ATS Hiring Signal Radar

Monitor target accounts on public Greenhouse, Lever, and Ashby boards, then turn job-posting changes into explainable hiring-intent signals for SDR, RevOps, and founder-led sales workflows.

This Actor helps prioritize when to investigate an account and what hiring evidence to reference. Hiring activity is a sales-research signal, not proof that a company will buy.

Quickstart: monitor 3 target accounts

Start with three public ATS boards and a hard company cap. A small run typically completes in under five minutes, but ATS availability can affect run time.

{
  "companies": [
    {
      "name": "Stripe",
      "domain": "stripe.com",
      "boardUrl": "https://boards.greenhouse.io/stripe",
      "tags": ["target-account", "fintech"]
    },
    {
      "name": "Spotify",
      "domain": "spotify.com",
      "boardUrl": "https://jobs.lever.co/spotify",
      "tags": ["target-account", "media"]
    },
    {
      "name": "Linear",
      "domain": "linear.app",
      "boardUrl": "https://jobs.ashbyhq.com/linear",
      "tags": ["target-account", "dev-tools"]
    }
  ],
  "changesOnly": false,
  "triggerPack": "all",
  "minConfidence": 0.3,
  "maxCompaniesPerRun": 3,
  "emitSignalsOnFirstScan": false,
  "snapshotNamespace": "starter-pack-sales-accounts"
}

What to expect

1. First run = baseline. The Actor records current jobs. With emitSignalsOnFirstScan: false, each successfully processed company is labeled as a baseline rather than a new signal.
2. Later runs = deltas. Reuse the same snapshotNamespace, set changesOnly: true, and run daily or weekly. Dataset rows appear only for companies whose jobs changed; RUN_SUMMARY is still written for the run.
3. Route completed runs. An Apify ACTOR.RUN.SUCCEEDED webhook can notify Make, Zapier, or your own endpoint, which can fetch the Dataset and RUN_SUMMARY before posting selected fields to Slack or a CRM.

Optional immediate demo vs. normal monitoring

The quickstart above is the recommended monitoring path: emitSignalsOnFirstScan stays false, so the first run is a clean baseline.

To inspect populated signal fields immediately, run one company once with emitSignalsOnFirstScan: true and a separate demo namespace:

{
  "companies": [
    {
      "name": "Stripe",
      "domain": "stripe.com",
      "boardUrl": "https://boards.greenhouse.io/stripe",
      "tags": ["target-account", "fintech"]
    }
  ],
  "changesOnly": false,
  "triggerPack": "all",
  "minConfidence": 0.3,
  "maxCompaniesPerRun": 1,
  "emitSignalsOnFirstScan": true,
  "snapshotNamespace": "starter-pack-immediate-demo"
}

Demo mode compares all current roles with an empty snapshot, so it can populate intentEvent, whyNow, and recommendedAngle while isFirstScan is still true. It is not a measured hiring delta or purchase-intent proof. Use the normal false path and a stable production namespace for scheduled monitoring.

Cost guardrail

At the current event prices, three successfully processed companies create up to $0.06 in company-scanned events. Actor-start events are separate and depend on selected memory (the current 4 GB default is about $0.0002 per run). These are estimates; check the current Pricing tab before scheduling larger account lists.

Why This Actor?

Raw job scrapers return listings. This Actor adds monitoring context.

Raw job data	This Actor
"Acme has 23 open roles"	"Acme added RevOps roles; evaluate a CRM or pipeline-tooling outreach hypothesis"
Job listing dump	Named hiring-intent event with confidence and evidence
One-time snapshot	Namespaced snapshots for scheduled change detection
Manual row filtering	signalScore, whyNow, and recommendedAngle fields for account research

No login or proxy is required for the three supported public ATS APIs. External provider availability and API behavior can still change.

Who Is It For

Primary buyer: B2B SDR, RevOps, or founder-led sales operators.

Use this Actor when your job is to monitor a defined target-account list and prepare an evidence-backed outreach hypothesis after public ATS changes. It is not positioned as a general-purpose recruiter job scraper.

Supported ATS Platforms

Platform	Method	Auth Required	Proxy Needed	Region
Greenhouse	Official Job Board API	No	No	Global
Lever	Official Postings API	No	No	US + EU
Ashby	Official Public API	No	No	Global

Only these three providers are supported in the current release. Workday, Rippling, BambooHR, and JazzHR are not currently supported.

Key Features

10 Intent Triggers Across 3 Packs

Pre-built detection patterns that convert raw job deltas into named, explainable events:

GTM Pack — Sales, Marketing, RevOps signals

Trigger	Fires When	Recommended Angle
first_revops_buildout	First Sales Ops / CRM / Salesforce role appears	CRM, pipeline visibility, forecasting
sales_team_expansion	3+ new sales roles posted	Sales enablement, lead gen, outreach tools
marketing_ramp	2+ new marketing roles posted	Marketing automation, ABM, analytics

Engineering Pack — Dev, Security, Data, AI signals

Trigger	Fires When	Recommended Angle
ai_platform_buildout	First ML/AI/LLM role appears	MLOps, GPU infra, vector DBs, LLM tooling
first_security_hire	First security role with no prior security team	SIEM, compliance, identity management
engineering_surge	5+ new engineering roles posted	Dev tools, CI/CD, cloud infra, observability
data_team_buildout	2+ new data roles posted	Data warehousing, ETL, BI platforms

Expansion Pack — Geo, Leadership, New Functions

Trigger	Fires When	Recommended Angle
geo_expansion	Jobs posted in previously unseen locations	Localization, compliance, local partnerships
leadership_hiring	2+ director/VP/C-level roles posted	Org restructuring, strategic planning
new_function_launch	Entirely new function (e.g. legal, finance) appears	Function-specific tooling, consulting

Snapshot Diffing

Automatically stores and compares job snapshots between runs:

• New roles added since last scan
• Roles removed (filled or closed)
• Net headcount change
• Function distribution shifts (engineering vs. sales vs. marketing...)

Signal Scoring

Each company receives a 0-100 signal score:

• high (70+) — Strong hiring momentum with clear intent triggers
• medium (40-69) — Notable changes worth monitoring
• low (15-39) — Minor activity detected
• none (0-14) — Stable or no meaningful changes

First Scan Behavior

For normal monitoring, keep emitSignalsOnFirstScan: false. The first scan stores a baseline without signals; later runs compare against that snapshot.

Setting emitSignalsOnFirstScan: true is an optional demonstration mode. It evaluates the current board against an empty snapshot and may populate signal fields immediately, but the row remains a first-scan baseline and is not a real period-over-period delta.

Input

Full Input Reference

Field	Type	Default	Description
companies	array	required	List of companies to monitor. Each needs name and ideally boardUrl.
changesOnly	boolean	false	Only output companies with changes since last run. Ideal for scheduled alerts.
triggerPack	string	"all"	Which triggers to run: all, gtm, engineering, or expansion.
minConfidence	number	0.3	Only report signals above this confidence (0.0 - 1.0).
maxCompaniesPerRun	integer	100	Cap companies per run to control cost. Max 1000.
emitSignalsOnFirstScan	boolean	false	Emit signals even on the first scan (normally suppressed for baseline).
snapshotNamespace	string	auto	Custom namespace for snapshot storage. Useful for separate monitoring contexts.

Company Object

Field	Type	Required	Description
name	string	Yes	Company name
domain	string	No	Company domain (e.g. stripe.com). Used for auto-detection.
boardUrl	string	No	Direct ATS board URL. Recommended for reliable detection.
tags	string[]	No	Custom tags for filtering (e.g. "competitor", "prospect", "portfolio")

Scheduled Monitoring Input

Use this input for the first baseline run. After it succeeds, keep the same snapshotNamespace and change only changesOnly to true for the scheduled Task.

{
  "companies": [
    {
      "name": "Stripe",
      "domain": "stripe.com",
      "boardUrl": "https://boards.greenhouse.io/stripe",
      "tags": ["target-account", "fintech"]
    },
    {
      "name": "Spotify",
      "domain": "spotify.com",
      "boardUrl": "https://jobs.lever.co/spotify",
      "tags": ["target-account", "media"]
    },
    {
      "name": "Linear",
      "domain": "linear.app",
      "boardUrl": "https://jobs.ashbyhq.com/linear",
      "tags": ["target-account", "dev-tools"]
    }
  ],
  "changesOnly": false,
  "triggerPack": "all",
  "minConfidence": 0.3,
  "maxCompaniesPerRun": 3,
  "emitSignalsOnFirstScan": false,
  "snapshotNamespace": "starter-pack-sales-accounts"
}

Output

Each successfully processed company produces one Dataset record unless changesOnly: true filters an unchanged company. Skipped or failed companies do not produce a result row.

Output Fields Reference

Field	Type	Description
company	string	Company name
domain	string	Company domain
tags	string[]	User-provided tags
provider	string	ATS provider (greenhouse, lever, ashby)
slug	string	ATS board slug
boardUrl	string	Full ATS board URL
intentEvent	string	Primary trigger ID (e.g. first_revops_buildout) or no_signal
intentLabel	string	Human-readable trigger label
signalScore	number	0-100 composite signal score
signalStrength	string	high, medium, low, or none
confidence	number	Confidence of the primary signal (0.0 - 1.0)
whyNow	string[]	Evidence array explaining why this signal fired
recommendedAngle	string	Suggested outreach angle or use case
allSignals	object[]	All detected signals (not just primary)
newRoles	number	New roles since last scan
removedRoles	number	Removed roles since last scan
netChange	number	Net headcount change
totalOpenRoles	number	Current total open roles
previousOpenRoles	number	Total open roles at last scan
isFirstScan	boolean	Whether this is the first scan for this company
functionDistribution	object	Role count by function (engineering, sales, data...)
newJobDetails	object[]	Detailed new job info (only in changesOnly mode)
scannedAt	string	ISO timestamp of this scan
snapshotAge	string	ISO timestamp of previous snapshot

Actual Baseline Output (public latest smoke)

The proposed three-company quickstart was run on public build 1.0.14 / v38fugaIfthTTrCTI on 2026-07-14. This selected-field Dataset row shows normal first-run semantics:

{
  "company": "Linear",
  "provider": "ashby",
  "intentEvent": "no_signal",
  "signalScore": 0,
  "whyNow": [],
  "recommendedAngle": "",
  "newRoles": 24,
  "totalOpenRoles": 24,
  "isFirstScan": true
}

Actual Immediate-Demo Output (unchanged runtime, local validation)

A one-company validation run used emitSignalsOnFirstScan: true. The current Stripe board was evaluated against an empty snapshot and produced these selected fields:

{
  "company": "Stripe",
  "provider": "greenhouse",
  "intentEvent": "sales_team_expansion",
  "intentLabel": "Sales Team Expansion",
  "signalScore": 84,
  "signalStrength": "high",
  "confidence": 0.95,
  "whyNow": [
    "Account Executive, AI Sales (Grower) posted",
    "Account Executive, Commercial Grower (Japanese Fluency) posted",
    "Account Executive, Commercial Hunter (Japanese Fluency) posted",
    "Account Executive, Cross Border China posted",
    "Account Executive - Enterprise, Grower posted"
  ],
  "recommendedAngle": "Sales enablement, lead generation, outreach tooling",
  "newRoles": 523,
  "totalOpenRoles": 523,
  "isFirstScan": true
}

These values are time-sensitive public job-board observations, not guaranteed future output. The demo row is still a baseline; it is not a later-run delta or proof of buying intent.

The following actual RUN_SUMMARY came from the public three-company baseline smoke and was stored separately in the Key-Value Store:

{
  "totalCompanies": 3,
  "processed": 3,
  "skipped": 0,
  "signalsDetected": 0,
  "errors": 0,
  "triggerPack": "all",
  "changesOnly": false,
  "completedAt": "2026-07-14T07:56:50.465Z"
}

Schedule Weekly Monitoring

Use one Apify Task so the input and snapshot context remain stable:

1. Create a Task with the Starter Pack input and a unique, stable snapshotNamespace for this account list.
2. Run the Task once with changesOnly: false. This baseline run records current roles. A normal first run does not imply that no hiring exists; it means no prior snapshot was available for comparison.
3. Change only changesOnly to true. Keep the same company list and snapshotNamespace.
4. Attach a daily or weekly Apify Schedule to that Task. A Schedule is time-based; it does not wait for a separate upstream workflow.
5. Add an ACTOR.RUN.SUCCEEDED webhook to Make, Zapier, or your own endpoint. The receiver can fetch:
   • Dataset items for companies with detected changes;
   • RUN_SUMMARY for processed/skipped/signal/error counts;
   • intentEvent, signalScore, whyNow, and recommendedAngle for Slack or CRM formatting.
6. Treat an empty Dataset as “no changed companies for this run,” not as a failed run. Check RUN_SUMMARY and the run status before deciding whether to notify a sales channel.

Use a separate snapshotNamespace for each distinct account list or monitoring workflow. Raising minConfidence to 0.5 can reduce lower-confidence signal rows, but does not guarantee sales relevance.

Pricing

This Actor uses pay-per-event (PPE) pricing with two existing events:

• company-scanned: $0.02 for each company successfully processed;
• apify-actor-start: $0.00005 per GB when the Actor starts (one event per GB, minimum one).

The company-scanned event is not charged for a company that is skipped or fails. The Actor-start charge remains separate.

Company outcome	company-scanned events
Company scanned and a result returned	1
ATS board not found or unsupported	0
changesOnly mode with no changes	0
Suspicious data drop awaiting confirmation	0
Processing error	0

Estimated company-scan cost at $0.02/event

Companies	Frequency	Estimated events/month	Estimated company-scan cost
50	Daily	~1,500	~$30/month
100	Daily	~3,000	~$60/month
200	Daily	~6,000	~$120/month
200	Weekly	~800	~$16/month

These estimates assume every listed company is successfully processed on every run. Actor-start events and any current platform charges are separate; verify the Pricing tab before enabling a large schedule.

Changelog

v1.0.0

• Initial release
• Greenhouse, Lever (US + EU), Ashby support
• 10 intent triggers across GTM, Engineering, Expansion packs
• Snapshot diffing with signal scoring
• First-scan baseline capture (false positive prevention)
• Retry/timeout handling with exponential backoff
• Namespaced snapshot storage for multi-context monitoring