B2B intent data: what it is, where it comes from, what it cannot tell you, and how to use it without being creepy.

Definition

B2B intent data is information about the online research behavior of companies and the people who work there, used to estimate which accounts are actively looking into a problem, a product category or a vendor right now.

It is built from signals such as content read on publisher websites, visits to your own website, review site comparisons and engagement with emails, ads or webinars. A provider or your own team turns those signals into topics and scores per account, so marketing and sales teams can see which companies are researching what.

Intent data is a probability, not a fact. It tells you that research activity on a topic rose at an account. It does not tell you that the account has a budget, a project or a decision date, and third-party data rarely tells you which person is doing the research.

What B2B intent data tells you, and what it does not

A lot of disappointment with intent data comes from expecting it to answer questions it was never built to answer. The table separates the two, as this page sees them.

Read it as a reason to look at an account now, not as a reason to pitch it. The account still has to match your ideal customer profile, and a person still has to confirm the need in a conversation.

Why B2B intent data matters for sales and marketing teams

Many B2B buyers research a problem, compare options and read reviews before they talk to any vendor. In this page's view, part of the decision may be shaped before a demo form is filled in. Intent data is an attempt to see that earlier research while the buyer is still forming a view.

• For marketing teams. Spend ad budget and content on in-market accounts, and choose campaign topics from what target buyers actually research.
• For sales teams. Decide which accounts a rep works this week, and which problem to open with, instead of working a list from the top.
• For RevOps and GTM leaders. Add a timing layer to account scoring and routing, so marketing and sales act on the same accounts at the same moment.
• For customer success. Spot existing customers researching competitors or alternatives, a possible churn risk worth a conversation.

None of this replaces fit. B2B intent data adds a "when" to the "who" your ideal customer profile already defines, and it is only as useful as the actions your sales and marketing teams take on it.

Types of intent data: first-party, second-party and third-party

Intent data is grouped by who collects the underlying behavior. The group decides how much you can trust it, what you can see, and which privacy rules you have to think about. Some vendors add labels such as enriched or derived intent, which describe processing on top of these three, not a new source.

First-party intent data

First-party intent data is behavior you observe on properties you control: pricing page visits, repeat visits, demo requests, product sign-ups, email clicks and webinar attendance. It is the most specific kind, because it shows interest in you, not just in a topic.

Its limit is reach. It only covers accounts that already found you. Anonymous website visits have to be matched to a company, and HubSpot, for example, documents that its buyer intent tool connects anonymous visitors to companies through their IP addresses. That match can be weak for remote workers, mobile networks and shared offices.

Second-party intent data

Second-party data is another organization's first-party data made available to you, most often from a review platform, a trade publisher, a professional network or an event organizer. You see which accounts researched your category, your product or a competitor on their site.

Third-party intent data

Third-party intent data is collected across many websites you do not own, then aggregated and sold or licensed by data providers. It shows research by accounts that have never visited your site. It is also the kind where you usually see least about how each signal was collected.

First-party vs third-party intent data

Many teams use both. First-party data tells you who is engaging with you, third-party data tells you who is researching the category before they reach you. The trade-offs are different.

How intent data is collected

This page describes third-party intent as a five-step pipeline. Providers differ in the details, and the quality of each step decides the quality of the final score.

Capture happens through tags, cookies or similar technologies on websites and apps. HubSpot says its tracking code collects website activity, IP addresses and other online identifiers. LinkedIn uses its Insight Tag for website visits, and one co-op provider, Bombora, describes a proprietary tag on member websites.

Resolution matches a visit to an organization, usually through the IP address of the network the visitor uses, or through a device or cookie identifier linked to a business profile. This step is where most of the error enters, because an address identifies a network, not a buyer.

Classification maps the content that was read to a list of topics. Each provider keeps its own topic taxonomy. HubSpot, for instance, calls its research topics proprietary categorizations of websites, not search engine keywords, so the same research can appear under different topic names in different products.

How two platforms document their own buyer intent features

Vendor help pages are the most concrete public description of how intent data products work. Two are summarized here only for what they say about their own products. This is not a recommendation or a ranking.

LinkedIn Sales Navigator Buyer Intent

LinkedIn's Buyer Intent FAQ says the account score combines more than 180 distinct insight signals. It groups them as LinkedIn activity such as reactions, comments and page follows, ad interactions, InMail responses, and website visits where the company has installed the LinkedIn Insight Tag.

Its Buyer Intent Alerts FAQ explains that alerts fire when employees or leaders at a saved account view your company page or website. It cites GDPR as the reason it limits detail, and hides an attribute when fewer than 10 employees at the account share it.

HubSpot buyer intent and research intent

HubSpot's buyer intent article describes company-level website activity: visits, unique visitors, last visit and top pages. It says the tool can also surface research across the web and company news such as funding, executive hires, layoffs, product launches and mergers.

Its research intent article describes the mechanism: when an anonymous IP visits a site in HubSpot's network, the visit is matched to a company, then to topics assigned to that site. A research level compares the past 30 days of activity with the previous 30 days.

Both descriptions show the same pattern this page describes: a tag captures activity, an identity step resolves it to a company, and a model turns it into a score. Neither page publishes accuracy figures, and this page does not quote any.

Bidstream data, co-op data and publisher data

Three common sources of third-party intent data work very differently, and the difference matters for both accuracy and privacy. One co-op provider, Bombora, describes all three on its intent data page, from its own commercial point of view.

Co-op data

Bombora describes a data cooperative as traffic, engagement and content consumption signals unified from a group of publishers, brands, websites and apps. The provider knows which sites participate and can set collection rules for them.

Co-op data is only as broad as its membership. Topics that member sites do not cover will show little research, even when buyers are active elsewhere. Ask how many sites contribute to the topics you care about, not only how many sites exist in total.

Publisher direct data

Publisher data comes from one publisher or a small set of sites in a niche, such as a trade media group or a review site. It can be deep on that subject and silent on everything else, and it can include readers who registered for gated content.

Bidstream data

Bidstream data comes from programmatic advertising. When a page loads an ad, a bid request describing the page and the visitor goes out to many potential bidders in real time. Some intent providers keep those bid requests and turn them into research signals. Jeluvi's bidstream data entry covers the mechanics in depth.

The UK Information Commissioner's Office reviewed real time bidding in its 2019 update report into adtech. It found that one website visit can result in a person's personal data being seen by hundreds of organizations, and said legitimate interests cannot be used for the main bid request processing.

That does not settle every product. It does mean you should ask any provider using bidstream data what consent was collected at the point of capture, and whether its contracts allow bid requests to be used for anything other than bidding.

Surge scores, intent topics and buyer intent keywords

A surge score shows that an account's research on a topic has risen above its own usual level. It is a relative measure: a large company that always reads about cloud security will not surge just because it reads a lot, only when it reads more than before.

Providers calculate it differently and do not always publish the method. Bombora says its tag lets it build historical baselines and identify behavioral changes. HubSpot documents a research level that compares the past 30 days with the previous 30. Ask any provider the equivalent question before you build rules on a score.

Vendors call an account with sustained research on your topics in-market. Treat in-market as a label for sales and marketing priority, not as proof that a buying process has started.

Buyer intent keywords vs intent topics

Buyer intent keywords are the words a buyer uses while researching: a problem, a product category, a competitor name. Intent topics are the provider's own categories that many keywords and pages roll up into. You usually choose topics from a list, and some tools let you start from your own keywords or domain.

• Pick few topics. Choose topics that describe the problem you solve and the category you sell in, not every keyword related to your industry.
• Watch combinations. One account surging on three related topics at once is a stronger signal than one topic alone.
• Check the trend. A score that rose over several weeks means more than a one-week spike, which can be one person reading one long report.
• Review topics quarterly. Remove topics that never produce meetings, and add the language your won deals actually used.

Intent to learn vs intent to buy

Research on a broad problem, such as "reduce churn", is often intent to learn. Research on a product category, comparisons, pricing and a competitor name sits closer to a buying decision. This page's advice: weight comparison and competitor topics above general education topics, and never treat either as a confirmed opportunity.

Buying signals data: where B2B intent data fits

Buying signals data is the broader family that intent data belongs to. A buying signal is any observable change suggesting an account may need what you sell. Intent is the research signal. Other signals come from what the company does and what it is.

The strongest accounts combine several of these: a good fit, a trigger event, and research on your topics. Intent data alone, without fit, sends reps to accounts that are curious but could never become customers.

Enriched intent: adding firmographic and technographic fit

Enriched intent is intent data joined with fit data about the account. Technographics add the tools a company runs, and firmographics add size, industry and region. Where each firmographic field comes from, and why providers disagree, is covered in firmographics sources.

The join is only as good as the weakest field. A surge resolved to the wrong company, or a headcount from an old filing, produces a confident score on bad inputs. The wider picture of B2B data quality, decay and hygiene applies to intent feeds too.

Which buying signals sales teams should act on

Not every signal deserves the same response. Sales teams get more from buying signals data when each signal type has an agreed action, so a prospect showing strong signals gets a person and a weak signal gets content.

• Strong: direct engagement. A demo request, a pricing page visit from a target account, or a reply asking about pricing. Route to a rep the same day.
• Strong: decision makers appear. Several people from one account engage with you, including leaders who sign off on budget.
• Medium: trigger events. Funding, a new leader, an acquisition or a change of tools. Research the account and start a relevant sequence.
• Medium: category research. A third-party surge on your topics at a fitting account. Research first, then a problem-led first touch.
• Weak: single visits. One blog visit or one email open. Keep in nurture and watch for more signals.

Signal strength also depends on who the prospect is. The same pricing page visit means more from a decision maker at a fitting account than from a junior prospect at a company outside your market. Weigh buying signals by fit and role before sales tools turn them into tasks.

Account-level vs person-level intent

Much third-party intent data works at the account level. It says a company is researching a topic, not who at that company is doing it. That is partly technical, since IP addresses identify networks, and partly a privacy choice.

Person-level signals mostly come from first-party data or from platforms where people are logged in: a known contact who opened an email, filled a form or reacted to a post. Some products claim to identify anonymous individuals. Treat that claim carefully, for accuracy and because naming a person makes it personal data.

In practice, account-level intent tells you which company to work, and your research, lead enrichment and the Sales Navigator account view tell you which people in the buying group to contact.

The limits of intent data

Intent data is useful and noisy at the same time. Knowing where the noise comes from keeps you from trusting a score more than it deserves. The points below are this page's assessment, not measured rates.

• Research is not buying. Students, analysts, journalists, job seekers and your own competitors read the same content as buyers. LinkedIn says its alerts try to separate job seekers from buyers, which shows the problem is real.
• IP resolution breaks. Remote work, VPNs, mobile networks and shared offices make some visits impossible to match to the right company.
• Big companies look busy. Large accounts produce research on many topics, so they can show surges without a real project behind them.
• Topics are models. Content is classified into topics automatically, and a page about one thing can be counted as research on another.
• Coverage has gaps. Research on sites outside the provider's network, in private communities, or in conversations with peers is invisible.
• Timing lags. Scores compare windows of days or weeks, so by the time a change appears the research may already be weeks old.

Intent data, privacy and GDPR

B2B does not mean privacy law stops applying. Intent data is built on online behavior, and once behavior can be linked to a person or a device, it is personal data under the GDPR and similar laws.

The GDPR defines personal data as information relating to an identifiable natural person, including by an online identifier. Recital 30 names IP addresses and cookie identifiers as examples, and notes they can be combined to create profiles.

Consent and cookies

The ICO's guidance on cookies and similar technologies says you must tell people about cookies, explain what they do, and get consent given through a clear positive action. It says behavioral tracking needs particular care to get clear and specific consent.

For first-party intent data, that means your website tracking should only run for visitors who accepted it. For third-party data, it means asking the provider how consent was collected at every site that fed the signal.

B2B contacts are still people

The ICO's business-to-business marketing guidance says the UK GDPR applies to direct marketing even in a business context, for example when you hold the name of the person who represents the business. Business contacts can also object to your marketing.

This page is not legal advice. Ask your privacy counsel how these rules apply to your data, your providers and your markets.

Intent data under US privacy law: the FTC and California

In the US, federal enforcement and California law both reach the raw material of third-party intent data: browsing data, bid request data and the companies that sell it.

What the FTC has said about browsing and bid request data

In 2024 the FTC announced an order against Avast over charges it sold re-identifiable browsing data, collected through its software, without adequate notice or consent. The order bans selling or licensing web browsing data for advertising purposes.

An FTC technology blog post on mass data collectors states that browsing and location data are sensitive, even without names. It adds that contract clauses against re-identification are often insufficient on their own, and that data use must match the purpose it was collected for.

In December 2024 the FTC announced a proposed order against Mobilewalla, a data broker it said kept information from bid requests even when it did not win the auction. The proposed order bans collecting or retaining auction data for any purpose other than taking part in the auction.

The FTC finalized the order in January 2025.

Those cases involved consumer location and browsing data, not B2B topic scores. They still show what the agency examines: whether people knew, whether they consented, and whether data collected for one purpose was resold for another.

California: the CCPA and data brokers

The California Privacy Protection Agency's CCPA FAQ says personal information can include internet browsing history and inferences about a person's preferences. It says CCPA rights cover contacts at business customers and vendors, and that the business-to-business exemption expired on December 31, 2022.

The same FAQ describes a right to opt out of the sale of personal information and of its sharing for cross-context behavioral advertising. Intent programs that sync audiences to ad platforms should check how those opt-outs reach every system that holds the data.

California also regulates data brokers, defined as businesses that knowingly collect and sell the personal information of consumers they have no direct relationship with. The agency's data broker page says brokers must register yearly and, from August 1, 2026, process deletion requests from a central mechanism.

Whether a given intent provider is a data broker depends on what it collects and sells. Ask it directly, and check the public registry the agency maintains before you sign.

How to use intent data responsibly

Responsible use protects buyers and protects your results, because outreach that feels like surveillance gets ignored or reported. Six practices cover most of it.

• Never reveal the tracking. Do not write "I saw you were researching X". Use the signal to choose the account and the topic, then lead with a relevant problem.
• Keep it at the account level. Use intent to prioritize companies, and reach individuals through normal prospecting with a documented lawful basis.
• Ask about the consent chain. Get the provider's written answer on sources, consent, bidstream use and data broker registration before you sign.
• Honor objections everywhere. An opt-out in your email tool should also remove the person from scoring and ad audiences.
• Use data for the purpose given. Do not reuse signals collected for one purpose in a way people would not expect, a theme the FTC stresses.
• Keep less, for less time. Store the topics and scores you act on, and let old signals expire.

Using intent data in ABM

Intent data works at the account level, which is why it fits account-based marketing. In an ABM strategy, intent helps in three places: choosing the target list, deciding which accounts to activate this month, and tailoring the message.

• List selection. Among accounts that fit the profile, add those showing research on your topics to the tier that gets the most attention.
• Activation. Move surging accounts from the watch list into active campaigns: targeted ads, sales outreach and executive contact.
• Messaging. Use the topics the account researches to choose the case study, the pain point and the content you send.
• Measurement. Compare engagement and pipeline for accounts activated on intent with accounts activated on fit alone.

Intent should not replace the shared account list. It reorders attention inside that list, so marketing and sales still work the same accounts, only in a better order. Turning a signal into a written rule is covered in intent based targeting.

Using intent data for sales outreach timing

For sales teams, intent data answers one question: which accounts deserve a touch this week. It is a timing tool for outbound lead generation, not a replacement for research or a reason to send more email.

Once an account passes the fit check, the rep finds the likely owner of the problem, checks for a trigger event, and starts a sales cadence across email, LinkedIn and phone. The first message names a problem the topic suggests, never the tracking behind it.

Speed matters most for first-party signals, such as a pricing page visit from a target account. Third-party surges can build over weeks, so a thoughtful touch within the same week beats an instant generic one.

Getting intent data into your CRM and lead scoring

Intent data only changes behavior when it shows up where reps and marketers already work. A separate dashboard that reps never open cannot change what they do.

• Account fields. Write the top surging topics and a score to the account record in your CRM, with the date they were last updated.
• Lead scoring. Add intent as one input to account and lead scoring, next to fit and engagement, so it can raise a lead's priority but not qualify it alone.
• Alerts and routing. Send an alert to the account owner when a fitting account surges or a target account visits the pricing page.
• Audiences. Sync surging accounts to ad platforms and email lists, and remove them when the surge ends or a person objects.
• Reporting. Tag opportunities that were intent-sourced, so the pipeline and win rate comparison is possible later.

Prioritizing accounts with fit and intent

A simple way to put intent to work is a two-by-two of fit and intent. It keeps reps from chasing curious accounts that could never buy, and from ignoring great-fit accounts that are quiet for now.

Scores feed the same logic as lead qualification: intent can raise an account's priority, but a conversation still confirms need, authority, budget and timing.

Types of intent data providers, and what to ask them

Providers are easiest to compare by the source they rely on. This page names categories only, with no ranking; many products combine several categories.

Questions to ask an intent data provider

Ask every provider the same questions, in writing, and test the answers against your own data.

• Sources. Which sites or feeds produce the signals for my topics, and is any of it bidstream data?
• Consent. How was consent collected where signals were captured, and can you document it?
• Status. Are you registered as a data broker anywhere, and how do you process deletion and opt-out requests?
• Method. How is a surge calculated, over what window, and against what baseline?
• Resolution. How do you match visits to companies, and how do you handle remote and mobile traffic?
• Taxonomy. Can I see the topic list, and can I build custom topics from my own keywords?
• Back test. Can you score my last closed-won and closed-lost accounts before I sign?
• Delivery. How does the data reach my CRM, how often, and at what level of detail?

Measuring whether intent data works

Compare accounts worked on intent with a similar group worked without it, over at least one full sales cycle. Everything else is a vendor dashboard.

Small samples mislead here, because a few large wins can make any group look better. Read the results the way you would read any data insight: check the sample, look for a second explanation, then decide what changes.

Common mistakes with intent data

• Telling the prospect you tracked them, which turns a relevant message into a privacy complaint.
• Working surging accounts that do not fit your profile, because a score looks urgent.
• Tracking dozens of broad topics, so large companies seem to surge every week.
• Sending the whole surge list into one automated sequence instead of researched outreach.
• Buying third-party data without asking where it came from and how consent was collected.
• Judging the data by vendor dashboards instead of your own meetings, pipeline and wins.
• Letting marketing and sales use different intent scores, so they chase different accounts.

In a sequence

The template below turns an intent signal into a first email without mentioning the signal. It names a problem the research topic suggests, gives one relevant proof point, and asks one question. It was written for this page.

Subject: {{problem}} at {{companyName}}

Hi {{firstName}},

Teams like yours at {{companyType}} companies often look at {{problem}} when {{trigger}}.

We helped {{similarCustomer}} with {{outcome}}, mainly by {{approach}}.

Is {{problem}} something you are working on this quarter, or is someone else closer to it?

{{senderName}}

Frequently asked questions