---
title: "How does intent data work?"
description: "**Canonical URL:** https://jobspipe.dev/answers/how-intent-data-works"
canonical: https://jobspipe.dev/answers/how-intent-data-works
last_updated: 2026-09-18
---

# How does intent data work?

> Intent data works by collecting signals of research or spending behaviour, tying each signal to a company, mapping it to a topic, and scoring whether that company's interest is rising above its normal level. The result is a list of accounts showing unusual activity around subjects you sell into, delivered to a CRM or advertising platform so sales and marketing can prioritise them.

**Canonical URL:** https://jobspipe.dev/answers/how-intent-data-works
**Topic:** Hiring signals (https://jobspipe.dev/intent-data-providers)

The pipeline has four steps, and each one limits what the output means. Collection gathers events: visits to your own site, activity on review sites, article reads across publisher networks, or public actions such as posting a job. Resolution ties an anonymous event to an organisation, usually from a network address, a cookie or an email domain, and remote work and shared networks make this the weakest link. Classification assigns content to a topic taxonomy, so the quality of the topics decides how specific the signal can be. Scoring compares recent activity with the account's own baseline and flags a surge. What comes out is an account-level probability, not a named buyer and not a confirmed project. A surge can be a student, a competitor or one curious employee, which is why teams combine sources before acting.

Hiring signals skip the two weakest steps. A job posting names the company, so nothing has to be resolved, and it states the function, seniority and tools in plain text, so little has to be inferred. JobsPipe exposes postings through POST /v1/jobs/search: company_name_or watches an account list, job_title_or and skills_or define the pattern, and posted_at_gte returns only what is new since the last run. Webhooks and the MCP server deliver the same matches to a workflow or an agent.

To put intent data to work, choose the buying stage you most need to see, pick a source type that covers it, and check every score against closed deals before trusting it. For hiring signals, run a search against your target accounts in the sandbox, which needs no key, then schedule it on the free tier. The patterns worth watching are described at jobspipe.dev/hiring-signals-api.

## Related questions

### How is intent data collected?

From tracking on a vendor's own properties, from review sites and publishers that record what their audiences read, from networks of publishers that pool that behaviour, from advertising auction data, and from public sources such as job postings. Each method covers a different part of the market, and providers usually blend more than one into the scores they sell.

### How accurate is intent data?

It varies by source. Review-site activity is close to a purchase decision and tends to be the most actionable. Topic surges from content consumption are broader and noisier, because reading is cheap and company matching is imperfect. Hiring signals are declarative and low in noise but lead a purchase by longer. Test any source against your own won and lost deals.

### What is an intent surge or spike?

A surge is a period when an account's activity on a topic rises well above that account's usual level. Providers compare against a baseline so that a large company that always reads a lot does not swamp a small one that has just started researching. A surge shows a change in attention. It does not confirm that a buying project exists.

### How is a hiring signal different from behavioural intent?

Behavioural intent infers interest from what people at a company read or compare. A hiring signal is a statement the company made itself: it published a role, under its own name, with budget behind it. It fires earlier and needs no identity matching, but it only covers needs that show up as hiring. The JobsPipe page on hiring signals lists the patterns.

## More answers

- [Where does intent data come from?](https://jobspipe.dev/answers/where-does-intent-data-come-from)
- [What is headcount?](https://jobspipe.dev/answers/what-is-headcount)
- [What companies are hiring?](https://jobspipe.dev/answers/what-companies-are-hiring)

Data behind these answers is published as shares only; every figure can be reproduced against the JobsPipe API (https://jobspipe.dev/jobs-api).

---
AI agents: the full machine-readable index of this site is https://jobspipe.dev/llms.txt?src=md-twin - API quickstart, no-key sandbox (POST https://api.jobspipe.dev/v1/sandbox/jobs/search), MCP server, pricing. Free key: https://jobspipe.dev/signup