Intent Data: the complete reference for performance marketers.
Buyer intent signals are everywhere. Most of them are noise. This is the practical guide to telling the difference, choosing a provider, and turning intent into audiences that perform.
01
Definition
Defining intent data.
Intent data is behavioral evidence that someone is actively researching, evaluating, or about to buy a product or service. It’s the digital equivalent of overhearing a prospect say, “We need a new CRM by Q3.” Instead of guessing who might be interested, you act on signals from people already showing interest.
Those signals come from real activity: searches performed, articles read, pages visited, products compared, content downloaded. When stitched together and matched to identities or accounts, they tell you not just what someone is interested in, but where they are in the buying journey.
Intent data turns marketing from a broadcast into a response. You stop interrupting people who don’t care, and start showing up for the ones who already do.
02
Taxonomy
The four types of intent data.
Not all intent is created equal. Most confusion in this category comes from conflating these four data types. Each has different sources, different costs, and different applications.
Source
First-party intent
Behavior on properties you own: your website, your app, your emails. Highest quality, highest accuracy, lowest scale.
Source
Third-party intent
Behavior aggregated from publisher networks, review sites, and content syndication. Broad reach, variable signal strength.
Granularity
Account-level intent
Signals attributed to a company or domain. The default for B2B. Useful for ABM and sales prioritization.
Granularity
Person-level intent
Signals attributed to an individual. Higher precision, harder to source compliantly, gold for activation.
First-party vs. third-party at a glance
| Dimension | First-party intent | Third-party intent |
|---|---|---|
| Source | Your own digital properties | External publishers, review sites, partner networks |
| Accuracy | Very high. You control collection | Variable. Depends on provider methodology |
| Scale | Limited to your audience | Broad. Millions of signals across the open web |
| Best for | Retargeting, lead scoring, personalization | Top-of-funnel discovery, market sizing, prospecting |
03
Mechanics
How intent data actually works.
At a high level, every intent data product does the same four things: collect signals, attribute them to an entity, score them, and make them activatable. Where providers differ is the breadth and quality of each step.
1. Signal collection
Providers ingest raw events: page views, search queries, content consumption, ad interactions, form submissions. The best providers collect from many surfaces, not just one publisher network.
2. Identity resolution
Raw events get tied to a known entity: a company (account-level) or an individual (person-level). This is the hardest step, and the biggest differentiator between providers.
3. Scoring & topic modeling
Behavior gets organized into topics (think: “electric vehicles,” “HRIS software,” or 36,000 more granular variations) and scored against historical baselines. A 10× spike in research activity on a topic relative to baseline is a strong intent signal.
4. Activation
Intent only matters if you can act on it. Modern intent platforms push audiences directly into the channels you already use, Meta, Google, LinkedIn, programmatic DSPs, CRM, sales tools, without manual exports.
04
Application
Five ways teams use intent data.
01
Top-of-funnel audience generation
Build paid media audiences from people actively researching topics related to your category. Replaces guesswork demographic and interest-based targeting.
02
Account-based marketing prioritization
Sort your target account list by who’s actively in-market right now. Sales reaches out when interest is highest, not on an arbitrary cadence.
03
Competitive conquest
Identify audiences researching named competitors. Run ads, send mail, or cue sales outreach at exactly the moment a switch is on the table.
04
Lead scoring & qualification
Layer external intent signals on top of internal activity to surface MQLs that are actually ready, not just nominally engaged.
05
Content personalization
Adapt website experience, email content, or ad creative based on the topics a visitor is researching elsewhere.
05
Evaluation
Choosing an intent data provider.
Most procurement decisions in this category go sideways for the same handful of reasons. Use this checklist before signing anything.
The buyer’s checklist
-
Where do the signals come from? If a vendor can’t name their sources clearly, walk away.
-
Is it account-level, person-level, or both? Match this to whether your motion is B2B ABM or person-based marketing.
-
What’s the activation surface? Data that doesn’t flow into your ad platforms or CRM is data that won’t get used.
-
How is privacy handled? Confirm GDPR, CCPA, and consent management, not just that they exist on paper.
-
What does refresh frequency look like? Daily refreshes beat weekly. Real-time beats both.
-
Can you trial it on your accounts? Demand a real test against accounts you already know and see if the signals match reality.
06
Reference
The intent data glossary.
Quick definitions for the terms you’ll see on every vendor site, RFP, and category report.
- Account-Level Intent
- Intent signals attributed to a company or domain rather than a specific person. The default unit of measurement in B2B intent data.
- Behavioral Intent
- Intent inferred from observed actions, page views, downloads, searches, as opposed to declared interest.
- Buyer Intent Score
- A composite score representing how likely an account or individual is to be in an active buying cycle for a given category.
- Contextual Intent
- Intent signals derived from the context of content being consumed in real time, without relying on third-party cookies or persistent identifiers.
- Custom Intent
- An intent topic or audience defined by the buyer, typically using their own keywords, competitor names, or product terms.
- Declared Intent
- Intent expressed explicitly by the buyer. For example, filling out a demo request or selecting a product category in a survey.
- First-Party Intent
- Intent collected from properties the buyer owns and operates, such as their website, app, or email program.
- Identity Resolution
- The process of stitching together fragmented behavioral signals to a single, persistent entity. A person or an account.
- In-Market
- An account or individual currently displaying behavior consistent with active buying research for a defined category.
- Intent Topic
- A discrete subject area used to group intent signals (e.g. “HRIS software” or “electric SUV.”) Top providers maintain tens of thousands.
- Lookalike Audience
- An audience modeled to share traits with a known seed audience. Lookalikes built from intent-qualified seeds outperform demographic seeds.
- Person-Level Intent
- Intent attributed to a single individual. Higher precision than account-level, but harder to collect compliantly.
- Predictive Intent
- Forward-looking intent scoring that uses historical patterns to forecast which accounts will enter a buying cycle.
- Privacy-Safe Intent
- Intent data collected and activated without third-party cookies or PII, typically using contextual or first-party methods.
- Real-Time Intent
- Signals delivered with little to no latency between event and availability, measured in seconds or minutes, not days.
- Surge Detection
- Algorithmic detection of unusual increases in intent activity for a specific topic relative to a baseline.
- Third-Party Intent
- Intent collected from publisher networks, review sites, or syndication partners outside the buyer’s owned properties.
- Zero-Party Data
- Information a person intentionally and proactively shares, like preferences, intentions, context. Adjacent to but distinct from intent data.
07
FAQ
Frequently asked questions.
Lead data tells you who someone is. Intent data tells you what they're thinking about right now. The two work best together — intent identifies who's in-market, lead data tells you how to reach them.
Yes — but the methodology matters more than ever. Contextual and first-party intent approaches don't depend on third-party cookies, which is why most modern providers have moved away from cookie-based identity stitching entirely.
Firmographics describe what an account looks like (industry, size, geography). Intent data describes what they're doing right now. Best-in-class targeting layers intent on top of firmographics.
Yes — modern intent platforms push directly into HubSpot, Marketo, Salesforce, Klaviyo, and most ad platforms. If you're still exporting CSVs, you're using the wrong tool.
Accuracy varies wildly by provider and methodology. The honest answer: trial it against accounts you already know are in-market and see if the signals match. Any vendor that won't let you do that probably knows their data won't hold up.
Pricing models range from per-topic, per-account, per-record, to flat-rate platform pricing. A useful framing: divide annual cost by expected pipeline influenced. If the ratio doesn't clear a 5× return inside 12 months, the math doesn't work.
Try it on your own data
Turn intent into audiences in one click.
Untitled gives performance marketers access to 36,000+ intent topics and pushes audiences directly into Meta, Google, TikTok, and Streaming TV. 14-day free trial. No credit card.
