Clay
Where lists get found, enriched, and qualified. The data engine behind every campaign, with FullEnrich in the waterfall.
In our stackLast reviewed July 17, 2026
What Clay actually is.
Clay is a data and enrichment platform built around a spreadsheet-style table. Each row holds a company or a contact. Each column runs a lookup, an AI step, or a push into another tool. Instead of relying on one data vendor, Clay routes a column through a waterfall of outside providers until something returns a usable email, phone, or firmographic field. Claygent, its built-in AI agent, works inside that same table for open-ended research when structured lookups come up empty.
It doesn't run itself. Waterfalls, conditional logic, and Claygent prompts need someone who keeps rebuilding the table as providers drop off, pricing changes, or a campaign's targets shift. Leave that upkeep undone and coverage quietly drops, or credit spend quietly climbs instead.
What it does
- Runs each field through multiple data providers until one returns a result.
- Claygent, the built-in AI agent, does open-ended web research inside a table row.
- Table logic supports conditionals, formulas, and branching, like a spreadsheet.
- Syncs enriched records back into a CRM or a data warehouse.
- Pushes qualified audiences straight to LinkedIn, Meta, or Google ad platforms.
- Tracks job changes and similar buying signals as an ongoing feed.
- Exposes an HTTP API and webhooks so outside code can trigger a table.
Where it sits in the pipeline.
- AI & Intelligence
- Data & Enrichment
- Scraping & Custom Data
- Email Infrastructure & Deliverability
- Outreach & Sending
- LinkedIn & Multichannel
- Signals & Intent
- Calls & Meetings
- CRM & Pipeline Ops
Runs well with
The tools we'd run next to Clay, and what each pairing covers.
Its job inside client systems.
Clay is the enrichment layer. We build wide target lists, then run them through waterfalls that stack providers so coverage stays high and cost stays reasonable, with FullEnrich handling the email step. Qualification happens here too. Claude scores and personalizes on pre-filtered subsets of a few hundred rows, never a raw list of thousands. Qualified prospects post back and move to SmartLead. We're on the legacy Explorer plan, which still beats the newer two-currency pricing at our volume.
Pricing, without the fiction.
From the Clay pricing page · verified July 17, 2026
- 100 data credits/mo, 500 actions/mo
- Unlimited seats and tables
- Up to 200 rows per table
- Multi-provider waterfall enrichment
- Enrich with Claygent
- Bring your own API key
- 15,000 actions/mo, 3,000 data credits/mo (base tier, expandable to 200,000 actions/50,000 credits)
- Unlimited seats
- Enrich phone numbers
- Unlimited search with Audiences, job-change and signal tracking
- Up to 50,000 rows per table
- Launch email campaigns via integrations
- 40,000 actions/mo, 6,000 data credits/mo (base tier, expandable to 200,000 actions/50,000 credits)
- Unlimited seats
- Auto-sync and enrich CRM/data warehouse
- HTTP API integrations, webhook-triggered signals
- Web intent signal tracking, 1 ads audience included
- Priority support
- 200,000+ actions/mo, 100,000+ data credits (custom)
- Unlimited seats
- Unlimited ad audiences, unlimited search/imports with Audiences
- SSO and role-based access control (RBAC)
- Dedicated growth strategist, custom CSA and DPA
- Data warehouse integrations, 15-minute source sync frequency
The sticker price ($167/mo Launch, $446/mo Growth) is not the real number - actual cost is set by data credit consumption, and every provider call in a waterfall (checking 3-4 vendors to find one working email) burns multiple credits per row, not one. Teams that build wide waterfalls or lean hard on Claygent's AI research agent will blow through the included 3,000-6,000 monthly data credits well before the action allowance runs out, then pay a 30% premium to top up. Seats and tables are unlimited on every tier, so cost doesn't scale with headcount - it scales entirely with how aggressively tables are built and run, which makes it cheap for light, well-designed workflows and expensive for sloppy or overly broad ones.
Choose it, or skip it.
Choose it if
- Teams with a dedicated ops or growth person willing to spend real hours building and maintaining tables, waterfalls, and AI prompts - Clay is a workbench, not a plug-and-play tool.
- GTM teams that need to pull from many data providers at once and want one place to waterfall between them instead of paying for and stitching together five separate point tools.
- Agencies and in-house teams running account or lead research at moderate volume who can absorb the $167/mo Launch plan (3,000 data credits, 15,000 actions) or justify Growth at $446/mo.
- Teams already comfortable with spreadsheet-style logic (formulas, conditional runs) who want to add AI enrichment and outbound automation on top of that mental model.
Skip it if
- Solo founders or small teams that just want clean contact data fast - the learning curve and setup time will cost more than the tool saves at low volume.
- Teams with no one who owns GTM ops day to day - Clay tables need upkeep (broken waterfalls, credit burn from bad enrichment logic) or they quietly degrade.
- Anyone expecting a fixed monthly bill - actual spend depends on data credit consumption per enrichment, which is hard to forecast until you've run it.
If not Clay, then what.
- Apollo.iofree tier (no card) + paid plans from ~$49-59/mo/user (annual billing)
Choose Apollo when you want an integrated database plus sequencing in one login rather than a build-it-yourself workbench pulling from 150+ separate vendors.
- FullEnrichcredit-based; from $55/mo at 1,000 credits
Choose FullEnrich when you only need waterfall contact enrichment (not full workflow orchestration, AI agents, or table logic) at a simpler, narrower price point.
- BetterContactFree tier (50 credits, no card) + paid plans from $15/mo (200 credits); pay only for valid/verified results
Choose BetterContact when the job is strictly bulk email/phone waterfall enrichment on a CSV and you don't need Clay's table-based automation layer.
Asked in real client work.
What actually eats through the credits - seats, contacts, or something else?
Two separate meters. Data credits get spent when Clay pulls a record from one of its 150+ marketplace vendors (email, phone, firmographic data); Launch includes 3,000/mo and each credit starts around $0.05 with volume pricing. Actions are platform operations - enrichment runs, AI model calls, table writes, integrations - and Launch includes 15,000/mo at under $0.01 each. A waterfall that checks 3 providers for one contact can burn several data credits in one row.
What happens if we run out of credits mid-month?
You can buy top-up data credits at a 30% premium over the base rate on Launch/Growth plans. Unused credits roll over on Launch and Growth, capped at 2x your monthly allotment. Enterprise plans roll over up to 15% of the prior year's purchased credits if they renew at an equal or higher commitment. Either way, credits don't bank indefinitely.
Is there a real free tier to test it before paying?
Yes - Free gives 100 data credits and 500 actions per month, capped at 200 rows per table. That's enough to build and test a workflow, not enough to run a live campaign; Launch (50,000 rows/table) is the first plan sized for production use.
How hard is the AI-agent piece (Claygent) to actually use well?
Clay's own onboarding series devotes multiple lessons to it (prompt engineering with a named framework, conditional AI runs, web research agents), which signals it's not point-and-click - getting reliable output requires writing and iterating on prompts inside the table, similar to prompt engineering work anywhere else.
Do seats or number of users affect the price?
No - seats and tables are unlimited on every paid tier. The bill is driven entirely by data credit and action consumption, so a 20-person team on Launch pays the same $167/mo as a 2-person team, provided both stay under 15,000 actions and 3,000 data credits.