Intent Data vs Buying Signals: Which One Actually Finds You Deals?

Intent data tells you a company might be researching your category. A buying signal tells you something happened, on a date, reported by a source you can open in a browser. Both are useful, but not for the same job. Intent data earns its keep prioritizing a named account list you already have. It struggles at discovery, because it resolves to companies rather than people, infers rather than evidences and sits on a data supply that has been shrinking for years. If you're deciding whether to spend on an intent subscription this year, the question isn't which is better in the abstract. It's which job you're actually trying to do.

How intent data is actually collected

Most of the confusion about intent data comes from not knowing where it comes from. (For the other side of the comparison, the complete guide to B2B buying signals covers what counts as a signal and why.) There are three sources of intent data, and they are not equal.

Bidstream data is the exhaust of programmatic advertising. Every time an ad auction runs, information about the page and the visitor passes through the pipes. Aggregators collect it at enormous scale. It's the cheapest and broadest source, and the most compromised. Privacy enforcement and browser changes have been squeezing it for years, and it was never designed to identify business buyers in the first place.

Co-op or publisher data comes from a consented network of B2B publishers who share what topics readers engage with. Narrower, better provenance, more expensive.

First-party intent is what happens on your own properties: your pricing page, your docs, your comparison pages. It's the highest-quality intent signal that exists, you already own it and it's routinely the most under-used asset in the building.

Notice that only the third one involves your prospect doing something about you. The other two tell you somebody at a company engaged with a topic.

The four structural limits

None of these are fixable by choosing a better vendor. They're properties of the approach.

It resolves to accounts, not people. Intent works by matching an IP address to a company. That's the ceiling. You learn that somebody at a 400-person manufacturer looked at warehouse automation content. You don't learn who, so you still guess which of eleven plausible people to email. Gartner describes B2B buying as a loop through six buying jobs that pulls in new stakeholders as it goes, and account-level data cannot tell you which of them did the reading.

Remote work broke the matching. IP-to-company resolution assumes people work in offices on corporate networks. VPNs, home broadband and shared gateways all degrade it. The model was built for a working world that no longer exists.

It infers, and inference can't be checked. This is the one that quietly kills adoption. A rep gets an account "showing high intent," opens it, and finds nothing they can point at. There's no artifact: no announcement, no filing, no job post. Just a score. After a few of those, reps stop opening the tab. The data may be fine; it's unusable because it isn't inspectable.

A topic is not a pain. "Interested in supply-chain software" isn't a reason to call and it isn't a sentence you can put in an email. The rep still has to invent the relevance, which is the hard part of the job and exactly the part the tool didn't help with.

Where intent data genuinely wins

I'd rather make this argument honestly than win it cheaply, so here's the case for buying intent data.

You run ABM against a fixed, named list. Forrester's research found ABM programs deliver higher ROI than non-ABM marketing across every region it measured, and disciplined prioritization is part of why. If your world is 300 target accounts and the question is which 30 to work this month, intent is real input. You're not asking it to find anything. You're asking it to rank a list you already trust. That's the job it's good at.

You have enough volume for the noise to average out. Intent is probabilistic. At 50 accounts, a false positive is a wasted week. At 3,000, the error rate washes out and the aggregate direction is informative.

You're measuring account-level engagement over time. Trend matters more than any single reading. An account whose engagement has climbed for six straight weeks is telling you something even if no individual data point is trustworthy.

You already have first-party intent and aren't using it. Before spending on third-party data, instrument your own site. Who's on the pricing page twice this week is worth more than any purchased topic surge.

Where public buying signals win

Buying signals work differently, and the difference is that a human published the underlying fact.

Discovery. A signal surfaces companies that were never on your list. That's the part intent structurally can't do. A company has to already be researching for intent to see it, and only around 5% of any market is doing that at a time.

Attribution to a person. "Named a new Chief Revenue Officer" comes with a name. You know who to contact and why they specifically care.

A message that writes itself. The event is the relevance. You don't have to manufacture a reason for the email; you have to be disciplined about how you reference the one you have.

Verifiability. A rep can open the source before dialling. That single property does more for adoption than any accuracy improvement.

A note on the statistics in this debate

While researching this piece we tried to trace the numbers that get quoted in every article on trigger-event selling. Three of the most repeated could not be traced to any primary source: a widely cited figure about new executives making a purchase within their first 100 days, another about half of all revenue going to whoever responds to a trigger first and a conversion-rate comparison between signal-led and intent-led outreach. They appear in dozens of vendor blog posts, each citing the last, with the trail ending nowhere.

We're not saying they're false. We're saying nobody publishing them knows whether they're true.

That's an odd foundation for a category whose whole promise is better evidence. It's also the reason this piece leans on the 95-5 research from the LinkedIn B2B Institute and Ehrenberg-Bass, on Gartner's published work on B2B buying behavior, and on numbers from our own platform that we can show our working for. If a statistic can't be opened in a browser, it fails the same test we apply to signals.

How to decide

The short version: if you already know who your accounts are, intent data helps you sequence them. If your problem is that you don't have enough of the right accounts, intent data is the wrong tool and no amount of budget fixes that.

Most teams under 100 people have the second problem and buy for the first.

The stakes are not small either. Harvard Business Review pegged the cost of bad data at more than $3 trillion a year in the US alone, and an intent feed you cannot inspect is bad data you cannot even detect.

The two aren't mutually exclusive

The strongest setup uses both, in the right order. Public signals find and prioritize accounts that were never on your radar. First-party intent tells you when one of them starts paying attention to you. Third-party intent, if you have the volume and the list to justify it, sorts the known world.

What doesn't work is buying a topic feed and calling it a prospecting strategy. The feed can only ever show you the small slice of the market that's already looking. It's the slice your competitors are also staring at, where you arrive third.

The alternative is to work out which public events reveal the pain you solve, and go looking for them. Which events actually fire, and how often, is a question we can answer with real numbers rather than intuition. Or you can see what the method looks like when it runs automatically, and how it compares to the intent-led tools you may already be evaluating.