What Is Intent Data in B2B and When Is It Worth Using

    Short answer

    The four types, how signals are collected and scored, what they cannot tell you, and a market-size rule for deciding whether to bother.

    Artem Smirnov
    Artem Smirnov

    Last updated · 10 min read

    Artem Smirnov in a dark suit against a charcoal studio backdrop, next to the line 'WHAT IS INTENT DATA. Useful in big markets only'

    What is intent data, in plain terms?

    Intent data is a record of online research that suggests a company may be looking into a problem, or a product, like yours. A provider watches what people at an account read and compares it with that account's normal level. When the research rises, it flags the account.

    A hypothetical example: four people at a civil engineering firm spend one week reading comparison pages and pricing content about project management software, far more than usual. The provider marks the firm as surging on that topic. You learn the company name and the topic. You do not learn who the four readers were.

    Real reports look the same. Bombora's Summary Company Surge report lists each account by company name, domain, average composite score and topic count. HubSpot's Company Surge signal, powered by Bombora, fires when a tracked company's research spikes on topics you set. It shows the topic, research level, category and top metros. No names of people.

    Intent data exists because buyers do most of their homework alone. In Gartner's 2024 research, 632 B2B buyers answered questions in August and September. The survey showed 61% would rather buy without any sales rep involved, and most choose to do their own research through digital channels (Gartner). Intent data tries to watch that research from the outside. I see the same thing from the seller's side: nobody buys from an email, they buy after they audit you.

    In sales, intent data helps decide who to contact and when. In marketing, it helps decide which accounts see which content and campaigns.

    Short verdict: in a big market, intent data helps me decide who to contact first. In a small one, my team skips it.

    What are the four types of intent data?

    TypeWhere the signal comes fromWho owns itTypical precisionExample signal
    First-partyYour website, your emails, your productYouHighest, but only for people who already found youThree visits to your pricing page from one company in a week
    Second-partyAnother company's own data, shared through a partnershipThe partner, shared with youHigh for that partner's site, blind everywhere elseAn account compares your product with a rival on a review site
    Third-partyPublisher networks and the bidstreamThe providerDirectionalMore people at an account reading articles on your topic than usual
    DerivedA score built from the other threeWhoever builds the modelOnly as good as its inputsAn account scored "high" this week

    Vendors draw these lines differently. This is the split I use, and the precision column is my judgment, not a measured figure.

    First-party intent is activity on things you own. It is the most direct signal you can get, with one hard limit: it never shows a company that has not found you yet.

    Second-party intent is someone else's first-party data, passed to you by agreement. Review sites are the usual case. They see who looks at your profile and your competitors' profiles, and they sell that view to you.

    Third-party intent is collected on sites you do not own. It comes through networks of publishers or through the bidstream, which is the data passed between websites and buyers of ad space each time a page loads. Its appeal is reach, because it can surface accounts that have never heard of you. The cost is precision. It points in a direction more than it proves anything.

    Derived intent is a model. It weighs how many signals arrived, from where and how recently, and turns them into one label or number per account.

    How do providers collect and score intent signals?

    The basic mechanic is the same almost everywhere. A provider learns how much an account normally reads on a topic. When recent reading climbs clearly above that baseline, the account gets flagged. Two providers publish enough documentation to show what that looks like in practice.

    Per G2's documentation, its buyer intent data comes from research on four review sites: G2, Capterra, Software Advice and GetApp. Signals come from actions such as viewing your product profile or your pricing, comparing you with a competitor, or browsing alternatives.

    Demandbase's documentation says it draws on the bidstream, B2B content consumption and engagement on your own website. It also notes that its strength thresholds can change as the model is updated.

    G2Demandbase
    SourcesResearch on G2, Capterra, Software Advice and GetAppBidstream, B2B content consumption, your website
    Level collectedOrganizationOrganization (account)
    LabelsBuying stage: Awareness, Consideration, Decision. Activity level: Low, Medium, HighStrength: low, medium, high
    RefreshStage and activity updated dailyViews aggregated weekly
    Threshold to knowHigh activity: engage immediately. Medium: engage within a weekAt companies with 51-200 employees, High means 8 or more people reading
    Surge ruleNo numeric rule published; stage and activity are recalculated from page engagement and signal type"Trending" means signals in the last 7 days are far above the account's own average of the previous 8 weeks (Demandbase's cutoff is two standard deviations)

    What can intent data not tell you?

    A surge tells you an account may be researching. It does not tell you who is reading, why they are reading, or whether any money exists.

    Who. Forrester's Q1 2023 survey of intent data users found that the top execution challenge is identifying the specific contacts to target inside accounts that show intent. The reason given is that few providers can deliver signals from known contacts (Forrester).

    Finding the decision-maker is still your job. Go after the titles that actually make the call in that niche, not a vague "manager". Titles like Assistant, Associate and Manager come off the list.

    Why. Many people read content in your category. Students, journalists, competitors and your prospect's current vendor all add to the count, so some "high" weeks are noise. Demandbase's own documentation calls its intent probabilistic.

    Whether. Reading says nothing about budget or timing. A team may research a topic for an internal report, for next year's plan, or to push back on the vendor it already has.

    The list itself. If your titles are wrong and your data is unchecked, a surging account just receives a bad email sooner. That is why the list is usually the problem before the email is.

    When nobody replies, check the list before you rewrite a single email. Copy cannot fix a bad list, and a surge flag cannot either. I have seen this in our campaigns. Two companies sent an identical message. One got 2% replies, the other 7.5%.

    Do companies that use intent data actually get results?

    In the same Forrester survey, more than 85% of intent data users reported business benefits. The benefits named most often were higher outbound response rates and better sales prospecting results. More than 70% use several providers, and nearly half draw on at least three sources. 30% plan to raise spend, while only 10% plan to cut it.

    The gaps matter more. Only half apply the spend to speeding up pipeline, and fewer than a third use it for customer use cases. More respondents expected a benefit than reported getting it. The biggest shortfall appeared in sales applications and in revenue growth overall.

    Every respondent already pays for intent data, so the sample leans friendly. The survey is from 2023, and the blog post does not give its sample size.

    When is intent data worth using?

    My team uses intent data when the total market is large, and leaves it out when the market is small.

    In a big market, many accounts are researching at any moment, so a surge helps you choose which of thousands to contact this week. In a small market, the signal covers so few companies that planning around it leaves most of the market untouched.

    Surges are rare by design. Bombora's own thresholding guide counts an account as spiking only at a score of 60 or more on a topic, a statistically significant rise over its own baseline. So only a subset of monitored accounts counts as surging. Bombora also offers a report that compares monitored domains in your list with domains showing intent.

    Picture 2,000 target companies where only about 10 show intent signals. Skip intent data there, because the other 1,990 companies still need to hear from you.

    Your marketDoes intent data help?What to do with itWhy
    Large addressable market, with many accounts researching at any momentYes, as a layerUse surges to set the order of outreach inside a list you would contact anywayEnough accounts surge each week to change who hears from you first
    Small market of a couple of thousand named companiesUsually notSkip it and contact the whole list with one strong offerSignals land on a handful of accounts, and the rest would be ignored
    Niche where buyers choose suppliers through referrals, tenders or trade relationshipsNoTreat any signal you happen to get as a bonusThere is little online research to observe

    Fix the offer and the list first. Then let a surge decide who gets the email on Monday.

    I treat a short list of surging accounts the same way I treat 50 to 100 hand-picked leads a month. Both are a bet that the right person happens to be free, ready and interested this month, and a coin flip is not a plan. You need quantity and quality. A surge list gives you a little quality and almost no quantity.

    If you are weighing a paid tool, see these price tiers and a buy-or-skip check for small teams.

    What should you do when your market is too small for intent data?

    1. Write one offer for the whole market. Aim it at this one market, not at everyone, because a message for everyone lands with no one in particular. The offer has to name this market's problem clearly enough that a company with no surge this week still wants the call. Running outbound when your market is small or running out is its own problem.
    2. Start from your clients. Begin with past and current clients, then search for businesses with the same profile. That gives you a list built on proof instead of on reading habits.
    3. Watch public triggers. A new senior hire, a new office or a newly announced project can do part of the job intent data does, at no data cost. Here are free trigger signals a small team can watch.

    Here is a made-up example for a company that rents site lighting to construction contractors. It is written to a contractor's operations manager:

    Subject: Lighting for night shifts on your sites

    Hi Priya, when your crews work nights, who sorts out temporary lighting for each site? We rent lighting towers per site, so nothing sits in your yard between jobs. If night work is on your next program, would a 15-minute call next week be worth it to compare cost per site?

    It never mentions what the reader researched, so it makes sense whether or not the account surged this week.

    Account-level signals describe companies. Once a named person and an email address sit next to the signal, you are handling personal data, and data protection rules apply.

    Under the GDPR, the basis to look at is legitimate interests. Article 6(1)(f) permits processing when it is needed to pursue a legitimate interest held by the controller or by another party, unless the person's own rights and freedoms override them (GDPR on EUR-Lex).

    Recital 47 states that a legitimate interest can cover direct marketing. It also asks for a careful assessment of whether the person could reasonably expect that use when the data was collected.

    People can object to direct marketing at any time, including profiling linked to it. Once someone objects, you must stop. You also have to tell them about this right clearly, at the latest in your first contact. When data reaches you from a third party, Article 14 requires you to tell the person where it came from. That matters when a provider supplied the contact.

    First-party website intent has its own rule. Article 5(3) of the ePrivacy Directive says storing or reading information on a user's device needs consent, given after clear and comprehensive information about the purpose. The exceptions are narrow: storage needed only to transmit a communication, or storage strictly necessary for a service the user explicitly asked for (ePrivacy Directive on EUR-Lex).

    A tracking tag that sets cookies or reads the visitor's device to identify companies falls under this rule.

    For the UK, the ICO's legitimate interests guidance was updated on 23 March 2026 for the Data (Use and Access) Act. It covers the three-part test and has sections on direct marketing and on business-to-business contacts (ICO).

    Our own rule for Germany, Austria and the Czech Republic: LinkedIn outreach first, and cold email only where we can show a compliant basis.

    This is not legal advice. Get your own counsel, and use a filled-in legitimate interest assessment as a starting draft.

    Not sure your market is big enough for intent data? Book a call and bring your company count. I will tell you whether to buy or skip.

    Questions founders ask about intent data

    How can I test whether a provider's signals fit my pipeline?

    Ask for history before you sign. Demandbase, for example, retains intent history going back as far as 13 months. Take the deals you won and lost in that period and check whether those accounts were flagged before the first meeting. If won deals show nothing, or half your market reads "high", the signal is not tracking your buyers.

    Do different intent data providers sell different signals?

    Yes. G2 sells research done on its own review sites, Bombora's Company Surge data also feeds tools like HubSpot, and Demandbase blends its own sources with outside feeds. Ask where each signal starts before you compare prices. Demandbase, for example, runs its own sources and also takes in feeds from Bombora, G2 and TrustRadius, which it labels "Intent Surge". Two providers can therefore show you the same G2 research under different names, so get each one's source list before you pay for both.

    What first-party intent signals can I collect for free?

    Your inbox and CRM already hold some. Look at replies that asked about price, prospects who forwarded your email to a colleague, deals that went quiet after a proposal, and old contacts who changed company. None of these needs a website tag, and together they show which accounts already know your name.

    How does intent data fit into lead scoring?

    Treat it as a tiebreaker that sits below fit. Score industry, company size and the right decision-maker title first. Then let intent move a fitting account up the queue. An account that is surging but does not fit should never outrank one that fits and is quiet, because a fit account can still buy next quarter.

    The order of work is offer, then list, then intent as a tiebreaker. In a market of a couple of thousand companies, skip the tool and contact everyone. Fix the funnel first. Then scale.

    Want to see what we would build for your company?

    I am Artem Smirnov. I help established B2B companies build and run the go-to-market system behind their pipeline, from outbound to AI search visibility. Book a call and we will see what fits yours.

    Artem Smirnov
    Artem Smirnov

    I help established B2B companies build and run their go-to-market systems. 9 years in outbound, 10,000+ campaigns.