Intent Data in EMEA: What Providers Can and Can't See

Most intent-data demos look great. The example accounts are large, English-speaking and heavy on online research. The heatmap glows.

Then you load your own target list. A few hundred mid-sized companies in Germany, the Netherlands and Poland. Half of them barely show up.

That's not necessarily a bad vendor. It's the method. Every kind of intent data sees a specific slice of buyer behaviour, and in EMEA some of those slices are thin.

This guide is about methods, not vendors. Where I name a vendor, it's as a well-known example of a method, using only what that vendor says in its own documentation. I'm not ranking anyone, and nothing here says how well a specific product covers your market. That's what the two-week test at the end is for.

What "intent data" actually means

Intent data is evidence that someone at a company is researching a topic, a category or a vendor. That's it.

It comes in two kinds:

• Third-party intent. Research that happens on other people's websites: publisher networks, review platforms, ad exchanges.
• First-party intent. Research that happens on your own channels: your website, your webinars, your emails.

Vendors often bundle a third kind of data next to it: hiring and technology data. Strictly speaking, that's not intent. It's context, or triggers. I've included it because it's sold in the same packages and it often carries more weight in EMEA than topic intent does.

One thing to keep in mind for every method below. Intent tells you someone is looking. It rarely tells you why. That's why I treat it as one layer in a stack, never as the reason on its own. The model I use to combine layers is in Signal-based ABM in EMEA.

1. Publisher co-op and bidstream topic intent

How it works

A network of B2B publishers and websites shares content consumption data. A tag on each site records which topics readers engage with. The activity is mapped to a company, usually via the visitor's IP address and other identity data, and compared against that company's normal baseline. When a company reads much more about a topic than usual, it "surges".

Bombora is the commonly cited example of the co-op model. According to its own site, it has direct relationships with co-op members and places a proprietary JavaScript tag on their websites, and it tracks more than 25,300 intent topics [1]. Bombora also states that it doesn't use bidstream data [2].

Bidstream intent is the other variant. It's derived from the data that flows through programmatic ad auctions when a page loads an ad: the page URL, the device, the IP address. Some providers model topic interest from that stream.

What it can see

Topic-level research across many sites, aggregated per company. At its best, it shows a company reading about a problem before it visits any vendor.

EMEA blind spots

• Language. A co-op sees the sites that are in it. If your buyers research in German, French, Polish or Arabic on local trade media, the question is whether those publishers are in the network and whether the topic model understands the language.
• Mid-sized private companies. A company with 300 employees produces far less measurable research than one with 30,000. Surge models compare a company to its own baseline. Thin data means a noisy baseline.
• Remote work and IP mapping. Mapping a reader to a company gets harder when people work from home on residential connections.
• Consent at the source. In the EU, reading or storing information on a visitor's device needs consent unless it's strictly necessary (ePrivacy Directive, Art. 5(3)) [3]. So the quality of the consent chain on publisher sites matters. For bidstream data, the CJEU held in 2024 that the consent string used across much of European ad tech (the TC String) can be personal data [4]. That puts the whole chain under GDPR scrutiny.
• Small-company identifiability. Company-level intent for a five-person firm can point at one person. Then it's personal data.

What to ask the vendor

• Which publishers in my countries and languages are in the network? Can you share a list or at least counts per country and language?
• How do you map activity to a company, and what happens with home-office and mobile traffic?
• How is consent collected on the source sites, and can you document it?
• Do you use bidstream data anywhere in the product?
• Is there a minimum company size or activity threshold before you report a surge?

2. Review-site category research

How it works

A software review platform sees which companies browse which categories, products and comparisons. It sells that activity back to the vendors listed there.

G2 is the best-known example. Its documentation lists signal types such as views of your product profile, your pricing page, alternatives pages, category pages and comparison pages that include your product [5]. G2 says it collects Buyer Intent data at the account level, so a signal doesn't mean a specific contact generated it [5]. It reports a Buying Stage and an Activity Level per company, updated daily, and which signals you receive depends on your plan [6]. Its documentation also says the data includes activity from Capterra, Software Advice and GetApp [6].

What it can see

Late-stage software research. Someone comparing your product with a competitor, or reading about alternatives to the tool they already have, is usually further along than someone reading a general article.

EMEA blind spots

• Only software, only listed categories. If you sell services, hardware or something without a review category, there's nothing to see.
• Only that platform. It sees research on its own sites. Buyers who ask peers, a consultant, a local IT partner or an industry association leave no trace there. In my experience that's common in the German Mittelstand and in relationship-driven markets like the Gulf.
• Language and local habits. Whether buyers in your markets use English-language review sites for your category is a question worth checking with your own customers, not assuming.
• Low volume at mid-sized companies. Fewer researchers means fewer signals. A quiet month may just mean nobody logged in.

What to ask the vendor

• How many companies in my countries showed activity in my category over the last 90 days?
• How do you identify the company behind a visit?
• Which signal types are in my plan, and which aren't?
• Is there a minimum threshold before a small company is reported?

3. Reverse-IP website visitor identification

How it works

A script on your website collects the visitor's IP address and matches it against a database of IP ranges known to belong to companies. The output is a list of companies that visited, which pages they viewed and when.

Leadfeeder is a well-known example in Europe. The company traded as Dealfront from 2023, after Leadfeeder and Echobot merged, and moved back to the Leadfeeder brand in 2026 [7]. Its documentation says the tracker matches visitor IPs against its IP-to-company databases, sets cookies such as _lfa (one year), and supports a consent manager mode where the visitor's consent status is stored [8]. Leadfeeder's own blog says an IP address can be personal data under GDPR, which is why it still involves personal data at the point it's processed [9].

What it can see

Companies visiting your own site. Which pages, how often, how recently. That's your own audience, so the relevance is high.

EMEA blind spots

• Consent. In the EU, a script that stores or reads information on the visitor's device needs consent unless strictly necessary [3]. The CJEU confirmed in Planet49 that this applies whether or not the information is personal data [10]. The EDPB goes further in its guidelines: even tracking based on the IP address alone can fall under the same rule [11]. So in practice, you only see the visitors who accept. In Germany that's §25 TDDDG, covered in GDPR and buying signals.
• Home offices and mobile networks. A residential or mobile IP doesn't belong to a company. Those visits stay anonymous or get misattributed.
• Shared and proxied traffic. Companies that route traffic through cloud security proxies or VPNs can show up as the proxy provider, or not at all.
• Small companies and sole traders. If a company is one person, identifying the company identifies the person.
• It only sees your site. It tells you nothing about companies that haven't found you yet.

What to ask the vendor

• How does the script behave before consent and after a "no"? Does anything load?
• What happens to the raw IP, and how long is it kept? Is there IP truncation?
• How are residential, mobile and proxy IPs handled?
• Are you a processor or a controller for this data? Is there a DPA?
• Do you ever match a visit to a named person? If yes, on which legal basis?

4. Job-post and technographic data

How it works

Job-post data comes from crawling job boards and company career pages, then classifying roles, functions, seniority and the tools named in the ad.

Technographic data comes from scanning public website code for traces of the tools a company runs: analytics, marketing automation, chat widgets, e-commerce platforms. Some providers add other sources on top.

What it can see

Hiring is one of the clearest triggers there is. A first hire in a function, a hiring wave, or an ad that says "migration to" a new system. Technographics show which tools are visible on the website and when they change.

EMEA blind spots

• Local job boards. A lot of mid-sized hiring in EMEA happens on StepStone, Welcome to the Jungle, Pracuj.pl, InfoJobs, Bayt or the company's own career page, not on LinkedIn. Coverage depends on which of those a provider crawls.
• Language. "Kreditorenbuchhaltung", "contrôleur de gestion" and "specjalista ds. fakturowania" need to be classified correctly to be useful. Ask how titles in your languages are mapped.
• Back-office tools are invisible. Website scans see what runs on the website. ERP, finance, HR and most internal systems don't leave traces there. If you sell to finance or operations, technographics will miss most of what matters.
• Reposted and evergreen ads. A role that's been "new" for six months isn't a signal. Check how duplicates and reposts are handled.
• Personal data in ads. Job ads often name a recruiter or hiring manager. Ask whether those names are stripped.

What to ask the vendor

• Which job boards and career pages do you crawl in my countries?
• How do you classify job titles in German, French, Polish or Arabic?
• How do you detect reposts?
• Which technologies can you detect, and from which sources besides public website code?
• How fresh is the data? When was a typical company in my list last checked?

5. First-party intent

How it works

Your own data. Website visits at company level, webinar sign-ups and attendance, demo requests that never got booked, newsletter replies, event conversations, product usage if you have a free tier.

What it can see

Companies that already know you and are spending time with you. It's the most reliable layer, and the one most small teams never set up properly.

EMEA blind spots

• Consent banners. Everything that runs through a script depends on consent in the EU. Your data shows the visitors who accepted, not all visitors.
• Email tracking. In France, the CNIL's 2026 recommendation says measuring email opens to optimise campaigns needs consent. Details in GDPR and buying signals.
• Volume. A small team in a niche market may get a handful of meaningful visits a week. That's still useful, but it won't fill a pipeline.
• Scattered data. Website, webinar tool, newsletter, event lists and someone's inbox. If it isn't joined at company level, nobody sees the pattern.

What to ask yourself

• Can I see company-level engagement across all my channels in one place?
• Which pages actually signal buying for my offer: pricing, integrations, implementation, comparison?
• Is consent set up correctly in every country I sell into?

The five methods at a glance

What "EMEA coverage" in a sales deck can mean

"We cover EMEA" can mean a lot of things. A database that contains company records for most countries. A network that has some publishers in each region. A product that's sold to customers in Europe.

None of those tells you how many signals you'll get for your 300 target accounts in Austria and Poland. Only a test on your own list does.

How to test a provider in two weeks before you sign

Most vendors will run a trial or a backtest if you ask. Here's the version I'd run.

Week 1: look backwards

1. Take your last 50 won deals. Export the company domains and the date the first real conversation started.
2. Ask the vendor to run them through historical data. For each company: did it show a signal, and when?
3. Count how many it flagged before you knew. A signal that showed up after your first call is a report, not intent. The useful number is the share of won deals with a signal before first contact, and how far ahead.
4. Split by country. If the UK looks great and DACH is empty, that's your answer for DACH.

If the vendor can't backtest, run the next week twice as carefully.

Week 2: look at today

1. Run your current pipeline and your target list. How many target accounts show any signal at all? How many in each of your top three countries?
2. Check 20 signals by hand. Pick at random. Does the company fit your ICP? Is there anything in the real world that backs the signal up: a job ad, a leadership change, a visit to your site? Or is it a university, an ISP or a company that was never going to buy?
3. Ask sales about the top 10. Would they call these companies this week based on what the tool shows? If they need another hour of research per account to know why, the cost of the tool is higher than the price.
4. Check the consent and contract answers. Send the questions above before the demo. Vague answers are an answer too.

What a good result looks like

There's no universal threshold. Context matters: a company selling to large UK banks will see more than one selling to family-owned manufacturers in Bavaria.

What I'd look for: a meaningful share of past wins flagged before first contact, signals that survive the manual check, and coverage in the countries where you actually sell. If one of those is missing, the tool may still be worth it for one market and not for another.

So which one should you buy?

Maybe none yet.

For a small team selling into DACH, Benelux or CEE, I'd start with first-party engagement and public triggers: registries, local job boards, regulation deadlines. They're cheaper, they're easier to explain to your DPO, and they often carry more weight than topic intent in those markets. I've listed free sources in the buying signals most teams miss.

Add third-party intent when you know which signals predict a purchase for your offer, and when a two-week test shows it sees your market. In English-heavy markets with lots of online research, that may be early. In the Mittelstand, it may be never.

Whichever method you use, a signal is a reason to look at a company, not permission to contact a person. Contact rules differ a lot by country. Cold email needs consent in some EMEA markets and not in others, and LinkedIn messages are a grey zone: personal one-to-one messages are widely used, mass automation is where the risk is. The country breakdown is in GDPR and buying signals.

I combine these sources for B2B teams across EMEA, weighted per market, and deliver the result every Monday: the companies showing buying signals that week, why now, who to talk to and suggested messaging. If you want to see what that finds in your market, ask for a free sample of 3 companies.

References

1. Bombora, Our data. bombora.com
2. Bombora, "Bombora's Company Surge intent data still works despite challenges like remote work and the cookieless future". bombora.com
3. Directive 2002/58/EC (ePrivacy Directive), Art. 5(3). eur-lex.europa.eu
4. CJEU, IAB Europe v Gegevensbeschermingsautoriteit, C-604/22, judgment of 7 March 2024. eur-lex.europa.eu
5. G2 Documentation, Salesloft integration (Buyer Intent signal types, account-level collection). documentation.g2.com
6. G2 Documentation, Buyer Intent. documentation.g2.com
7. Leadfeeder, Company (history of Leadfeeder, Echobot and Dealfront). leadfeeder.com
8. Leadfeeder Help Center, What is the Leadfeeder tracker. help.leadfeeder.com
9. Leadfeeder, "Website visitor tracking in a post-GDPR world: what B2Bs need to know". leadfeeder.com
10. CJEU, Planet49, C-673/17, judgment of 1 October 2019. eur-lex.europa.eu
11. EDPB, Guidelines 2/2023 on Technical Scope of Art. 5(3) of ePrivacy Directive, version 2.0, adopted 7 October 2024, paragraphs 54 and 55. edpb.europa.eu