B2B intent data.

B2B intent data is behavioral data that signals a company is in-market: researching a topic, comparing vendors. It comes in two forms: first-party (activity on your own site) and third-party (research across the wider web, sold by providers). Most of it is reported at the account level, meaning it tells you a company is interested, not which person.

That last sentence is the whole problem.

I run a contact-level advertising platform, so read this knowing where I sit. But the account-level limitation isn't my opinion. It's printed in the docs of every provider I'm about to break down.

Here's the trap. You buy intent data. A dashboard lights up: "Acme Corp is surging on account-based marketing." Your SDR opens Acme. There are 200 employees. The dashboard cannot tell them which one did the research. So the rep guesses, picks the VP of Marketing from a title filter, and sends a cold email that has nothing to do with what the actual researcher was looking at.

That's not a signal. That's a coin flip with extra steps.

This article is the honest version of the intent data conversation. What the signals actually are. What first-party, third-party, and "marketing intent data" really mean. The big providers, compared. And why the fix isn't more intent — it's knowing who.

If you've read my piece on contact-level marketing, you already know the thesis: start with named people, build everything around reaching them. Intent data is supposed to feed that. Mostly, it doesn't — yet.

---

What is intent data.

Intent data is behavioral evidence that a buyer is moving toward a purchase.

Someone downloads a whitepaper on data security. Someone reads three competitor comparison pages in a week. A company's research on "ABM platforms" spikes well above its normal level. Each of those is a behavior that suggests interest. Bundle enough of them together and you get what providers sell as "intent."

The promise is simple: stop guessing who's in-market, and spend your time on the 3-5% of accounts that are actually researching right now.

That promise is real. Prioritization works. The part that breaks is everything that comes after.

Because almost all of it is measured at the account level. The provider sees research happening, ties it to a company through IP and cookie matching, and reports "Acme is interested." It doesn't know which human did the researching, and in most cases legally won't say.

So you know a building is warm. You don't know which window has the light on.

---

Intent signals.

An intent signal is a single observed behavior. The data is the collection of them, scored.

Here are the ones that matter, roughly in order of how much they're worth:

→ Pricing page visits — someone reading your pricing is closer to a decision than someone reading your blog.

→ Competitor comparison views — "you vs them" pages are read by people building a shortlist.

→ Review-site activity — visiting your G2 or Capterra profile, comparing alternatives, reading reviews.

→ Repeat content consumption — the same account researching the same topic across multiple sessions.

→ Topic surges — a company's research on a topic jumping above its 12-week baseline.

One signal alone is noise. A CFO reads one article on a Tuesday and forgets it by Wednesday. What providers actually score is a cluster — multiple signals from the same account in a short window, weighted against that account's normal behavior. Bombora calls the spike a "surge." Demandbase calls its keyword version "trending" intent. Same idea.

The strongest signals are first-party and behavioral — someone doing something on a property you own. The weakest are modeled and third-party. Someone's research inferred from bidstream data on sites you'll never see.

Most teams over-trust the weak ones because they have more volume. More volume, less truth.

---

First-party intent data.

Intent data splits into two halves, and they fail in opposite directions. The first is first-party.

First-party intent is behavior on properties you own. Your website, your emails, your product, your ads.

It's the most accurate intent you can get. If someone visits your pricing page twice and opens three emails, that's a real human doing real things in response to you. No modeling, no inference.

The catch: it only covers people already engaging with you. First-party intent can't tell you about the 97% of your market that hasn't found you yet. And on most B2B sites, the visitor is anonymous — you see the behavior, not the name. (Putting names on that anonymous traffic is its own discipline; I cover it in website visitor identification.)

Third-party intent data.

Third-party intent is research activity captured across other websites and sold to you by a provider. This is the other half — and the one most people mean when they say "buying intent data."

Bombora runs a co-operative of around 5,000 B2B websites and tracks content consumption across them, mapped to more than 12,000 topics. 6sense and Demandbase both offer Bombora's feed inside their platforms. ZoomInfo went the other way — it acquired Clickagy and builds its intent from bidstream data across 300,000+ publisher domains. Different plumbing, same output: research activity attributed to companies.

Third-party intent gives you reach: visibility into accounts that have never touched your site.

But it has three structural problems, and you should know all three before you sign a contract:

→ It's modeled. Bidstream and co-op data is inferred and matched, not observed directly. Match accuracy varies.

→ It's lagged. Bombora's Company Surge measures a 3-week window against a 12-week baseline. By the time a surge surfaces, the research is already weeks old.

→ It's account-level. And this is the one nobody wants to talk about, so let's talk about it.

Why account-level intent underdelivers.

Here's the line straight from 6sense's own compliance docs: "6sense does not track individuals — all insights are derived at a company level."

Read that again. The most sophisticated intent platform in the category tells you, in its own documentation, that it can't tell you who the person is.

So picture what actually happens at your company.

The dashboard says "Acme Corp is surging on cloud security." Great. Acme has 200 employees. The research could have been:

→ The CISO building a shortlist.

→ An intern writing a report for a class.

→ A procurement analyst doing a routine renewal check.

→ Someone who left the company last week.

You have no idea which. So your rep does what reps do: they pick the most senior plausible title, write a generic "noticed you might be exploring security solutions" email, and send it. The actual researcher never sees it. The person who gets it has no idea what they're talking about.

That's the gap. Account-level intent is a great prioritization tool and a terrible targeting tool. It tells you which doors to knock on. It can't tell you who's home.

The lag makes it worse. "Someone at the company googled something adjacent three weeks ago" is not a reason to interrupt a stranger today. The window where that interest was hot has often already closed.

I'm not saying account-level intent is useless. I'm saying it does one job, sorting accounts, and gets sold as if it does two.

---

Marketing intent data.

"Marketing intent data" usually means intent signals used to drive marketing actions rather than sales outreach — feeding audiences, triggering campaigns, scoring leads.

This is where account-level intent runs into a second wall: you can't advertise to an account.

Ad platforms target people. LinkedIn, Meta, Google — every one of them matches on individuals, not companies. So when your intent provider hands you a list of surging accounts, you still have to convert "Acme is interested" into "here are the specific people at Acme to reach," and then get those people onto a platform at a usable match rate.

Most teams do this with a title filter and a hunch. They take the account, guess the persona, and run ABM display ads to a broad demographic audience hoping the right person sees it.

That's not intent-driven marketing. That's intent-flavored spray-and-pray.

The fix is to resolve the account down to named contacts, then reach those contacts directly. I break the targeting mechanics down in contact-level targeting, and the paid distribution side in contact-level advertising.

---

Intent-based targeting.

Intent-based targeting means using intent signals to decide who you reach and what you show them.

Done at the account level, it looks like this: pull surging accounts, push them into an ABM ad campaign, target by firmographic + title. The signal informs which accounts but not which people, so the targeting layer is still demographic guesswork underneath.

Done at the contact level, it looks different: a named person takes an action, and you reach that person with content matched to it.

The difference shows up in the numbers. A raw CSV of accounts-turned-titles uploaded to an ad platform matches roughly 20-50% of the list. Most of your audience never sees the campaign. With identity enrichment that maps business identities to the personal identifiers platforms actually match on, that climbs to 70-99%. Same intent, but now it reaches the people instead of evaporating into a low match rate.

That's also where the waste comes from. When your targeting is "everyone with this title at these surging accounts," you pay to reach hundreds of people who aren't the researcher. Tightening targeting to the actual contacts is where teams see up to a 90% reduction in wasted ad spend. You stop paying to reach people outside the buying group.

The point isn't to throw out account-level intent. It's to use it as the input (which accounts), and contact-level identity as the targeting layer (which people), so the signal turns into something a campaign or a rep can act on.

---

Third-party intent data providers, compared.

Here's the honest version. I sell a contact-level product, so I'm not neutral. But everything in this table is from the providers' own documentation, not my marketing.

The thing they share: all four surface intent primarily at the account level. None of them, on their core intent product, hands a rep a named person who did the research.

A few things worth pulling out.

Bombora is the water supply. 6sense and Demandbase both pipe in Bombora data and add their own graph on top — ZoomInfo is the exception, running on its own bidstream feed. When you compare the rest on third-party intent, you're often comparing wrappers around the same source. I wrote a full breakdown in Bombora intent data.

G2 got a lot stronger in early 2026. G2 acquired Gartner's Digital Markets — Capterra, Software Advice, and GetApp — and folded their buyer activity into one Buyer Intent view. G2 reports up to 2x more signals and 36% more in-market accounts from the combined dataset (G2 newsroom). It's the most decision-proximate signal in the category. It's also still account-level, and limited to buyers who use review sites.

Demandbase shipping a "person-based intent" product is the tell. When the account-based pioneer starts selling person-level signals as the upgrade, you know account-level alone wasn't closing the loop.

For the full vendor list and where each one fits, see intent data providers. For the ABM-platform angle on these same vendors, 6sense competitors covers the orchestration side.

---

What contact-level intent fixes.

Contact-level intent ties a signal to a named person, not a company.

Not "someone at Acme researched ABM." Sarah Chen at Acme read your pricing page, clicked your LinkedIn ad, and came back the next day for the case study.

That's the version a rep can act on without guessing. They know who. They know what that person looked at. They know the order it happened in. The cold email writes itself, because it's not cold anymore.

This is the bridge between account-level intent and pipeline. You can — and should — buy account-level intent to decide which accounts to work. Then you resolve those accounts to named contacts and watch their behavior: which person on the buying committee engaged, with what, when.

I wrote the full argument in contact-level intent data — it's the piece I'd read next if this post hit a nerve. It covers the two types of contact-level intent signals and how they feed campaigns and sales.

Two things this unlocks that account-level can't:

You see the whole buying committee, by name. Not "Acme engaged," but "the champion engaged, the CFO didn't, the CISO read the security page twice." That's the difference between knowing a deal is warm and knowing exactly where it's stalled. Most of the dark funnel — the research happening where you can't see it — stays dark precisely because account-level tools can't name the people inside it.

You can act on first-party signals immediately. When the person who visited is named, there's no 3-week lag and no title-filter guess. The signal and the contact are the same record.

The results follow from that. Reaching the full buying committee by name instead of betting on one champion is where teams see 320% ROI on contact-level ABM versus 180% on account-level, and 85% higher close rates. When you warm named contacts with ads before sales reaches out, email reply rates climb 470% and reps book 2.2x more meetings. None of that comes from better intent data. It comes from knowing who the intent belongs to.

---

How to actually use intent data.

You don't have to pick one. Use the layers for what each is good at.

→ Account-level intent for prioritization. Buy Bombora, 6sense, or G2 to sort your TAM. Which accounts are surging? Work those first. This is the job account-level intent does well — let it.

→ Contact-level identity for activation. Resolve the prioritized accounts to named contacts. Now you know who to reach, not just where.

→ First-party signals for timing. Track what your named contacts do on your site and ads. This is real-time and person-specific — the highest-quality signal you'll ever get.

→ Content matched to the stage. A surging account isn't a buying account. Most are still problem-aware at best. Feed them content that fits where they actually are, not a demo request.

The mistake is buying account-level intent and treating it like it's contact-level. It isn't. It tells you the account is warm. Your job is to find the person and reach them, and that's a different tool.

---

Go deeper.

Intent data is one input into the larger contact-level marketing system. The signal only matters if you can act on it — by name.

The POV bridge:

→ Contact-level intent data — why account-level intent underdelivers and what person-level signals look like. Read this next.

The providers:

→ Bombora intent data — an honest review of the feed most providers resell, with pricing.

→ Intent data providers — the full list, compared.

The activation layer:

→ Contact-level advertising — how to turn named contacts into reachable audiences across LinkedIn, Meta, Google, Reddit, and X.

→ Dark funnel B2B marketing — where buyer research hides, and why account-level tools can't see into it.