Cold Email Reply-Rate Optimizer
An A/B scoring sheet that ranks your variants by positive reply rate — with built-in AI prompt formulas to write the next test.
TL;DR
Test cold-email variants, score them on the metric that actually matters — positive reply rate, not raw replies — and use nine built-in prompt formulas to generate the next round. Aimed at the 15–25% top-quartile target, and honest that copy alone won't get you there.
Google Sheets & Excel + any AI tool · 30-day guarantee
The problem
Reply rates are falling — and you're measuring the wrong number.
The 2026 average cold-email reply rate is roughly 3–6% and dropping, as inboxes saturate and AI-generated outreach floods every prospect. Sending more, faster, isn't working.
So teams chase reply rate — and optimize toward a vanity number. A subject line can double your replies and fill your inbox with “unsubscribe.” Without separating positive replies from the noise, you can't tell which variant actually works.
A 14% reply rate that's mostly “no” is worse than an 11% rate that's mostly “tell me more.” If your scoreboard can't see that, it's optimizing you toward the wrong winner.
What's inside
A scoreboard and a prompt library.
Two tools in one sheet, pre-loaded with three example variants so it works the moment you open it.
Start Here
The honest method, the benchmark in plain terms, and how to read the scoreboard.
A/B Scoreboard
Enter each variant's results; it ranks them by positive reply rate, bands them against benchmarks, and flags bad deliverability — with a reply-vs-positive chart.
Prompt Formulas
Nine copy-paste AI prompts: ICP, hooks, subject lines, body, CTA, a Day 3/10/17 follow-up sequence, a reply diagnostic, and a next-test writer.
Benchmarks & Method
What 'good' means in 2026, the exact formulas, and the sources behind every number.
How it works
Test, score, improve — on a loop.
Log the results
Enter each variant's delivered, replies, positive replies, meetings, and bounces.
Read the winner
It ranks by positive reply rate, bands each against benchmarks, and flags weak deliverability.
Write the next test
Use a prompt formula to turn the result into a clean, single-variable experiment.
It checks the foundation, not just the copy
If a variant's bounce rate is too high, the sheet flags it ⚠ Fix list — because a reply rate under ~2% is almost always targeting or deliverability, not wording. The optimizer won't let you A/B-test copy on a broken list.
Try it
Rank your variants on positive reply rate
Winner (positive rate)
B — Question CTA
deliverability-passed
Winner's positive reply rate
4.4%
On target
Most raw replies
C — Discount hook
14.0% reply rate
Deliverability flags
1
bounce ≥ 3%
C — Discount hook pulled the most replies, but B — Question CTA wins on positive reply rate — that's the variant actually working.
This is the live engine. Your variants here reset when you reload. The kit saves every test, charts reply vs positive over time, flags weak deliverability, and gives you nine AI prompt formulas — including the next-test writer that turns each result into a clean single-variable experiment.
Get the kit — $39| Variant | Delivered | Replies | Positive | Bounces | Positive rate | Verdict |
|---|---|---|---|---|---|---|
| 2.4% | Minimum viable | |||||
| 4.4% | On target· winner | |||||
| 1.4% | ⚠ Fix deliverability |
The gap between the two bars is the trap — variants that score high on raw replies often score low on positives.
What you'll see
The signal, not the noise.
The real winner
The variant that wins on positive reply rate — not the one that just pulled the most replies.
Reply vs positive, side by side
A chart that exposes the variants that get lots of replies but few good ones.
The next test, written for you
Prompt formulas that turn each result into a clean, single-variable next experiment.
A deliverability check
A flag when a variant's bounce rate is too high to trust the numbers at all.
Who it's for
Clear about the lane. No inflated promises.
Built for you if…
- Founders and SDRs running outbound who want to test, not guess
- Agencies managing cold campaigns for multiple clients
- RevOps and growth teams that report on reply quality, not vanity metrics
- Anyone using AI to write cold email who needs a way to score what works
Not for you if…
- Anyone wanting a tool that promises 25% reply rates from copy alone
- Teams with a broken list or domain — fix deliverability first
- People after a full sending platform; this is the scoring and prompt layer
Pairs well with
Lift the reply rate, then convert what replies.
More replies only help if the pipeline behind them holds. The Lead Response-Time Revenue Calculator shows what slow follow-up costs, the MQL-SQL Definition & Handoff Kit keeps sales and marketing agreed on what’s qualified, and the Lead Source ROI Scorecard tells you which sources are worth the outreach.
Common Questions
The questions SDRs and founders actually ask before A/B-testing cold email.
It's a spreadsheet with two parts: an A/B scoreboard that ranks your cold-email variants by positive reply rate, and a library of nine built-in AI prompt formulas for generating the next round of hooks, subject lines, bodies, and follow-ups. You test, score, and improve in one place.
It's a top-quartile target, not a baseline. Recent benchmarks put the 2026 average around 3–6% and falling, with 15–25% reached by the best campaigns through tight targeting, strong hooks, and multi-touch follow-ups. The sheet labels that band honestly as a stretch target you climb toward — and it's clear that copy alone won't get you there; targeting, deliverability, and offer move the number more.
Because a 14% reply rate that's mostly 'no thanks' is worse than an 11% rate that's mostly 'tell me more.' Raw replies are a vanity metric — in good campaigns only about 60–70% of replies are positive. The scoreboard shows raw reply rate and positive reply rate side by side and picks the winner on positive replies, so you don't optimize toward the wrong number.
Nine model-agnostic prompts you copy into any AI tool: an ICP sharpener, a timeline-hook generator, a subject-line A/B set, a personalization-line writer, a concise body writer, low-friction CTAs, a Day 3/10/17 follow-up sequence, a reply-quality diagnostic, and a next-test hypothesis writer. Each has a verify reminder so you check the output before sending.
Google Sheets or Microsoft Excel for the sheet, and any AI tool (Claude, ChatGPT, Gemini) for the prompts. No coding. It comes pre-loaded with three example variants so it works the moment you open it.
Honestly, no — and the sheet says so. If your reply rate is under about 2%, the cause is almost always targeting or deliverability, not wording, and the optimizer flags a variant whose bounce rate is too high to trust. Fix the foundation first; then the A/B scoring and prompts help you climb.
Yes — a 30-day money-back guarantee. If it doesn't help you run cleaner tests and lift your positive reply rate, request a refund within 30 days of purchase.
Get the optimizer
Optimize for “yes,” not noise.
The full sheet — scoreboard, chart, nine prompt formulas, and the benchmark method — one payment, yours for good, lifetime updates.
- A/B scoreboard + reply-vs-positive chart
- 9 built-in AI prompt formulas
- 30-day money-back guarantee · lifetime updates