The Complete Guide to Intent Data

Fit data tells you who could buy. Intent data tells you who is buying right now. The two get lumped together as "data," but they answer different questions and decay on different clocks.

A company's headcount and industry stay true for months. A pricing-page visit is true for hours. The value of an intent signal lives almost entirely in how fast you act on it: the same event worked in an hour and worked in a month is not the same signal. Most teams treat intent as one more column to buy and wonder why it never moves pipeline. This guide defines intent data, separates it from fit, and shows where it comes from and how to act on it.

Fit data vs. intent data: what each one actually answers

Intent data and fit data are not two grades of the same thing. They answer different questions and feed different decisions. Fit decides whether an account belongs on your list; intent decides whether you reach out today.

Fit is the slow-moving description of an account: its industry, size, revenue band, region, and tech stack. Intent is the fast-moving evidence that the account is in a buying window right now. You build a target list on fit. You sequence and prioritize that list on intent. Confuse the two and you either spam your whole ICP with no timing, or you chase hot signals from companies you would never want as customers.

Fit is the filter; intent is the trigger. You need both: intent without fit floods reps with in-market companies who will never close, and fit without intent leaves you guessing when to call.

First-party vs. third-party intent data

Intent data comes from two places, and the difference decides how much you can trust it. First-party intent is behavior you observe directly: someone visited your pricing page, downloaded your whitepaper, opened a sequence, or used a feature in your product. You own it, it names a specific company or person, and it is the highest-confidence intent you will ever get.

Third-party intent is behavior observed somewhere else and sold to you: a research panel reports that accounts matching a topic are "surging" across the web. It covers companies that never touched your site, which is its whole value, but it is aggregated, often account-level rather than person-level, and by the time a topic shows up as a surge the account is usually already evaluating several vendors. The practical rule: first-party intent is rarer and hotter, so act on it fast and personally. Third-party intent is broader and cooler, so use it to widen the top of your funnel and prioritize accounts, not to trigger same-day outreach.

That gap, the signals that don't exist in off-the-shelf tools, is where custom first-party and search-based intent earns its keep. Anyone can buy the same topic-surge feed. Almost nobody is monitoring the specific behavior that predicts a purchase in your category, and the strongest programs stack both.

The intent signal types (and what each one tells you)

Not all intent is equal, and the most common mistake is treating every signal as the same heat. A funding round and a second pricing-page visit are both "intent," but they sit at opposite ends of the funnel and demand different plays.

The tiles that rise to the top, repeat pricing-page visits and content engagement, are first-party and late-funnel: small in volume, high in confidence, and worth a same-day, personal touch. The tiles that settle near the bottom, like a topic surge, are broad and early: good for deciding which accounts to warm up, weak as a reason to call today. Funding, hiring, leadership changes, and technographic shifts sit in between, and they double as the buying signals that tell a rep what to say, not just when to say it.

Intent decays: the freshness window is the whole asset

An intent signal has a half-life, and a stale one is worse than no signal because it sends a rep in cold to a moment that already passed. Treating intent data as a static field you enrich once a quarter is the single fastest way to waste it. The question is never just "is this account showing intent," it is "is this signal still inside the window where a play lands."

A pricing-page visit acted on within hours is one of the highest-converting moments in B2B; the same visit surfaced a week later is a cold record. A funding round gives you a few weeks before every other vendor has worked the list. This is why intent data has to run as a standing monitor with an alert, not a quarterly enrichment pass: the value is in catching the event while it is warm and routing it to a rep the same day.

How to source intent data

Intent data has no single source, and that is the point. The teams that win on intent assemble it from several layers rather than buying one feed and calling it done.

First-party intent is already in your stack; you just have to capture it. De-anonymize website visitors, pipe in form fills and content downloads, and pull product-usage events from your warehouse. Third-party intent comes from research panels and topic-surge providers: useful for reach, lower in confidence. The richest layer is the one you build. It runs monitors over hiring pages, funding and news, technographic changes, and review-site activity, plus custom queries that surface the niche behavior only your team knows to watch.

Clay sits across all three. Its Web Intent tracking captures first-party visits: you install a tracking snippet on your site, and Clay de-anonymizes the traffic into named companies. Clay's default Signals monitor the high-value business events out of the box: job changes, new hires, news and fundraising, promotions, and brand mentions. Custom Signals let you watch any digitally accessible source on a set schedule. A careers page or a trust-center change becomes a standing monitor. Each hit is then enriched, so a rep can act without doing the research themselves.

Custom signals are where the edge compounds, because they run on data your competitors have no structured way to watch. A search query you write once becomes a standing intent monitor.

That single prompt turns a vague "they might be growing GTM" hunch into a sourced, rep-ready row, which is the only form of intent data that moves pipeline.

How to act on intent data without burning the signal

Sourcing intent is the easy half. The hard half is acting on it before the window closes, and acting on it with enough context that the outreach does not read as creepy or generic. A raw intent signal is an alert, not an action: "Acme visited pricing twice" is a row a rep still has to research, and that research gap is where most intent data quietly dies. The fix is to attach the work to the signal so the record arrives ready to send.

Routing also depends on how much intent has piled up on one account, not just whether a single event fired. A lone funding round or a single pricing-page visit rarely justifies a same-day call; the same account with three or four signals firing in the same two weeks is a different account, and that convergence is the thing worth routing on.

When a signal fires, the right contact, a verified email, the firmographics, and a researched talking point should already be on the row before a human sees it. Then the signal routes: a high-confidence first-party hit goes straight to a rep with a same-day alert; a broader third-party surge feeds an account-prioritization pass and a warmer nurture rather than a cold call. Intent is also one input, not the whole decision: it feeds lead scoring as the timing axis, and the play it triggers lives in your outbound motion. This compounding effect, signals captured, scored, enriched, and routed automatically, is what lets a small team run an intent-driven motion at the scale of a much larger one.

The lift did not come from buying more intent data. It came from orchestrating the signals Oyster already had into a single motion that reached accounts the team had previously left untouched.

Where to start with intent data

Don't start by buying a feed. Start with the play, then find the signal that should trigger it. Run your last 20 closed-won deals through one question: what observable event happened in the 90 days before each deal opened? Those recurring events, a new exec, fresh funding, a champion who had used you before, a competitor they had outgrown, are your highest-value intent signals, and you already have proof they predict revenue. Stand up monitors for those first, wire each one to a specific play and a refresh cadence that matches its freshness window, and only then widen into broader third-party surge data.

Build the fit side in parallel so intent has something to filter against: define your ICP and your target account list, layer intent on top as the timing signal, and let the two work together. From there, the same machinery scales across B2B prospecting and into building lists by tech stack, where a technographic change is itself an intent signal worth monitoring.