Home › Guides › ATS hiring signals

Hiring signals from ATS job boards: what companies tell you before they announce it

Every company that hires runs an applicant tracking system, and almost every ATS hosts the careers page on a public URL. That makes a company's open roles the most honest, earliest-available signal of where its budget is moving — months before the funding announcement or the product launch. This guide covers how to detect which ATS a company uses, what each board actually exposes, and how to turn a job list into an intent score.

Why job boards beat most intent data

Most "buying intent" products infer interest from web traffic or content consumption. Job posts are different in three ways:

• They are declarations, not inferences. A company posting a Director of Revenue Operations role has decided to build that function. There is no probability model in between.
• They are early. You hire the team before you launch the initiative. The job post precedes the announcement by a quarter or more.
• They are budget-bearing. Headcount is the biggest line item in most companies, and each new function drags tooling spend behind it.

The catch is that this data is scattered across seven platforms and nobody tells you which one a given company uses.

The seven boards, and how they differ

If you already know the platform, the dedicated scraper gives you its deepest fields. If you do not — the normal case when working from a target list — an ATS-detecting scraper resolves the domain first and then returns roles using the same schema whichever board it lands on. Outputs from all eight concatenate into one table.

What a job board actually gives you

Across the seven systems, fields fall into three tiers. These percentages come from our own measured runs, not from vendor claims:

A worked row from a real run looks like this:

{
  "companyName": "Ramp",
  "jobTitle": "Security Engineer, Cloud",
  "department": "Engineering",
  "location": "New York, NY (HQ)",
  "isRemote": true,
  "employmentType": "FullTime",
  "postedAt": "2026-04-07",
  "openRoleCount": 148
}

Note openRoleCount: the company's total, carried on every row. It lets you compute intent without pulling the whole board.

Two data-quality traps

Job board data looks clean and often is not. Two problems we hit and fixed are worth knowing about, because they affect anyone reading these boards:

1. Entity-encoded HTML in descriptions. Some boards deliver the description as escaped HTML (<div>). If you decode entities after stripping tags, the decode step recreates tags that never get removed — and every description ships full of visible markup. In one measurement, 50 of 50 rows were affected before the fix; 0 of 50 after. Decode first, strip second.
2. Prefixed department names. Greenhouse prefixes department names with an internal numeric id — 8611 Security Analytics. Half the rows in our sample carried that prefix. If you group by department without stripping it, the same team splits into multiple buckets and your intent score is wrong.

Turning roles into an intent score

Raw counts are noise. A usable score combines three things:

1. Department mix. Map roles to functions and ask which function is growing. Sales hiring predicts sales-tool purchases; data-engineering hiring predicts infrastructure spend.
2. Velocity. Sample the same board weekly. New roles this week versus the trailing average separates expansion from backfill — a company replacing departures looks identical to one expanding if you only look once.
3. Firsts. The first role of a kind is the strongest signal in the dataset. A company's first security engineer, first RevOps hire or first role in a new city marks a threshold being crossed.

Join the result to firmographics on the company domain and you have a target list ranked by what a company is actually doing, not by what a database says it is. Company Enrichment supplies the firmographic side; LinkedIn Hiring Signals gives a second, independent read on the same companies.

Regional coverage

The seven platforms skew Anglo-American and European. For other markets the signal exists but lives elsewhere — in India, for example, tech hiring concentrates on local boards, which is why India Tech Hiring Signals reads Instahyre and reports company-week hiring velocity rather than individual postings. The principle transfers; the source does not.

Combining hiring with other timed signals

Hiring is one of several public signals that carry a date, and the dated signals are the ones worth building a workflow around. Each sits at a different distance from a buying decision:

The combination beats any one of them. A company that raised recently and is hiring for the function you sell into is a materially stronger prospect than one doing either alone — and both facts are public, dated, and free to observe.

Frequently asked questions

How do you find out which ATS a company uses?

Resolve the company domain to its careers page and inspect where the job list is served from. Most platforms host or embed from recognisable domains. Automating this matters because you rarely know in advance which of the seven a target runs.

Why is hiring data a buying signal?

Headcount is the largest budget line, and teams buy tools for the people they hire. Job posts also appear months before public announcements, because the team has to exist before the initiative launches.

Why do posting dates go missing?

Not every ATS publishes one. Around 75% of rows in our runs carry a posting date; the rest simply do not expose it. Build the score so a missing date degrades it rather than breaking it.

Is it legal?

Company job boards are published for candidates and indexed by search engines. The rows are role listings and company-level activity — business information. Descriptions occasionally name a hiring manager, which is content the employer chose to publish.

Detect any ATS & pull roles → See all 87 datasets