three SDRs and has no way to tell them who to call” is a signal, because it names a person, a pressure, and a gap.
Intent data vs trigger events
These get sold as the same thing and behave completely differently.
Intent data is inferred. A vendor observes that people at a company read content about your category and infers interest. It is probabilistic, usually anonymised to the company level, and it tells you that someone somewhere in a 2,000-person organisation looked at something. It is useful for prioritising a list and close to useless for writing a message, because you cannot reference it without sounding like you are watching them.
Trigger events are observed. A person was hired. A round was announced. A role was posted. A competitor was dropped from the stack. These are facts, attached to names, that you can reference openly because the company published them.
Intent data answers which accounts to look at. Trigger events answer who to contact and what to say. Teams that buy the first and skip the second end up with a better-sorted list and the same generic email, which is why the intent budget so often fails to show up in booked meetings.
The five signal types that actually predict a deal
Leadership change. Someone new took over the function that owns your problem. This is the strongest single signal in B2B, because new leaders are evaluated on visible change and have a short window — roughly their first two quarters — in which changing tools is expected rather than suspicious. Settled leaders defend the status quo they built; new ones are looking for reasons to replace it.
Hiring activity. Job descriptions are the most honest documents a company publishes, because they are written to attract someone who can fix a specific problem. A posting that lists responsibilities matching your product’s job is a company telling you, in writing, that they have a gap and have budgeted headcount against it. Volume matters too: a team going from two reps to eight has an onboarding and process problem it did not have last quarter.
Funding or a new mandate. Money creates deadlines. A round, a new product line, a market expansion, or a reorg all put named people under pressure to deliver something specific on a timeline. The signal is not the money itself — it is that someone now owns an outcome they cannot hit with their current setup.
Tech-stack change. Adopting a tool adjacent to yours means they have accepted the category and are actively building in that direction. Dropping a competitor is better still: they have a budget line, an internal advocate who owns the problem, and recent evidence that the current answer is not working.
Public commentary. A post, a talk, a podcast, or a conference question where someone describes your problem in their own words is the highest-conversion signal available, and the least used. It gives you the person, the pain, and their exact language for it, which is most of a good first message already written.
Where to find each signal
Most of this is public and free, which is why the constraint is time rather than access.
LinkedIn covers leadership change and public commentary — filter by time in role to surface recent moves, and read activity feeds rather than profiles, because the feed tells you what someone is thinking about now. Company careers pages and job boards cover hiring, and the requisition usually names the hiring manager, who is often your champion. Funding and mandates show up in press releases, the company blog, and earnings calls for public companies. Tech-stack changes surface in job descriptions (the required-tools list), engineering blogs, and integration or partner pages.
The pattern across all of these: the company already published the signal. Nobody needs a data vendor to find out that a VP started last month. What they need is the time to check consistently across a few hundred accounts, which is exactly the part that does not scale by hand.
How to turn a signal into an outbound angle
A signal is not a message. The mistake is opening with the observation — “I saw you raised a Series B” — which tells the recipient you have an alert set up and nothing else. Everyone with the same alert sent that email the same morning.
The structure that works:
- Name the consequence, not the event. Not “congrats on the round” but “hiring six reps in a quarter usually breaks whoever owns territory assignment.”
- Connect it to their job specifically. The consequence has to be one they own. A VP of Sales and a Head of RevOps feel the same funding round differently.
- Make one concrete claim. What changes if this is solved, stated in a number or a specific outcome, not an adjective.
- Ask for something small. A signal earns you a reply, not a purchase order. Ask a question they can answer in one line.
The test: if you could send the same message to a company with a different signal, the signal is decorative. It should be structurally impossible to reuse.
How to avoid acting on noise
Not every change is a signal, and chasing all of them is how teams end up busy and unproductive. Three filters:
Is it recent enough to be unresolved? Signals decay. A leadership hire is strongest in weeks four to twenty — early enough that the new person is still forming their plan, late enough that they have stopped drinking from the firehose. A funding round is strongest in the first two quarters. Past that window, the decision has usually been made without you.
Does it attach to a person you can reach? A signal about a company with no identifiable owner for the problem is not actionable. Park it.
Would this matter if you sold nothing? This is the honest check. If a reasonable person inside that account would agree the change creates a real problem for them, it is a signal. If the connection only makes sense from where you are standing, it is a pretext, and it will read as one.
Signals also compound. One is a reason to reach out; two or three in the same account within a quarter is a strong indication that something is genuinely in motion, and those accounts deserve disproportionate attention.
How to run a weekly signal review
Signal-led prospecting only works if it is a habit rather than a burst of enthusiasm after a good quarter. The teams that sustain it run a short recurring review with a fixed shape.
Set a standing hour at the start of the week. Work a defined account set — the segment you own, not the entire market — and check the same five signal types in the same order every time, because consistency is what makes the absence of a signal informative. Note what changed since last week and nothing else; the review is for detection, not for writing messages.
Then triage into three buckets. Act now is a fresh signal with a named person and a clear consequence, and it gets outreach that day. Watch is a signal that is real but early — a role posted but not filled, a reorg announced but not staffed — and it gets a date to re-check. Discard is everything that failed the noise filters above, and it matters that you discard it explicitly rather than leaving it on a list to guilt you.
Two rules keep the habit alive. Cap the act-now bucket at a number you can genuinely follow up on in a week, because an unworked signal is worse than an unfound one — it decays while sitting in your queue. And record which signal types actually converted, so after a quarter you know whether leadership changes or hiring activity produce meetings in your market, rather than in general. Most teams find one or two types carry the bulk of the results, and that finding is worth more than any vendor’s list.
How Alfa fits
Signal-led prospecting fails for a boring reason: it works, and it does not scale by hand. Checking a few hundred accounts every week for five signal types, tying each one to the right person, and drafting an angle that references it specifically is more research than any rep can sustain alongside actually selling. So the discipline survives for the top ten accounts and collapses everywhere else.
Alfa starts from what you sell and who you sell to, then watches market movement continuously and turns it into a live stream of accounts, the likely champion inside each, and the reason they may care now. The research step that used to gate the workflow runs in the background instead.
The judgment stays with the rep. You still decide which signals matter for your market and what to send. Alfa removes the part where finding them is the bottleneck.
Signals only pay off inside a defined set of accounts. Our similar companies lists show what that set looks like for a named company, including which candidates were dropped because they had been acquired or had shut down.
Once you have a signal, the next question is who inside the account to point it at — that is how to find internal champions. And if your signals are firing across accounts that were never a fit to begin with, the problem is upstream: start with how to define your ICP.