Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for growth marketers

Subscriber Growth Experiment Agent

Find signup changes that reliably raise subscriber growth

Subscriber Growth Experiment Agent: what goes in, what the agent does and what you get

What it does

Newsletter growth stalls when writers change many things at once and cannot tell what worked. This agent runs one growth experiment at a time. It pulls signups and conversion by source for the last 30 days and picks the source with the most traffic and the lowest conversion. It proposes one change, such as a new form headline or a referral reward, and writes the variant. The writer approves it before it goes live. It tracks signups for control and variant, excluding single-day spikes over three times normal traffic. It needs at least 300 visitors per variant before judging. If the sample is too small or the difference unclear, it extends the test a week, and after two extensions proposes a bolder variant instead of declaring a winner. Results go into a growth log with the numbers. Edge case: a spike from one viral mention is excluded so it does not fake a win.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
ApprovedYes, continueNo 1 STARTS WHEN Previous experiment closed 2 USES A TOOL Pull signups and conversion by source for the last30 days 3 DOES Pick the weakest high-traffic source and propose onechange 4 YOU APPROVE Writer approves the variant before it goes live 5 USES A TOOL Track signups for control and variant 6 CHECKS THE RESULT Is the sample large enough and the difference clear? If not: extend the test by a week, or propose a boldervariant after two extensions. Back to step 5. 7 DOES Write result and numbers into the growth log 8 RESULT Next experiment proposed
Read the steps as a list
  1. Previous experiment closed
  2. Pull signups and conversion by source for the last 30 days
  3. Pick the weakest high-traffic source and propose one change
  4. Writer approves the variant before it goes liveThe agent waits here for your OK.
  5. Track signups for control and variant
  6. Is the sample large enough and the difference clear?If not: extend the test by a week, or propose a bolder variant after two extensions. Back to step 5.
  7. Write result and numbers into the growth log
  8. Next experiment proposed

How it decides

It picks the source with the most traffic and lowest conversion to test next, and needs a minimum sample before calling a result.

  • Test one change at a time
  • Need at least 300 visitors per variant
  • Exclude single-day spikes over three times normal traffic

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Minimum visitors per variant (default 300)
  • Maximum test length (default 6 weeks)
  • Sources allowed for testing
  • Where the growth log lives

What keeps you in control

It always asks you first

  • Putting a variant live
  • Changing the default form after a win

Hard limits

  • Never changes live forms without approval
  • Never buys subscribers or lists

It stops when

  • Done: result logged
  • Stop: traffic too low to reach sample in six weeks

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensThe blog footer form had 9,000 monthly views and 0.8% conversion. The agent tested a headline naming the weekly benefit. After one week the variant showed 1.1% on 280 views, below the 300 minimum, so the check failed and it extended. After two weeks it was 1.3% against 0.8% on 650 views each. The writer approved it as the default.

More agents for growth marketers