Automatically test and improve cold email copy with Claude

Runs your cold email campaigns as ongoing experiments, using AI to test new copy and promote the best performer.

How the work actually flows

It repeats. Repeats new challenger restarts test cycle; a person is alerted when a step fails.

Pattern: Structured Loop (21)

flowchart TD trig(["runs every few hours"]):::trigtime s0["check campaign performance"]:::task s1["compare best vs challenger"]:::task s2["pick winning version"]:::task s3["write new challenger copy"]:::task trig --> s0 s0 --> s1 s1 --> s2 s2 --> s3 lp{"until reply rate drops too low"}:::gate s3 --> lp lp -. "new challenger restarts test cycle" .-> s0 lp -->|"finished"| out out[/"continuously optimized email copy"/]:::out pay{{"reply rates improve without manual testing"}}:::pay out --> pay esc(("Alerts a person")):::human s2 -. "if it fails" .-> esc esc -.-> out classDef task fill:#e7f6fe,stroke:#34b8f0,color:#2c2a29 classDef svc fill:#f6f8fa,stroke:#7c8795,color:#2c2a29 classDef mi fill:#e7f6fe,stroke:#0079a8,color:#2c2a29,stroke-width:2px classDef human fill:#fff,stroke:#0079a8,color:#0079a8 classDef store fill:#f6f8fa,stroke:#0079a8,color:#2c2a29 classDef trig fill:#00a4eb,stroke:#0079a8,color:#fff,font-weight:bold classDef trigtime fill:#00a4eb,stroke:#0079a8,color:#fff,font-weight:bold classDef trigdata fill:#8ad4f5,stroke:#0079a8,color:#06314c,font-weight:bold classDef gate fill:#fff,stroke:#e8a23d,color:#6b4708,font-weight:bold classDef out fill:#1f9d6b,stroke:#167a53,color:#fff,font-weight:bold classDef pay fill:#06314c,stroke:#021f33,color:#fff
Starts itA stepA personRepeat or finishResultPayoff
Build size
Standard

A mid-size build with several tools working together.

Business functions
Email AutomationSpreadsheet & Database Ops
Connects
ClaudeInstantlyGoogle Sheets
Featured in

The problem it solves

You write cold email copy once, send it, and hope it works, but you rarely have time to systematically test new versions against the old ones. Winning subject lines and openers get lost because nobody is tracking what actually improves reply rates over time.

Who it fits

Sales and marketing teams running ongoing cold email campaigns who want their copy to keep improving.

How it works

  1. Every few hours, the automation checks your live campaign performance in Instantly
  2. It compares your current best email against a challenger version
  3. Once enough replies have come in, Claude reviews the results and picks a winner
  4. Claude writes a new challenger with a specific reason for the change
  5. If reply rates ever drop too low, it pauses and alerts you instead of continuing
What you get

Reply rates your campaigns climb to

Your cold email campaigns keep testing new copy against your best performer and automatically promote whichever version gets more replies, pausing if results dip.

What you get

A constantly improving cold email campaign, with a log of every test, winner, and reasoning in Google Sheets.

What you need

An Instantly account, an Anthropic Claude API key, and a Google Sheets account.

We can build this. But should you?

The hard question is not how to build it. It is whether this is the right thing to build first.

That is what a Fractional Chief AI Officer figures out with you, before anyone writes a line of code.

Let's Talk Strategy

Related automations

Back to the AI Playbook