Clay
Clay is a go-to-market data platform that pulls contact and company records from 200+ providers under one contract, runs AI research agents over them, then pushes clean, scored data into your CRM, your ads and your outreach.
What is Clay?
Clay is a go-to-market infrastructure platform built around a spreadsheet-like table: every row is a company or a person, and every column can trigger an enrichment, an AI agent or an action. It assembles four layers. The data layer is a marketplace where more than 200 data and AI vendors sit under one Clay contract, at wholesale rates the platform displays provider by provider, together with Audiences, intent signals and waterfall enrichment — the ability to query several providers in sequence for the same field and pay only once a result comes back. The agent layer is Claygent, an AI researcher with web access that reads sites, PDFs, job posts and financial documents and writes structured output; Clay's own counter shows over five billion Claygent runs. Account Research Agents extend that across a whole segment, keeping persistent memory per account and writing back to the CRM after human approval, with a full reasoning trace and per-agent cost and error tracking. The orchestration layer covers reusable Functions, AI formatting, webhooks, an HTTP API and two-way sync with HubSpot, Salesforce, Pipedrive, Close, Loxo, Snowflake, Databricks and BigQuery. The execution layer ships a native sequencer plus ad audience sync to LinkedIn, Meta, Google, Reddit and Bing. Two Chrome extensions, an MCP server for sales reps' AI tools and a CLI/API plugin for coding agents round it out. Pricing runs on two meters — actions for platform usage, data credits for what you buy from vendors — while list building and data cleaning cost nothing. The key point is what Clay is not: it holds no data of its own and replaces neither your CRM nor your sending tool. It is the layer that sources, enriches and routes between them, which is why customers such as OpenAI, Anthropic, Figma, Canva, Intercom, Rippling and Mistral AI describe it as their orchestration layer rather than another point solution.
What it does
- Buy and cross-reference data from 200+ providers under a single Clay contract
- Chain several providers in a waterfall and pay only when a record is actually found
- Send an AI agent to research anything on the web and return it as a structured column
- Track buying signals: job changes, promotions, hiring, company news and web intent
- Keep a CRM and a data warehouse clean and in two-way sync, reverse ETL included
- Score, qualify and route inbound leads automatically
- Build ad audiences and push them to LinkedIn, Meta, Google, Reddit and Bing
When to use Clay / When not to
A quick filter to help you decide if Clay is the right fit.
When to use Clay
- GTM and revenue operations teams that own lead enrichment, scoring and routing around a CRM
- Growth leads at startups running the entire outbound motion single-handedly
- Outbound and GTM engineering agencies managing campaigns for several clients at once
- Demand generation and lifecycle marketers building ad audiences and inbound scoring from live signals
- Data and analytics teams accountable for CRM hygiene, TAM sizing and warehouse synchronisation
When not to use Clay
- Anyone shopping for a CRM or an email sending tool: Clay plugs into those rather than replacing them
- Teams that need a few hundred leads once in a while, for whom a $167/month entry tier and a credit budget are hard to justify
- Organisations handling health records or cardholder data, which the terms forbid without a separate agreement
- Users who want results without a learning curve: Clay assumes somebody will build and maintain the workflows
- Mobile-first or non-English-speaking teams, since there is no mobile app and no localised interface
How to use Clay
A typical end-to-end flow, from setup to results.
- Create a workspace on app.clay.com — the 14-day trial needs no credit card
- Build a list: import a CSV, connect your CRM, or source from LinkedIn, Google Maps, GitHub and websites, all free of credits
- Run basic enrichment and cleaning from the marketplace to fill in contact details and firmographics
- Stack several providers into a waterfall on the fields that matter, so you only pay when a value is found
- Describe an AI research task in plain language; Sculptor helps turn it into a Claygent prompt
- Connect the agent to its tools — web research, document reading, business context — then test on real rows and compare versions
- Deploy the agent to a table or an Audience and manage every deployment centrally from Claygent Builder
- Wire up signals and triggers: job changes, hiring, company news, web intent or your own webhooks
- Push results out through two-way CRM sync, your warehouse, the HTTP API, the native sequencer or a third-party sending tool
- Train the team through Clay University, the Clay 101 course, live cohorts and the certifications, starting with 20 to 30 rows at a time
Pros & Cons
Pros
- One contract instead of a dozen: the FAQ claims two to three times ZoomInfo's coverage at a fifth of the cost or less
- Transparent billing — Clay names the provider behind every data point and passes on wholesale rates
- Waterfalls bill only on a result, and credits are refunded when a provider refunds Clay for invalid data
- List building, prospecting and data cleaning consume no credits, and seats are unlimited on every plan
- Genuinely extensible: bring your own API keys, or reach any tool with an API through webhooks and the HTTP API
- SOC 2 Type II, GDPR, CCPA and ISO 27001 badges, plus a contractual 99.9% uptime target with service credits
- An unusually dense learning ecosystem: Clay University, live cohorts, certifications, a Slack community and an expert directory
Cons
- Steep entry point: the first paid tier is $167 per month on annual billing, and Growth jumps to $446
- Two meters to forecast — actions and data credits — with per-provider rates that make budgeting genuinely hard
- The free plan and the trial cap tables at 50 to 200 rows, which limits any realistic evaluation
- Credit rollover is tight: twice the monthly allowance on monthly plans, only 15% of unused credits on annual ones
- All fees are non-refundable and subscriptions auto-renew at the then-current rate
- Support is self-serve below Enterprise: priority queue starts at Growth, dedicated Slack and a growth strategist are Enterprise-only, as are SSO and RBAC
- Ageing and inconsistent documentation: a 2020 privacy policy, 2024 terms, mailto links still pointing at the old clay.run domain, and provider counts that swing between 100+ and 200+ across pages
Pricing & Plans
Clay maintains a permanent free plan — 500 actions and 100 data credits per month, unlimited seats and tables, capped at 200 rows per table — alongside a 14-day trial of the Pro feature set that includes 1,000 credits and requires no credit card. The cheapest paid tier, Launch, is advertised at $167 per month on annual billing, bundling 180,000 actions a year with 30,000 data credits; the same plan costs $185 per month when billed monthly. Growth starts at $446 per month annually, and Enterprise is quoted on request. All prices are stated in US dollars.
- Free — no cost — 500 actions and 100 data credits per month
- unlimited seats and tables
- multi-provider waterfalls
- Claygent
- bring-your-own API key and the sequencer
- capped at 200 rows per table
- with no phone number enrichment
- Launch — from $167/month billed annually
- $185/month billed monthly — for small teams automating prospecting workflows
- starts at 180
- 000 actions and 30
- 000 data credits a year and adds phone enrichment
- unlimited Audiences search
- job-change and hiring signals
- email campaigns through integrations and reusable Functions
- Growth — from $446/month billed annually
- $495/month billed monthly
- flagged as the recommended tier — for teams with CRM-based workflows
- starts at 480
- 000 actions and 72
- 000 data credits a year and adds automatic CRM and warehouse sync
- 250k Audience imports
- any HTTP API integration
- Enterprise — custom pricing — for organisations running GTM systems at scale
- unlimited Audiences search and imports
- SSO
- advanced role-based access control with per-workbook credit budgets
- a dedicated growth strategist
- onboarding and co-building
- a dedicated Slack channel and bi-annual business reviews
- Expansion tiers on annual billing let you scale Launch from 180
- 000 actions a year at $54/month up to 2.4 million at $486/month
- and Growth from 480
- 000 at $185/month up to 2.4 million at $765/month
Data, GDPR & hosting
A consolidated view of how Clay handles your data.
GDPR overview
The privacy policy devotes a section to GDPR: Clay offers Data Processing Agreements on request through privacy@clay.com, and those DPAs embed standard contractual clauses as the transfer mechanism. EEA residents are told they can complain to their local supervisory authority, access, correction and deletion requests get an answer within thirty days, and the product itself lets you export and permanently delete data. The enterprise page carries a GDPR badge next to SOC 2 Type II, CCPA and ISO 27001, with a Trust Center for the SOC 2 report, and the policy states that no personal information has been sold in the preceding twelve months. Two gaps deserve mention: no Article 27 EU representative and no data protection officer is named anywhere, and the privacy policy still carries a September 2020 date.
Who owns the data?
You keep ownership of everything you upload. Submitting content grants Clay Labs a worldwide, non-exclusive, royalty-free, transferable and perpetual licence to use, reproduce, modify, adapt, publish, translate, distribute and display it, for the stated purpose of providing and improving the service, with sub-licensing allowed so cloud providers can store it. The terms then narrow that considerably: Clay claims no ownership of uploaded data, makes it visible only to your workspace and the Clay team, and commits to using it for nothing beyond delivering the service. Aggregated performance data about how you use the account is exploited to improve the product. Feedback you send is licensed irrevocably, with no attribution and no compensation.
Reuse rights
Nothing in the terms requires Clay's permission to reuse what you produce. Data leaves through two-way CRM sync, data warehouse sync, webhooks, the HTTP API, the native sequencer or any third-party sending tool, and the privacy policy confirms you can export your data and delete it permanently. The real caveat sits upstream: Clay holds no data of its own, so every enriched record comes from a marketplace provider whose own conditions travel with it. Clay says it is highly selective about the providers it integrates and works only with GDPR-compliant companies, while pointing you to each provider's privacy documents. Compliance of the outreach itself, including consent, opt-out and do-not-call handling, remains your responsibility.
Data retention & training
Hosting summary
Clay publishes remarkably little about where data physically sits. No hosting country and no region is stated anywhere on the pages reviewed. The privacy policy describes technical and organisational safeguards without naming a location; the terms add an information security programme, prompt breach notification and administrative, physical and technical controls, and allow sub-licensing to cloud providers so content can be stored — again without naming them. Payment data is explicitly stored by Stripe, and international transfers rely on standard contractual clauses inside the DPA. No subprocessor list is published, and the Trust Center that might carry one renders entirely in JavaScript. A DNS lookup on 9 August 2026 resolved the domain to an Amazon CloudFront node in Zurich, which indicates a content delivery edge rather than a storage jurisdiction and should not be read as Swiss hosting. What is certain: the company is registered in Brooklyn, New York, and the terms are governed by New York law.
Things to keep in mind
Risks and trade-offs to weigh before adopting Clay.
- Mass prospecting cuts both ways: Clay argues for fewer, better-targeted emails, but the same engine makes unsolicited outreach at scale trivially easy
- Compliance of the outreach — consent, opt-out, do-not-call, GDPR towards the people you contact — sits with you, not with Clay
- Enriched personal data comes from third-party providers whose legal basis for collection you cannot verify from inside Clay
- Costs are hard to predict: two meters, per-provider rates, capped rollover and non-refundable fees make overspending easy
- Lock-in is a genuine question once Clay becomes the layer between your CRM, your warehouse and your sending tools, however well export is documented
- Delegating judgement to agents can silently discard viable prospects; the reasoning trace helps but is not an audit
- Automating prospect research can quietly erode a sales team's own understanding of its market
Setup & Integrations
Technical difficulty
Getting started needs no code: create an account, import a CSV or connect a CRM, add enrichment columns. The spreadsheet-style interface keeps that first hour approachable for an ops profile. Difficulty arrives with waterfalls, reusable Functions, webhooks, the HTTP API and two-way CRM and warehouse sync, and the MCP server or CLI plugin assume a technical user. Clay is explicit that it is primarily self-serve below Enterprise, which is precisely why it built Clay University, live cohorts, certifications and an expert directory — and why a whole job title, GTM engineer, has grown up around operating it well.
Deployment
Integrations
Behind Clay
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Clay.
Frequently asked questions
What exactly does Clay do?
Does Clay own its own database?
How much does Clay cost?
How do Clay credits work?
Do unused credits roll over?
What does Clay integrate with?
What does Clay do with my data?
Is Clay GDPR compliant, and is a DPA available?
Is there a referral or affiliate programme?
Is there a mobile app, and what support do I get?
Should you pick Clay?
Clay has done something few tools manage: it turned a procurement problem into a product. Instead of signing with ZoomInfo, then a phone provider, then an intent vendor, you sign once and buy from more than 200 of them at wholesale rates, paying only when a waterfall actually returns a value. Layer AI agents that research the open web on top, add two-way sync into the CRM and the warehouse, and you get the orchestration layer that customers like Figma describe sitting between Salesforce and Snowflake. The customer list — OpenAI, Anthropic, Canva, Intercom, Rippling, Verkada, Mistral AI — and a $5 billion valuation reached in January 2026 suggest the positioning holds.
The price of that power is real. Entry is $167 a month on annual billing, budgeting means tracking two meters whose rates vary by provider, fees are non-refundable and rollover is capped. Below Enterprise you are largely on your own: SSO and role-based access control are gated, and the sheer existence of paid cohorts, certifications and a partner network says plainly that Clay rewards someone willing to learn it properly. The documentation shows its age too — a privacy policy dated 2020, no published subprocessor list, no stated hosting country, and provider counts that drift between pages.
The verdict is a matter of volume and ownership. If a team has enough pipeline to justify the spend and someone whose job is to build the workflows, Clay pays back quickly and the case studies are unusually concrete. If you need a few hundred leads occasionally, or expect it to work untended, a simpler and cheaper tool will serve you better. Treat the free plan and the 14-day trial as a genuine test, and check the credit arithmetic against your own volumes before committing to a year.
- Choosing a selection results in a full page refresh.
- Opens in a new window.