AI prompt

Build a Clay pipeline to Attio that pays for phones only on reply

Tools: Attio ▪ Clay ▪ Lusha

Phones are the expensive part of a prospect list, and many of them never get dialed. This play builds the list in Clay with Lusha, runs the outreach on email first, and reveals a direct dial only for the people who reply. Every record lands in Attio, enriched and current, without anyone copying a row.

It runs on the native Lusha integration in Clay and Clay’s native Attio integration. You’ll need a Lusha API key, a Clay account and an Attio workspace admin to connect it.

How the pipeline works

Find in Clay with Lusha → verified emails → signals for the reason to write → personalize → send → positive reply → reveal the phone → upsert to Attio

Build it step by step

1. Find the people. In Clay, add a table from Lusha and choose Find People Lookalikes from your best customers, or bring in any list you already have: a CRM export, a Clay search or a CSV.

2. Add verified emails only. Add the Enrich Person enrichment from Lusha and map the LinkedIn URL or email. Use it for emails at this stage. Phones come later, for the people who earn them.

3. Add the reason to write. Add Lusha signal enrichments such as Enrich Company Jobs Growth, Enrich Company Headcount Growth or Enrich Company News. Use them to rank the table and to give every row a reason to reach out now.

4. Personalize. Add an AI column in Clay that writes the first line from the signal: the hiring push, the new funding, the new leader. Skip rows with no signal, or send them a lighter version.

5. Send. Push the qualified rows to your sequencer from Clay.

6. Reveal phones on reply. If your sequencer can send replies to a webhook, point it at a Clay webhook table. For each positive reply, run Lusha Enrich Person again to get the direct dial, so the rep can call while the conversation is warm.

7. Upsert to Attio. Add the Attio action in Clay and choose upsert, matching on email for people and domain for companies. Turn on auto-update so new rows sync without anyone pushing them.

What you’ll get back

An Attio person record for every prospect who replied, with fields like these:

Name:        J.K.
Title:       Head of Revenue Operations
Company:     [company].com
Email:       j.k@[company].com (verified)
Phone:       +1 (•••) •••-••12 (direct dial, revealed on reply)
Signal:      Engineering hiring up, last 30 days
First line:  "Saw you're adding six engineers this quarter..."
Status:      Replied, positive

Example outputs in this play are illustrative. They reflect the structure, fields and format of real Lusha output, but were not pulled from a live session. Run the workflow with your own data to see live results.

Why it works

Typical outbound pays for every field on every row, then contacts a fraction of them. This flips it. Emails are cheap and you need them for everyone, so you get them up front. Phones cost more, so you reveal them only for prospects who have already said yes to a conversation. Lusha’s signals give each email a reason to exist, and Attio gets complete records instead of half-filled ones. The data is verified, GDPR and CCPA compliant, and never scraped from social networks. Credit rates are in how Lusha credits work.

FAQ

Do I need my own Lusha account to use Lusha in Clay?

Yes. Connect your Lusha API key inside Clay to use the Lusha enrichments.

Which credits does each step use?

Lusha’s Enrich Person and Enrich Company use your Lusha credits. Clay actions use Clay credits. Check the Lusha in Clay guide for the current split on signals and lookalikes.

Can Clay update existing Attio records instead of creating duplicates?

Yes. Use the upsert action and match on email for people or domain for companies.

What if my sequencer can’t send replies to Clay?

Run step 6 on a schedule instead: export positive replies to a Clay table once a day and reveal phones for those rows.

Can I use another CRM instead of Attio?

Yes. Clay pushes to Salesforce, HubSpot and others. Swap the Attio action for your CRM’s.

Related plays

Ready to run this?

One data connection. Works in Claude, ChatGPT, your CRM, or any agent you build.