Signal-based prospecting content is full of advice about which signals to watch. Almost none of it says how long each one stays worth acting on, which is the part that decides whether the play works.
The spread is enormous. Some signals are dead in 48 hours. Others carry for three months. Running them all through the same weekly batch means you are too late for half and needlessly early for the rest.
Here is the decay table, and what changes at each stage.
The decay table
| Signal | Peak window | Still useful | Dead after | Check frequency needed |
|---|---|---|---|---|
| Post engagement | 0–48 hours | 1 week | 2 weeks | Every few hours |
| Topic/keyword post | 0–48 hours | 1 week | 1–2 weeks | Every few hours |
| Competitor audience engagement | 0–72 hours | 2 weeks | 1 month | Every few hours |
| New job posting | 1–2 weeks | 6 weeks | When the hire starts | Daily |
| Hiring surge | 2–4 weeks | 2 months | 3 months | Daily |
| Job change | 2–8 weeks | 90 days | 4–6 months | Weekly, daily for priority |
| Promotion | 3–12 weeks | 90 days, soft edge | 6 months | Weekly |
| Headcount growth | 1–2 months | 3 months | 6 months | Weekly |
| New match on a saved search | Immediate | Indefinite | Does not decay | Daily |
That last row is the exception worth noticing: a new person matching your ICP criteria is not a timed signal at all. They matched and they still match. Everything above it is a window.
Why the windows differ
Three things determine how fast a signal decays.
How public it is. A LinkedIn post is seen by thousands, some of them your competitors running the same play. Engagement signals decay fast partly because everyone else is also acting on them. A promotion is visible but far less harvested.
Whether the underlying state persists. A job change creates a condition (new budget, new mandate, no incumbent relationships) that lasts months. A post reflects a passing thought.
Whether a decision is being made. A hiring surge signals a process underway with a natural end. Once the hire starts, the window closes because the problem is nominally solved.
What to do at each stage
Inside the peak window. Reference the signal directly. “Saw you’re building out RevOps” is natural when the posting went up last week.
Past peak, still useful. Reference the underlying situation, not the event. Not “I saw your post last month” but “most teams scaling outbound run into this.” The signal told you who to contact and roughly why; it no longer justifies mentioning it.
Past the useful window. Treat them as an ordinary cold lead. The signal is now just a targeting input, which is still worth something, they matched your criteria once, but it buys you no opening.
That progression is the practical point. A signal does not become worthless, it becomes a targeting input rather than a conversation starter.
Matching check frequency to decay
The failure mode is checking slower than the signal decays.
A 48-hour signal checked once daily loses roughly half its value, you are contacting people on day two of a two-day window. Checked weekly, it is worthless.
Conversely, checking a 90-day signal every six hours wastes nothing but produces no benefit.
Sensible frequencies:
- Engagement and topic signals: every few hours
- Job postings and new search matches: daily
- Job changes, promotions, headcount: weekly, with a priority daily subset for your highest-value contacts
A priority tier matters here. If you monitor 5,000 contacts weekly, a job change on your most valuable 100 can wait six days, unless those hundred are checked daily.
Stacking changes the maths
Two signals on the same account do not add, they multiply, and the combined window is set by the shorter one.
A company posting three SDR roles (six-week window) whose VP of Sales also just posted about pipeline problems (48-hour window) is a much stronger prospect than either alone, and you have 48 hours, not six weeks.
So when signals stack, act on the shortest clock.
What a decayed signal is still worth
Do not discard them. A contact who surfaced through a signal six months ago is still someone who, at one point, matched a condition worth noticing.
Three uses:
Re-watch them. Someone who engaged with a competitor’s content once will likely do it again. Keep them in a monitored list rather than a dead one.
Use them as ICP evidence. If signal-sourced leads from one category convert better, that tells you something about your ICP that your written definition may have missed.
Nurture rather than sequence. A decayed signal is a reason to stay visible, not a reason to send a five-step cadence.
The suppression problem
Fast-decaying signals create a specific risk: the same person surfacing repeatedly.
Someone active on LinkedIn may engage with three competitor posts in a fortnight. Without account-level suppression, that is three enrolments in three sequences, and they will notice.
Rules worth setting:
- One signal enrolment per person per 30 days, regardless of how many signals fire
- Account-level suppression once anyone at the company replies
- A total-touch ceiling across all campaigns, not per campaign
The short version
Signals decay at wildly different rates. Post engagement is dead in two weeks; a job change carries for three months. Running them through the same weekly batch means being late for the fast ones and over-eager on the slow ones.
Match check frequency to decay rate, hours for engagement, daily for job posts, weekly for job changes with a priority tier on your best contacts.
Reference the signal directly inside its peak window, the underlying situation after that, and nothing at all once it is dead.
When signals stack, the shorter clock wins. And suppress at the account level, because the fastest-decaying signals are the ones most likely to surface the same person twice.
Next: signal-based prospecting end to end, or job change signals.
Frequently asked questions
How long is a buying signal valid?
It depends entirely on type. LinkedIn post engagement is dead within two weeks, job postings carry about six weeks, and job changes and promotions stay useful for around 90 days.
How often should you check for buying signals?
Match the check to the decay. Engagement and topic signals need checking every few hours, job postings daily, and job changes or promotions weekly with a daily subset for high-value contacts.
What do you do with an expired signal?
Treat the person as an ordinary cold lead, the signal becomes a targeting input rather than an opening. Keep them in a monitored list, since contacts who surfaced once often surface again.
What happens when two signals fire on the same account?
They strengthen the case, but the shorter window governs timing. A fast-decaying signal alongside a slow one means you have days, not weeks.