Intent data does not identify buyers, it identifies companies where someone showed interest in a topic. Three families exist: your own data, public signals, and data bought from a third party. The first two are free and underused. The third is only worth it if you already have the organisation to work the queue, otherwise it produces one more list.
Intent data is information showing that an organisation is researching a topic: the content its people read, the searches they run, the product pages and review sites they visit. It is collected on your own channels, picked up from public sources, or bought from vendors who observe activity across networks of websites. In nearly every case it describes a company, not a person.
Intent data is sold as an answer to the question “who is going to buy”. It is not, and that is the source of most disappointment. What it indicates is that inside a company, someone took an interest in a topic close to yours. That is valuable information, and it is targeting information, not buying intent.
Understanding that distinction completely changes how you choose a source, and whether you should buy one at all.
What the data actually measures
Almost every intent data source works at company level, not person level. It observes aggregated activity, across a domain or a set of addresses, and infers interest from it.
Three practical consequences. You do not know who showed interest: it could be the decision-maker, an intern, or someone preparing a presentation. You do not know how far along the thinking is. And you do not know whether the interest is in you or in your category, which are not remotely equivalent.
That is why intent data is handled as one signal among others, with the same criteria of freshness, scarcity and proximity to the problem that we describe in signal-based selling. A three-week-old intent record is worth far less than a job posting published yesterday.
Family one: your own data
These are the warmest signals, the cheapest, and the ones most often left sleeping in a tool.
Repeat visits to your high-intent pages, a look at your pricing page, a mid-funnel content download, the reopening of a proposal sent two months ago. Each of these events indicates that thinking is under way, at someone who already knows you.
Most companies have this data and do nothing with it, for want of deciding who works it and under what rule. That is the first job, and it costs only a little organisation.
Family two: public signals
Free, accessible, and often more predictive than bought data.
Job postings give away almost every company priority: hiring a dedicated profile for a topic means having decided to invest in it. Funding announcements, leadership changes, press releases and executives’ own posts follow the same logic.
Their limit is volume and manual processing time. Their strength is relative scarcity: few competitors watch job postings, whereas everyone watches funding rounds, which makes the signal far less crowded.
Family three: bought data
It observes activity across a network of specialist sites and flags companies whose activity is rising on your topics.
It brings one thing the other two families do not: coverage of accounts that do not know you yet and produce no public signal. That is real, and it can justify the spend.
It does assume an organisation that is not always there. Someone has to look at the queue every day, a sequence has to be ready to fire, and clean matching to existing accounts has to be in place so you do not contact a current customer. Without those three things, you get an impressive dashboard nobody opens after three weeks.
Intent data providers in France and Europe
Bought intent data comes from a handful of mechanisms, and knowing which one a vendor relies on tells you more than its demo.
Publisher co-ops. Bombora, the best-known example, pools content consumption data from a cooperative of B2B websites, more than 5,000 according to the company, and flags the companies whose reading on a topic climbs above their usual baseline.
Review platforms. G2 sells buyer intent drawn from activity on its own site: companies viewing your profile, your category or comparisons with your competitors. The signal is close to a purchase, but it is only worth as much as your category’s presence on the platform.
Platforms that bundle data and activation. 6sense and Demandbase combine several intent sources with account scoring and advertising. You pay for the platform as much as for the data.
Website visitor identification. Tools such as Dealfront, born in 2022 from the merger of Finland’s Leadfeeder and Germany’s Echobot, identify the companies visiting your site and match them against European company registers. Strictly speaking it is your own data, made readable.
If you sell in France, the most useful public sources are free: the BODACC API for legal announcements (registrations, transfers, mergers, insolvency proceedings), the BOAMP for public tenders, INSEE’s Sirene register for new companies and establishments, and job postings. Coverage is the point to test with any paid vendor: plenty of datasets are richer on English-language content and large US accounts than on mid-sized French companies.
On compliance, company-level data raises few questions. As soon as a vendor supplies named people, the GDPR applies, and you should ask on what legal basis the data was collected and how people can object.
What intent data costs
The large vendors do not publish a price list. They quote on request, usually on annual contracts, and the price depends on the number of accounts tracked, the topics and the features. The most solid public indications come from purchasing platforms that aggregate real contracts. In October 2026, Vendr reported a median of about $25,000 a year for Bombora, across 35 purchases ranging from roughly $13,000 to $80,000, and about $62,000 a year for 6sense, across 385 purchases ranging from roughly $11,000 to $177,000. These are dollar figures, mostly from US buyers, and they move: read them as orders of magnitude, not as a quote.
The subscription is never the full cost. Add the daily time of whoever works the queue, the matching with your CRM, and the sequences to write for each signal. A data contract nobody works costs its full price and returns nothing.
The three questions before signing
What is the real freshness of the signal? Not the dashboard refresh rate, but the delay between the observed activity and the moment it reaches you. Beyond a few days the window closes and you arrive after your competitors.
What is the coverage on YOUR target? A source that is excellent on large American accounts will do nothing for you on mid-sized European companies. Ask for a sample on your own target accounts before signing, not a demonstration on theirs.
How many of your competitors use the same source? If the answer is “all of them”, the signal loses most of its value: the prospect will get five messages the same week, and you will be the third.
The trap of tooling before method
The most frequent mistake is buying data before validating which signals matter in your market. The tool then delivers a volume nobody can qualify, and the team concludes that intent data does not work.
The order that works is the reverse. First analyse what happened at your last twenty won customers in the three months before the first conversation. A pattern almost always appears, and it names the signals to watch. Then buy a source able to supply them, if the first two families are not enough.
That analysis assumes an account history kept properly, which points back to your revenue operations foundation, and it is often on that occasion that people discover the data is missing.
Measuring without being impressed
One test settles it, and it needs none of the vendor’s tooling.
Run two identical sequences for sixty days, one on accounts from the source, one on your usual targeting. Compare reply rates. If the gap is not clear, the source adds nothing, whatever the quality of its interface.
That test costs two months and it avoids an annual subscription renewed out of habit for three years.
Where to start
Look at what you already have. List the accounts that came back to your site more than three times this month, and those where a contact reopened an old proposal. Then call them.
That is intent data, it is free, it is fresher than anything you could buy, and in most companies we open up, nobody is working it.
Frequently asked questions
- What is intent data?
- It is data indicating that a company is interested in a topic: content consumption, searches, activity on specialist sites. It is almost always aggregated at company level, never at person level, which means it tells you where to look, not who to call.
- How much does intent data cost?
- The large vendors quote on request, usually on annual contracts priced by the number of accounts and topics tracked. Purchase data aggregated by Vendr in October 2026 put the median at about $25,000 a year for Bombora and about $62,000 for 6sense, with wide ranges. Your own data and public signals cost nothing beyond the time to work them.
- Is bought intent data worth the price?
- It is worth it only if you already have someone working the queue daily and a sequence ready to fire. Without that organisation it produces a dashboard nobody opens after three weeks, for an annual subscription.
- Where do you start with no budget?
- With your own data and with public signals. Repeat visitors on your high-intent pages, job postings at your target accounts, funding announcements and leadership changes are free, and they are often the most predictive signals.
- How do you measure whether a source is useful?
- Compare the reply rate on accounts from that source with your usual rate, over sixty days and on identical sequences. If the gap is not clear, the source adds nothing, however impressive the vendor's dashboard.