Signals in.
Clay tables full.
Your Clay workflows are only as good as the list they start from. Start them from companies showing a buying signal right now - accounts and people stream straight into your table with the reason to reach out already attached. No CSV exports, no size limits, no copy-paste.
How it works
Rows appear in Clay as they arrive.
Pick accounts or people in Hunch - a hand-picked selection or everything matching a filter - and send them to Clay. Each record streams into your table as its own row, every field in its own column, ready to work the moment it lands.
The difference is what's in the row. These aren't cold records from a static database - they're companies that just showed a buying signal, scored and researched, with the AI why-now brief riding along. Your enrichment and scoring columns start from warm.
Create a Clay webhook table
In Clay: new table → import data → Monitor Webhook. Clay gives you a unique URL.
Add it as a destination in Hunch
Settings → Webhook destinations → paste the Clay URL and name it after the table.
Send, and watch rows land
Select accounts or people → Send → your Clay destination. Rows stream into the table as they deliver.
Built for volume
Ten rows or ten thousand.
Small selections deliver instantly. Large ones run as a background push that checkpoints its progress and keeps going until every row has landed - you can close the tab.
Send all matching
Push an entire filtered view - every account in a segment, every person with a verified email - in one action.
Nothing gets dropped
If Clay throttles, Hunch paces itself and retries until every row lands. Your table ends up complete, not mostly complete.
Rows you can trust
Every row is cryptographically signed by Hunch, so if you route records beyond Clay your systems can verify exactly where they came from.
What's in a row
Every row lands with the reason to reach out.
Accounts arrive with the full firmographic picture and their live signal context. People arrive with title, seniority, a verified work email, and a LinkedIn URL. Every field maps cleanly to its own Clay column - nothing to unpack, nothing to parse.
The why_now field is the same AI-written brief your reps see in Hunch - which means your Clay table starts with the reason to reach out already written, not just a company name to enrich.
From there it's standard Clay: waterfall enrichments, scoring columns, AI prompts against the brief, and a push to wherever your sequences live.
{
"record_type": "account",
"name": "Talkdesk, Inc.",
"domain": "talkdesk.com",
"industry": "Software Development",
"employees": 1800,
"location": "San Francisco, CA",
"score": 91,
"is_hot": true,
"tags": "Enterprise",
"linkedin_url": "https://linkedin.com/company/talkdesk",
"short_description": "AI-powered contact center platform.",
"why_now": "VP of Sales promoted to CRO. 14 AE roles opened in 3 weeks...",
"latest_signal": "Sales Hiring Surge",
"latest_signal_at": "2026-07-07T09:01:43.217Z",
"pushed_at": "2026-07-07T09:14:02.005Z"
}Setup
Connected in two minutes.
If you can copy a URL, you can wire Hunch to Clay. The full walkthrough lives in the webhook docs.
01
Get the Clay URL
In Clay, create a table with a Monitor Webhook source and copy the unique URL it generates.
02
Add the destination
In Hunch: Settings → Webhook destinations → paste the URL and give it the table's name.
03
Send from anywhere
Accounts table, People table, the API, or MCP - pick records and choose your Clay destination from the Send menu.
Feed Clay with signals.
Start your Clay tables from companies that are actually in market, with the why-now already written.
See your signals