The Shelf Life of a Buying Signal: How Long Intent Data Stays Useful

Most outbound teams treat a buying signal like a lead. It goes on a list, and the list gets worked when someone gets to it. That is the wrong mental model. A signal is not a record. It is a timestamp attached to a claim about a company's situation, and that claim gets less true every week it sits in a queue.

The practical question is how much less true, and how fast. Search for an answer and you will find confident numbers that contradict each other. One article says intent signals lose half their value in seven days. Another says forty-eight hours. Another says forty-five days. None of them cite a study, and none of them distinguish between a pricing page visit and a regulatory filing.

So instead of one decay curve, this piece gives you a way to reason about the shelf life of any signal you find, a tiering model for the signal types B2B teams actually work with, and a short check to run before you send.

What Signal Decay Actually Means

Signal decay is the loss of predictive value in a buying signal as time passes between the moment the signal appeared and the moment you act on it. A signal does not expire on a fixed date. It loses accuracy as the situation that produced it moves on without you.

Three things cause that loss. The observable event resolves, so the hook stops being current. The priority behind the event gets absorbed into a project that is already underway, so you arrive after the requirements were written. Or the person you selected changes roles, so the message lands with someone who no longer owns the problem.

A fourth cause gets less attention. Public signals are public, so the value of a fresh one depends partly on how many competitors watching the same source already used it. That is one reason unstructured signals requiring interpretation tend to stay useful longer than signals delivered pre-packaged in a feed.

Why the Decay Numbers You Read Online Do Not Hold Up

The specific decay statistics circulating in intent data content are not measurements. They are estimates that got repeated until they sounded like findings.

There is real research adjacent to this question, but it measures something narrower. A Harvard Business Review study of 2,241 US companies found that the average response time to a web-generated lead, among firms that responded at all within 30 days, was 42 hours. A companion analysis of 1.25 million leads found that firms contacting a prospect within an hour were roughly seven times more likely to qualify that lead than firms that waited an hour longer.

Those findings describe inbound form fills, where someone just raised their hand and is comparing options in that session. A job posting is not that. Applying an inbound response curve to a public unstructured signal is a category error, and it sits underneath most of the forty-eight-hour advice.

Decay is also not a smooth downward line. Gartner's research on the B2B buying journey describes buying as nonlinear, with buying groups looping back through the same jobs more than once, and finds that 99% of B2B purchases are driven by organizational changes. Organizational change plays out over quarters. A signal tied to it can sit dormant for a month and become relevant again when the group loops back to requirements.

Every Signal Has Three Clocks, Not One

Before you decide whether a signal is stale, separate three things that age at different rates: the event, the priority, and the person.

The event clock measures how long the observable fact stays current. A funding announcement is news the day it publishes and old news a month later. This clock is usually the shortest.

The priority clock measures how long the business priority behind the event stays open. A round that closed in March is being spent through the rest of the year. A company that opened three revenue operations roles in the spring is still building that function in the fall. This clock is usually the longest, and it is the one that determines whether your argument still holds.

The person clock measures whether the contact you selected still owns the problem. This one decays independently of the other two and invalidates outreach completely when it runs out.

Most teams track only the event clock. They see the announcement stop trending, conclude the signal is dead, and drop it. What they lost was the hook, not the argument, and that distinction is the difference between discarding a signal and repositioning it.

A Shelf-Life Model for Common Buying Signals

Signals fall into four rough tiers based on how quickly the event behind them resolves. Use the tier to decide how fast you need to move, not whether the account is worth contacting.

Tier One: Hours to Days

These are reactive events with a built-in deadline. Requests for proposal and procurement notices, competitor complaints in public communities, a live pricing question in a forum. If a date is printed on the signal, that date is the shelf life. Miss it and the window closes rather than narrows.

Tier Two: One to Four Weeks

These are announcement-driven events. Funding rounds, product launches, new market or location pages, the start of a paid ad campaign. Regulatory filings sit here too, with a useful property: public companies generally have four business days to report a triggering event on Form 8-K, so a filing is close to real time when it appears and thoroughly picked over within a couple of weeks.

A funding round is the clearest example of the split between clocks. Tier two hook, tier four priority. The announcement stops being a reason to reach out quickly. What the money is being spent on does not.

Tier Three: One to Two Quarters

These are capability-building signals. Open roles, growth concentrated in a specific function, technology stack changes, leadership changes. A role stays open for weeks, and the work it was opened to do continues long after the seat is filled. A new executive is usually measured against a plan they set in their first quarter, which makes months two through five more useful than week one.

Tier Four: Structural

These reflect a durable condition rather than an event. Sustained headcount growth in a function, workforce tenure, technology fit, long-running category presence. They barely decay. They also say almost nothing about timing, which is why they belong in lead scoring rather than in the first line of a message.

The working rule that falls out of this: use your fastest-decaying signal as the hook and your slowest as the argument. A team that leads with tier four language sounds generic. A team that leads with tier one urgency and has nothing behind it sounds opportunistic.

Four Questions to Ask Before You Send

Run these before a signal-based message goes out, especially if the signal has been sitting in a list.

1. Is the event still verifiable today? Reopen the source. Job postings get pulled, pages get reverted, announcements get amended.
2. Has anything happened since that changes the interpretation? A funding round followed by a restructuring is a different message than a funding round alone.
3. Does the person still hold the role and the remit? Title changes and reorganizations quietly invalidate a well-researched message.
4. Would this still make sense if the prospect assumed I found the signal today? If the answer is no, the signal is being asked to do work it can no longer support.

The fourth question is the one that catches most stale outreach. Prospects are good at spotting a message that is pretending to be timely.

What to Do With a Signal That Has Aged Out

An aged signal is not worthless. It gets demoted from the hook to the context.

Here is what leading with a stale event looks like:

Congratulations on your Series B back in February. Wanted to see if you were looking at outbound tooling this quarter.

The date is doing all the work, and the date is old. The reader knows the announcement was public months ago, which tells them the message was assembled from a list rather than from any understanding of their situation. There is no argument underneath the greeting.

Here is the same underlying signal used as context:

You have had three revenue operations and lifecycle roles open since the spring, and your careers page now lists retention as its own function. That usually means expansion revenue moved from a side project to somebody's number. If that is roughly right, the piece most teams underestimate is how much account context the team needs before those conversations get productive. Worth fifteen minutes?

The second message never claims recency it cannot support. It uses an accumulated pattern to make a claim about a priority, frames that claim as a hypothesis rather than a fact, and gives the reader an easy way to correct it. That is the shape most signal-based sequences should take once the event is more than a few weeks old. You are reading public evidence, not internal plans, and prospects respond better to a stated hypothesis than to a confident guess about their priorities.

Why Signal Decay Is an Operations Problem, Not a Data Problem

Most teams do not act late because they misjudged a signal's shelf life. They act late because nobody was watching the clock.

The typical manual process makes this unavoidable. Someone reviews job boards and news on a Tuesday, exports what they find, hands it to whoever writes the copy, and the campaign launches the following week. By send time the freshest signal is ten days old and the oldest is a month. Nothing in that workflow records when each signal appeared, so nothing in it can tell you which messages went out too late.

Alsona is built around monitoring rather than list building. The platform tracks over 30 intent signals across hiring, advertising, technology, funding, filings, reviews, social activity, and company news, and many are defined as change rather than state: hiring velocity, social activity surges, review velocity spikes, recent ad activity, technology stack changes. A signal defined as a change carries its own timestamp, which is what makes decay measurable at all. From there the platform researches the account and turns that context into individualized LinkedIn and email messaging, with follow-up that stays tied to the original signal rather than restarting from a generic template. A follow-up that ignores the signal behind the first message resets the conversation to cold and wastes whatever freshness that touch had.

The Takeaway

Shelf life is not a property of intent data in general. It is a property of the specific event, the priority underneath it, and the person you picked. Track those three clocks separately, move fast on tier one and tier two signals, use tier three and tier four to build the argument, and stop treating a signal list as a backlog that will still be accurate next month.

Build outbound around signals you can actually date. See how Alsona turns monitored buying signals into timely LinkedIn and email outreach.

Frequently Asked Questions

How long does a buying signal stay useful?

It depends on the signal type. Deadline-bound signals like procurement notices are useful for hours or days. Announcement signals like funding rounds work as a hook for one to four weeks. Hiring and technology signals stay relevant for one to two quarters because the work behind them continues after the event.

Do intent signals really lose 50% of their value in a week?

No published study supports that figure, and the numbers circulating online contradict each other. The closest real research measures inbound lead response time, a different situation from a public unstructured signal. Reason about decay by signal type instead of applying a single blanket statistic.

Which buying signals decay the slowest?

Structural signals decay slowest: sustained headcount growth in a function, workforce tenure, technology fit, and long-running category presence. The tradeoff is that they say very little about timing, so they are better used for scoring and prioritization than as the reason for reaching out now.

Is it worth contacting a company if the signal is a few months old?

Often yes, but the message has to change. Move the event out of the opening line and use it as supporting context for a claim about the company's current priority. Leading with a months-old announcement signals that the message came off a list.

How do you keep signals from going stale before outreach goes out?

Shorten the distance between detection and send. That means continuous monitoring instead of periodic list pulls, recording the date each signal appeared, prioritizing by tier rather than account size alone, and drafting messages when the signal is detected rather than in a weekly batch.