Clay Table Template for Account-Based Outbound: 2026 Build Guide
By Puzzle Inbox Team · May 22, 2026 · 9 min read read
Build a Clay table template for account-based outbound: ICP firmographics, signal enrichment, contact waterfalls, and Smartlead handoff in one repeatable workflow.
A working Clay table template for account-based outbound starts with one row per account and ends with a Smartlead-ready CSV
The fastest way to run account-based outbound in 2026 is a Clay table template that treats each row as a target account, enriches firmographics and signals, fans out to 2-4 personas per account, and writes a campaign-ready CSV into your sending tool. This guide walks through the exact column structure, the enrichment order, and the handoff to Smartlead or Instantly.
Step 1: account row foundation
Start your Clay table template with one row per company. Required base columns: company_domain, company_name, tier (1/2/3), and signal_source. Import your tier-1 list from a CRM export, a sales-nav search, or a partner-supplied account list. Avoid mixing contact-level data here; that comes later in a child table or expanded rows.
Add firmographic enrichment via Clarbit, Apollo, or Crunchbase: headcount, funding stage, tech stack, and HQ region. Use a conditional run so you only burn credits on rows where tier equals 1 or 2.
Signal columns that drive sequencing
Account-based outbound only works when sequences reference a real signal. Add columns for: recent funding (via Crunchbase), job postings (via LinkedIn or Wellfound), tech adoption (via BuiltWith), and news mentions (via Google News API). Each signal feeds a conditional snippet later.
Step 2: contact waterfall per account
Use a Clay "find people" enrichment to pull 2-4 contacts per account, filtered by title regex (for example, VP|Director|Head of Marketing). Output a child table or use Clay's expand-rows feature so each contact becomes its own row referencing the parent account.
Layer an email waterfall: Apollo, then Anymailfinder, then Datagma, then Clay's native finder. This typically yields 70-85% verified emails on ICP-fit accounts. Verify with NeverBounce or ZeroBounce as the final step, and drop anything not marked valid or accept-all-safe.
AI personalization at the row level
Add a Claude or GPT-5 column that ingests the company description, the strongest signal, and the contact's recent LinkedIn post. Prompt it to produce a 1-sentence opener that references the signal and the persona. Keep it under 25 words to preserve deliverability and avoid bot-like patterns. For reply triage on the receiving side, Puzzle Inbox classifies AI-generated personalization replies cleanly.
Step 3: campaign-ready export
Add export columns that match your sending tool's schema: email, first_name, company, opener, signal_snippet, cta_variant. Use a Clay HTTP API push to Smartlead or Instantly directly, or schedule a Google Sheets sync that your cold email tool picks up.
Tag each row with the campaign name and the persona segment so you can split-test openers across personas without rebuilding the table.
Reusable template structure
Save the table as a Clay template so you can clone it per ICP. Keep the column order: identifier, firmographic, signal, contact, email, AI personalization, export. A consistent order makes debugging waterfalls and credit usage 10x faster across new accounts.
Step 4: monitoring and iteration
Add a feedback column that pulls reply outcomes from your sending tool weekly. This lets you score signals: funding signals might outperform job-posting signals 2:1 in one ICP and reverse in another. Iterate the template every 30 days based on which signal columns correlate with positive replies.
Pair the table with a unified inbox so reply data flows back without manual tagging.