Intent data

Updated November 10, 2026

Intent data is behavioral data that suggests a company or person is researching a topic, product or category, and may be moving toward a purchase. It comes from activity such as website visits, content reads and searches, collected on a seller's own properties or across third-party sites, and sales and marketing teams use it to decide which accounts to prioritize.

In short: Intent data is behavioral data, such as web visits, content consumption and searches, that sellers use to estimate which companies are researching a product category and to prioritize accounts for outreach.

How intent data works

For a sales team, intent data tries to answer which companies are researching this kind of product right now. It does that by watching behavior and inferring interest from it.

IBM describes intent data as the behavioral signals that indicate how likely a prospect is to make a purchase, including website activity, social media engagement, content consumption and search patterns 4. Gartner's Rick LaFond describes the same idea from the marketing side, with brands using engagement or behavioral data to anticipate a customer's journey needs 1.

A Salesforce overview of intent data describes four common types 2.

First-party intent data. Activity on the seller's own website and properties, such as visits, email opens and form submissions. It is the most accurate type but limited in scope.
Third-party intent data. Activity gathered from external publishers, review sites and industry websites. It gives a broader view but is less precise.
Keyword intent data. The search terms companies use, which can hint at buying stage.
Topic-based intent data. Engagement with content on specific subjects related to the product.

Third-party intent tools work in steps. They collect activity, identify the company behind it and then score it. The same Salesforce overview says the company is identified through IP address, device tracking and database matching, and that an algorithm assigns a score from 1 to 100 2.

Teams use the output to prioritize. An account with a high score moves up the list for outreach or ads, and the score often feeds lead scoring alongside fit data such as firmographics.

Intent data is closely tied to account-based programs. In Gartner's guidance on account-based marketing, behavioral data helps teams anticipate the journey needs of target accounts 1. In that setting the account list comes first, and intent helps decide which listed accounts deserve attention now.

What decides whether intent data is useful

Intent data is an inference, so its value depends on how good the inference is. Four variables matter most.

The identification method comes first. Matching anonymous activity to a company through IP addresses and device data is imperfect 2. Remote work, shared networks and privacy controls all weaken the match. A score attached to the wrong company is worse than no score.

The source comes second. First-party data shows activity on the seller's own site, which makes it more accurate but narrower 2. Third-party data shows research across the web, which is broader but less precise 2. Teams should know which kind they are buying and how the vendor collects it.

Timing comes third. Interest fades fast. A surge from last month may describe a decision already made. Like other records, signals suffer from data decay, so every score should carry a date and an expiry.

Transparency comes fourth. A number with no explanation is hard to act on. A rep can use an observed fact, such as a visit to the pricing page, in a message. A rep cannot easily use a high score without knowing what produced it.

Privacy rules apply as well. Tracking individuals across sites is regulated differently in different countries, so teams should check how a vendor gathers data and confirm it fits the rules where buyers are located.

Example of signals used where intent data is not available

A team without an intent data contract still wants a monthly list of accounts that look active. It uses public activity signals instead, following RichAPI's signal-based prospecting recipe 3. RichAPI states plainly that it does not sell intent data, with no bidstream, no IP-to-company de-anonymization and no topic surge score 3.

text

Monthly signal run
  linkedin_job_search                  job posts for the roles that use the product
  dedupe to companies                  keep the post date on each account
  linkedin_company_employees_search    find the people who own the problem
  email_finder on those people         misses return billed false
  email_verifier on the hits           route on the verdict
  expire stale signals                 posts after two weeks, jobs after a month

The recipe prices one weekly run. Searching 200 job posts at 0.1 credits per result bills 20 credits, and the posts dedupe to about 60 companies 3. Finding 3 people per company returns 180 results at 0.1 credits each, or 18 credits 3. The search legs bill per result on any successful response, including an empty one 3.

The recipe draws a clear line between a signal and intent. A company hiring implementation engineers is evidence about that company, and the inference about buying is the reader's to make 3. The output is a smaller account list for enrichment and outreach.

Signals expire on a schedule. The recipe suggests windows of two weeks for posts, a month for jobs and a quarter for ads, because a message built on a stale trigger reads worse than no message 3. The team stores each signal with its timestamp so expired accounts drop out of the next run automatically.

Intent data vs public activity signals

Both help decide which accounts to contact first. Intent data infers research behavior, often from tracking, while public activity signals record things a company did in the open, such as posting a job.

	Intent data	Public activity signals
What it answers	Which companies seem to be researching this category?	What did this company just do in public?
Input	Web visits, content reads and searches matched to companies	Job posts, ad libraries and company posts
Output	An intent score or surge by topic	A dated event tied to a named company
Cost driver	Vendor subscriptions, often by topic or volume	Per-result searches and the enrichment that follows
Main risk	Wrong company matches and unexplained scores	Reading buying intent into an unrelated event
Use it when	Ranking a large fit list by apparent research	Building a short list with reasons a rep can cite

Read the reference →

Frequently asked questions

What is the difference between first-party and third-party intent data?

First-party intent data comes from activity on a company's own website, emails and forms. Third-party intent data comes from activity across other sites, such as publishers and review sites, gathered by a vendor 2. First-party data is more precise, and third-party data covers more of the market.

How accurate is intent data?

It varies by vendor and method. Matching anonymous activity to a company is imperfect, and scores can reflect research by students, competitors or job seekers. Test a vendor against accounts with known outcomes before relying on its scores, and treat a score as a reason to look rather than proof.

Is a job post intent data?

Not in the usual sense. A job post is a public, dated fact about a company, and it suggests growth or a gap. Whether it means the company will buy a product is an inference the seller makes. Many teams use such signals alongside or instead of intent scores.

How should sales teams use intent data?

Use it to prioritize accounts that already fit the target profile, after the list is built from fit criteria. Route high-intent fit accounts to outreach sooner, refer to observable facts in messages rather than the score, and expire old signals so reps do not act on stale interest.

The bottom line

Intent data is behavioral data that suggests which companies are researching a product category, gathered on a seller's own properties or across third-party sites. It works best as a way to prioritize accounts that already fit. Check how a vendor identifies companies, keep every signal dated with an expiry and prefer facts a rep can cite over unexplained scores.