SiteVerdict — Hiring Signal & Buying Intent Scorer

Know which accounts are spending right now. Paste company domains or names and get a 0–100 hiring score, growth velocity, the teams that are growing, the tools they name in job ads, and an explained "buying intent" score for CRM, data, cloud, security, AI or your own category — straight from the companies' own public job boards.

No LinkedIn. No Indeed. No scraping of job sites. Only the official public job-board APIs that Greenhouse, Lever, Ashby, Workable and SmartRecruiters run for careers pages.

Three ways teams use it

1. Prioritise outbound. Score 500 target accounts every Monday; work the "hot" ones first. Each row says why in plain English ("4 open roles mention Salesforce/HubSpot; 2 are dedicated hires: 'Salesforce Administrator', 'RevOps Manager'; 3 posted in the last 30 days").
2. Catch the moment. Monitor mode remembers last week's roles and tells you when an account opens a burst of new roles or hires its first Head of Data — with a Slack or webhook alert.
3. Research a market. Which fintechs are hiring security engineers? Which SaaS companies are building data teams on Snowflake vs BigQuery? The tools-mentioned table answers it.

Before / after

You give:

{ "companies": ["ramp.com", "greenhouse:discord", "https://jobs.lever.co/spotify", "Hugging Face"], "intentTopics": ["crm", "data-platform", "security", "ai-ml"] }

You get (one row per company, trimmed):

{
  "company": "Discord",
  "hiringScore": 64, "hiringLevel": "hot",
  "openRoles": 49, "postedLast30Days": 17, "velocityLabel": "accelerating",
  "departments": [{ "department": "Engineering", "roles": 21, "share": 42.9 }, { "department": "Data & AI", "roles": 6, "share": 12.2 }],
  "leadershipRoles": 3,
  "toolsMentioned": [{ "tool": "Python", "roles": 9 }, { "tool": "Kubernetes", "roles": 4 }],
  "intent": [{ "topic": "ai-ml", "score": 81, "level": "high", "explanation": "Buying intent for AI / machine learning: HIGH (81/100). 9 open roles mention machine learning, llm ... 4 are dedicated hires ('Senior ML Engineer', ...)." }],
  "changes": { "newRoles": 6, "closedRoles": 3, "previousScore": 55, "scoreDelta": 9 },
  "alertTriggers": ["AI / machine learning intent 81 >= 70"],
  "highlights": ["49 open roles, mostly Engineering (21).", "Hiring is accelerating (17 new in 30 days vs 8 in the 60 days before)."]
}

(Numbers above illustrate the format; run sample mode to see the exact output on recorded real boards.)

What is scored

Signal	How
Hiring score (0–100)	volume (log-scaled open roles), velocity (roles posted in the last 30 days vs the 60 before), breadth (departments hiring), leadership hires (director/VP/C-level), and — in monitor mode — new roles since the last run. Weights are in the output (scoreBreakdown).
Velocity	accelerating / steady / slowing / unknown (when the board does not publish posting dates).
Departments & seniority	from job titles (sales engineers count as Sales, VPs as Executive, etc.) with the vendor department as a fallback.
Tools mentioned	70+ tools (Salesforce, HubSpot, Snowflake, dbt, AWS, Kubernetes, Datadog, Okta, OpenAI…) matched as whole words; ambiguous names (React, Rust, Spark, SAP, Claude) only when written as names.
Buying intent per topic	dedicated hires (topic in the title) weigh 3×, passing mentions 1×, recent postings +25%; saturating 0–100 score with level, evidence (up to 5 roles) and a one-line explanation. 11 built-in topics + your own ("Video: Mux, transcoding").
Changes	new/closed roles and score delta vs the previous run under the same monitor name.
AI brief (optional)	with YOUR Anthropic key: a 120-word sales brief per company (why now, growing teams, product fit, an opening line).

Input

Field	Example
companies	ramp.com, https://jobs.lever.co/spotify, greenhouse:discord, Hugging Face
intentTopics	crm, marketing-automation, data-platform, cloud-devops, security, ai-ml, customer-support, finance-erp, people-hr, payments, expansion
customIntentTopics	Video infrastructure: Mux, video streaming, transcoding
monitorStateName + onlyChanged	weekly schedule that only returns (and only charges for) accounts that changed
slackWebhookUrl / webhookUrl	alert when hiring score ≥ 75, any intent ≥ 70, or ≥ 5 new roles (all configurable)
enableAiBrief + anthropicApiKey	optional AI brief on your own key
sampleMode	try it free on recorded real job boards

Output and exports

• Dataset views: Hiring scores (CSV/Excel friendly), Buying intent (explained), No board / errors.
• OUTPUT record: run summary, leaderboard sorted by score, alert delivery status, charged events — an agent-ready JSON summary.

Pricing (pay per event)

Event	Price	When
company-scored	$0.02	a company with a readable public job board was scored
ai-explanation	$0.01	an AI brief was produced (your Anthropic usage is billed to your key)

Max cost per run: company data first. Each requested company's company-scored charge is held back before anything else. An AI brief (only when enableAiBrief is on and you supplied your Anthropic key) is generated only if the budget left after holding one company charge per still-unprocessed company also covers the brief. Inputs that turn out free (no public board, errors, unchanged) release their hold as soon as they are checked, so briefs later in the run can use that money. If a brief cannot be covered, it is not generated (your Anthropic key is not used) and not charged. The company row is still delivered in full with aiBrief: null, aiBriefStatus: "skipped_budget" and an aiBriefNote, and OUTPUT shows aiBriefsSkippedBudget; the run is not reported as stopped. A failed AI call is never charged either (aiBriefStatus: "error"). Only when the budget cannot cover every company does the run stop (stoppedAtChargeLimit: true); each company it could not score gets a free not-processed row, so every requested company appears in the dataset.

Not charged: companies without a public board, errors, unchanged companies in onlyChanged mode, sample mode. Examples: 100 accounts ≈ $2.00; a weekly monitor of 500 accounts where ~80 change ≈ $1.60/week; 100 accounts with AI briefs ≈ $3.00 + your Anthropic usage.

FAQ

Which companies work? Any company whose careers page uses Greenhouse, Lever, Ashby, Workable or SmartRecruiters (very common for tech, SaaS, fintech, scale-ups). Companies on Workday, iCIMS, Taleo, SuccessFactors or their own systems return a free "no-board" row.

How is the board found? Direct identifiers/URLs are used as given. For domains we read the home and careers pages (robots.txt respected, max 5 pages) for a linked ATS board, then — if allowed — probe the ATS APIs with a slug from the domain or name. Probe matches say so in discoveryNote; check the company name.

Is a high intent score proof of a purchase? No. It means the company is hiring people who will use or own that category. Treat it as a reason to reach out, not a fact.

Do you collect personal data? No. We read job postings (organisational data). Recruiter names are never collected. The AI step sends only computed signals and job titles.

Limits

• Only the five ATSs above. Unlisted/confidential roles are not visible.
• Posting dates are vendor-reported; some boards (e.g. Workable) publish fewer dates, which makes velocity unknown.
• SmartRecruiters descriptions need one detail call per posting (default max 25 per company).
• Website discovery skips pages blocked by robots.txt and never bypasses bot protection.

Sources & compliance

• Greenhouse Job Board API (boards-api.greenhouse.io), Lever Postings API (api.lever.co), Ashby Posting API (api.ashbyhq.com/posting-api), Workable public jobs widget API (apply.workable.com/api/v1/widget), SmartRecruiters Posting API (api.smartrecruiters.com/v1/companies/{id}/postings). These are the public, unauthenticated endpoints the vendors document/provide for publishing a company's own job listings on its careers site. Terms/doc URLs checked 11 Oct 2026: developers.greenhouse.io/job-board.html, github.com/lever/postings-api, developers.ashbyhq.com/docs/public-job-posting-api, help.workable.com (jobs widget), developers.smartrecruiters.com/docs/posting-api.
• Company websites are read only to find the job-board link: honest User-Agent SiteVerdict/hiring-signal-scorer (+hello@webrevenuescan.com), robots.txt respected, SSRF-guarded, low page limits.
• Never used: LinkedIn, Indeed, Glassdoor or any other job site.
• UK GDPR / PECR: output is organisational data about companies. If you use it for outreach, you are the controller: have a lawful basis (usually legitimate interests for B2B), respect PECR rules for electronic marketing to individuals and sole traders, and honour opt-outs.
• Disclaimer: hiring signals are inferences from public job ads, not facts about budgets or purchases. Not financial or investment advice.

Built by SiteVerdict (BharatFare Travel and Technologies Ltd, UK). Contact: hello@webrevenuescan.com.