People Data Labs

What it is

People Data Labs (PDL) is a developer-first people and company data provider. It sells the same graph two ways: as REST APIs (person and company enrichment, search, identify, plus free cleaner and autocomplete endpoints) and as bulk Data License feeds delivered to S3, GCS, Azure, Snowflake or Databricks. Its October 2026 release notes (v36.0) count 817.8 million resume records, 76.9 million company records and 51.1 million job postings, with 6.1 million job changes detected that month; the full person dataset sat at 2.42 billion records after a September deduplication pass removed about 40 million duplicate profiles. PDL is independent and venture-backed, and it ships versioned monthly releases, v34.0 in April 2026 through v36.0 in October.

What it is not: a prospecting app or an AI agent product. There is no list-building UI for reps, no sequencer and no official MCP server. PDL is the raw layer that other products, Clay among them, are built on.

Why it shows up in RevOps stacks

• It sits under the enrichment waterfall. Clay ships a built-in PDL connection and a bring-your-own-key option, and PDL took Clay’s “OG Data Partner” award in 2026. Many teams already pay for PDL data inside another tool without knowing it.
• Published per-record prices. Pro starts at $98 a month for 350 person credits, about $0.28 a match, and annual billing takes 20% off. Few people-data vendors publish a number you can model before a sales call.
• You pay for matches, not attempts. Enrichment costs 1 credit per successful match and misses are free, so a CRM backfill with a 60% hit rate pays for 60% of the rows.
• Quality signals you can filter on. PDLScores went GA in the July 2026 release (v35.0): profile_score flags whether a profile looks like a real person and activity_score whether it is maintained. Filtering out low-score profiles before outreach is cheaper than bouncing emails.
• Warehouse-native delivery. For data teams, a Snowflake or Databricks share beats calling an API row by row, and it is the only way to get the IP and job-posting datasets, which are Enterprise-only.

Pricing

Checked 2026-11-06 on peopledatalabs.com/pricing and the PDL billing docs:

• Free: up to 100 person credits and 100 company records a month. Emails, phones and addresses are obfuscated, so the free tier tests match rate, not contactability.
• Pro (self-serve): person data from $98 a month for 350 credits; company data from $100 a month for 1,000 records. Contact fields and CSV enrichment are included; premium field bundles are add-ons. Volume tiers lower the per-credit price as usage grows, but the tier rates are not published.
• Enterprise: custom annual contracts, all APIs, Data License feeds, additional field bundles and SSO/SAML.
• Billing rules that move the bill. Enrichment: 1 credit per match. Bulk enrichment: up to 100 credits per call, 1 per match. Search: 1 credit per profile returned. Person Identify: 1 credit per request, including no-match responses.
• What it costs in practice. Enriching 10,000 CRM contacts at a 65% match rate is 6,500 credits, about $1,800 at the entry Pro rate and less on annual or volume tiers. The same budget pulled through the Search API returns 6,500 profiles, matched to your filters or not. Model searches and enrichments as separate line items.

Best for

• RevOps engineers and data teams at Series A-to-enterprise companies who want to enrich CRM, warehouse or product data at a known per-record price
• Product teams embedding people or company data in their own app, who need a bulk licence and a redistribution clause rather than seats
• Clay-heavy GTM teams that want to bring their own PDL key to cut per-row Clay credit costs on high-volume tables
• Not for reps who want a UI to build lists and send sequences (use Apollo); not for verified mobile numbers at scale (use a waterfall like FullEnrich); not for real-time job-change webhooks and live re-fetch (use Crustdata)

Alternatives — and when to pick them instead

• ZoomInfo — the category’s share leader, with intent data, org charts and a rep-facing UI on custom contracts around $33,500 a year at the median (Vendr). Pick ZoomInfo when sales reps are the main users. Pick PDL when code or a warehouse is the consumer and per-seat pricing makes no sense.
• Apollo — the other share leader: database, sequencing and dialer from $49 a seat a month on annual billing. Pick Apollo when people do the prospecting; pick PDL when you only need the data.
• Crustdata — a younger agent-first data API, with live endpoints, social posts, Watcher webhooks and a hosted MCP server. Pick Crustdata when an AI agent calls the data and freshness at the moment of outreach matters. Pick PDL when you want published prices, SOC 2 and a mature bulk licence.
• Coresignal — self-serve from $49 a month with large historical employee and company datasets. Pick Coresignal when historical headcount and job-history depth is the use case; pick PDL when you need broad enrichment and identity resolution.
• Clay — the fastest-growing platform in the segment (a $115M Series D at a $7.1B valuation in September 2026), and less a rival than a layer on top: Clay runs PDL plus dozens of other providers as a waterfall. Pick Clay when a person designs the enrichment logic in a table and wants CRM sync. Buy PDL directly when one provider covers most rows and you can write the code.

Watch-outs

• Search bills per profile returned. A loose Search API query that returns 5,000 profiles costs 5,000 credits, whether or not they fit your ICP. Guard: test every query against the free sandbox first, set a size cap on every request, and use enrichment on known identifiers whenever you have an email, domain or profile URL.
• Credits don’t buy fresh data. PDL rebuilds its dataset in monthly releases, so a record can trail a job change by weeks. Guard: filter on activity_score and the record’s job-change timestamp, and re-verify any email through a verifier before it enters a sequence.
• Data-broker exposure lands on you. PDL is a registered data broker, and California’s Delete Act requires brokers to process deletion requests from the state’s DROP platform from August 2026. Your copy of the data inherits that obligation. Guard: wire PDL’s free Subject Request API into your deletion flow, and keep PDL record IDs next to stored records so you can purge them when PDL does.
• Field schemas change on a schedule. v36.0 deprecated funding_details.investing_companies and affiliated_profiles, which will be removed on 2027-02-16. Guard: pin the API version in your client, subscribe to the changelog, and add a monthly schema-diff check to whatever pipeline consumes PDL fields.
• No official MCP server. Community wrappers exist on npm and GitHub, but none comes from PDL. Guard: if an agent needs PDL data, call the API from your own tool definition with a scoped key and a credit cap rather than handing a third-party server your production key.