Intent signals and intent data: definitions and the difference

Vendors sell intent data and intent signals as if they were one thing. One tells you that a named company did something you can see and cite. The other is a score built from what unnamed people at a company read across a network of websites. This page separates the two, explains first-, second- and third-party intent, and shows how a small team can act on intent without drowning in it.

What is an intent signal?

An intent signal is an observable action by a specific company that suggests it is working on a problem you solve. The company posts a job ad for the role your product supports, adds or drops a technology, engages with content on a social network, or visits your pricing page. Each is a fact about one named company, with a date.

The value of an intent signal is that you can explain it. When you write to the company, you can say what you saw. That makes it usable in the first sentence of a message, and it lets the recipient check that the reason is real.

What is intent data, and where does it come from?

Intent data is an aggregated measure of research activity, usually scored per company and per topic. The best-known form comes from publisher co-ops. Bombora, for example, collects content consumption across a cooperative of B2B websites, maps it to companies, and flags a surge when a company's reading on a topic over the last three weeks exceeds its 12-week baseline. You get a company, a topic and a score. You do not get a person or an action.

Forrester's Kahnke calls intent data a broad term covering sources that are not equal. Forrester sorts providers into four models: traditional intent data providers, ABM platforms, campaign execution providers, and walled gardens that only see activity on their own properties. Before buying, ask which model you are paying for and what the score actually counts.

First-, second- and third-party intent: what is the difference?

The difference is who observed the behaviour. First-party intent happens on your own properties. Second-party intent is shared by another site about its own visitors. Third-party intent is collected across many sites by a provider and sold as scores.

Public company events, such as job ads, registry filings and leadership changes, sit outside this grid: they are published facts, not tracked browsing.

The mechanics are documented by the providers and the regulator. Bombora describes Company Surge as content consumption across a publisher co-op, scored against each company's baseline. G2's Buyer Intent documentation lists the page types it reports, such as Pricing, Alternatives and Compare. Forrester's analysis of provider types groups vendors into four models, and its list of common mistakes warns against treating intent as qualification. The EDPB Guidelines 2/2023 bring tracking pixels and similar techniques under Article 5(3) of the ePrivacy Directive, the starting point for any legal review of third-party intent.

Why does tracking everything produce noise?

Because every signal costs time to act on, and time is the scarce part. Leadspicker's guidance is blunt: tracking everything produces noise, and the constraint is capacity, not signal quality. A team that subscribes to fifty intent topics and every job-change alert gets a feed nobody reads. The good signals expire inside it, next to the bad ones.

Aggregated scores add a second layer of noise. A surge on a topic tells you that someone at a company read about it, but not who, not why, and not whether they can buy. Forrester warns that signals age quickly and do not replace qualification. A high score on a company outside your ideal customer profile is still a company outside it.

How do you act on intent without a big team?

Set a signal budget: decide how many signals the team can act on each day, then track only enough signal types to fill it. A worked example for a team of two:

1. Count capacity. Two reps can each give about an hour a day to signalled accounts. At roughly 25 minutes per account for research, the right person and a first sentence that cites the signal, that is about five accounts a day.
2. Pick signal types that fill five a day. Choose two or three close to the buying moment, such as a job ad for the role your product supports and a new executive in the buying function.
3. Filter by fit before timing. Accept signals only from accounts that match the ideal customer profile.
4. Route, do not report. Each signal reaches the rep as a task in the CRM or chat on the same day, with the contact attached.
5. Review after a month. Keep the signal types that produce replies. Drop the rest before adding new ones.

The numbers are an illustration, and yours will differ. The principle does not: the size of the team sets the number of signals, not the size of the data feed.

Frequently asked questions

What is the difference between intent signals and intent data?

An intent signal is a dated action by a named company that you can observe and cite, such as a job ad or a leadership change. Intent data is an aggregated score of research activity per company and topic, usually from a third-party network. Signals tell you what happened; intent data tells you that reading went up.

What are first-, second- and third-party intent?

First-party intent is behaviour on your own properties, such as your website or product. Second-party intent is shared by another site about its own visitors, such as a review platform. Third-party intent is collected across many sites by a provider and sold as scores.

Why does tracking every signal produce noise?

Each signal needs time to research and act on, and most teams can handle only a few a day. Tracking every type fills the feed faster than anyone can work it, so good signals expire next to bad ones. Track only as many signal types as the team can act on daily.