Hey, it's Christian from the RevGrowth newsletter.

I'm emailing you a 3-minute playbook on how to use AI to find exactly
why your cold email campaigns are underperforming — and fix them.

We used this exact process to analyze 4,055 scripts across our client
accounts. Today I'm giving you the prompts.

MOST PEOPLE HAVE NO IDEA WHY THEIR CAMPAIGNS AREN'T WORKING

They'll change the subject line. Adjust the send time. Swap out the
lead list. Try a new sequencer.

And the reply rates barely move.

The problem is they're guessing. They don't have a system to actually
diagnose what's wrong — so every iteration is a shot in the dark.

Last months ago I had Claude Code run a full analysis across every
email script in our EmailBison instance. 4,055 scripts across 1,160 campaigns.

What came back was the clearest picture of what actually separates
winning campaigns from losing ones we've ever had. A 42.5x gap between
our top and bottom performers. Clear patterns in CTA language, tone,
personalization depth, and value framing.

The best part? You don't need a developer to do this for your own
campaigns. You just need your data and the right prompts.

Here's exactly how to do it.

STEP 1 — EXPORT YOUR CAMPAIGN DATA

Pull your top 10 and bottom 10 campaigns from whatever sequencer you
use. Simply feed your sequencer API to Claude Code and ask for these campaigns with the following data to start:

  • The email script (all steps/follow-ups)

  • The reply rate

Separate them into two groups: top performers and bottom performers.
The threshold doesn't matter much — just make sure each group is
clearly above or below your average.

STEP 2 — RUN THE PATTERN ANALYSIS PROMPT

Have Claude Code analyze both groups and identify:

  1. What patterns, phrases, and structures appear consistently in Group
    A that are absent in Group B?

  2. How do the CTAs differ between the two groups? Categorize each CTA
    as: soft permission ask, direct call request, right-person redirect,
    or question opener.

  3. What does Group A do in the opening line that Group B doesn't?

  4. How does the value proposition differ? Does Group A quantify the
    outcome more specifically?

  5. What should I stop doing immediately based on what you see in Group
    B?

Be specific. Quote lines from the scripts where relevant.

STEP 3 — RUN THE CTA AUDIT PROMPT

Your CTA is the single highest-leverage element in the email. Run this
separately:

"Review these email scripts and for each one, identify:

  1. The exact CTA used (quote it directly)

  2. The CTA type: soft permission ask / call request / right-person
    redirect / question / value asset offer

  3. A 1-10 score for how low-friction the ask is (10 = easiest possible
    yes, 1 = high commitment ask)

Then rank them from highest to lowest friction and tell me which CTA
types correlate with my highest-performing campaigns."

When we ran this on our data, one pattern was impossible to ignore.

"Am I reaching out to the right person" showed up in 36.9% of our
bottom performers. It showed up in only 5% of our top performers.

That single finding changed how we write CTAs across every client
account.

STEP 4 — AUDIT INDIVIDUAL SCRIPTS

Once you know your patterns, use this prompt to fix your worst
performers one by one:

"Audit this cold email script against these four criteria:

  1. Does the opening line create immediate relevance — or does it feel
    like it could be sent to anyone?

  2. Does the email offer a tangible asset or a quantified outcome — or
    does it describe features and process?

  3. Is the CTA a soft permission ask or does it require a higher
    commitment (call booking, demo request, etc.)?

  4. Are there any phrases that signal mass blast — things like 'Am I
    reaching out to the right person', 'just wanted to reach out', 'would
    it be crazy if', 'give us a shot'?

Score it 1-10 overall. Then rewrite it."

Feed it your worst campaigns first. The delta between the original and
the rewrite will show you exactly where the script was losing people.

STEP 5 — BUILD YOUR OWN RULES LIST

After running steps 2-4, you'll have a clear picture of what works
and what doesn't — specifically for your offer, your ICP, and your
market.

Write it down as a set of rules. Ours looks like:

  • Never use right-person CTAs

  • Always quantify the value prop with a specific number

  • Open with a situation question, not a compliment

  • Offer an asset before asking for a call

  • Keep tone conversational, not formal

Every new script we write gets checked against that list before it
goes live.

That's the audit system. Four prompts, one clear output, a rules list
you actually use.

The teams doing this consistently are the ones compounding. Every
campaign teaches them something. Every iteration is informed.
Everyone else is still guessing.

If you want to skip the audit entirely and just start with scripts
that already passed this test — the Cold Email Script Vault has the
top 10% of scripts from our 2025 campaigns.

210 scripts across 12 industries. Every one already analyzed and
validated.

Grab it here: https://emailscriptvault.com/

Hope you found this email valuable.

Have a blessed day,

Christian

Keep Reading