As of 13 August 2026, AI can find the best-performing content in your email campaigns.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, power-user skill, and roughly 30 minutes until you can act on the result.

   How to actually do it

   1. Open your email platform's campaign reports and export the results for the campaigns you want to compare, including the content, audience, send date, delivered messages, clicks, conversions and unsubscribes where available.
   2. Write down the business goal and primary success metric, such as purchases, booked calls, registrations or replies, and note any changes in audience, offer, timing or tracking between campaigns.
   3. Remove personal data from the export, keep the campaign identifiers and content fields, and combine the relevant rows into one spreadsheet or CSV.
   4. Paste the prompt and upload or paste the cleaned campaign data into an AI chatbot, then ask it to analyse only the supplied campaigns and to separate observations from hypotheses.
   5. Open the original campaign reports and compare the model's counts, rates and rankings against the platform figures, correcting any calculation or column-mapping errors.
   6. Use the ranked findings to select one content element for reuse and create follow-up tests that change one element at a time while keeping the primary metric fixed.
   7. Record the test results in the same spreadsheet and rerun the analysis only after enough comparable campaigns have accumulated to distinguish a pattern from a one-off result.

   Prompt

   Analyse the email campaign data below and identify which content performed best for the stated goal.
   
   Business goal: [for example, purchases, booked calls, registrations or replies]
   Audience and segments: [describe them]
   Campaign period: [dates]
   Email platform and tracking limitations: [describe them]
   Success metric and any secondary metrics: [define them]
   Campaign data:
   [paste a table or upload a CSV with campaign name, send date, audience, subject line, preview text, body copy or key content, send volume, delivered, opens, clicks, click rate, conversions, conversion rate, unsubscribes, bounces and revenue if available]
   
   Do the following:
   1. Check the column names, missing values and whether the campaigns are genuinely comparable. State any limitations before drawing conclusions.
   2. Compare the campaigns using counts as well as rates, and do not treat open rate as the main success measure unless it matches the stated goal.
   3. Rank the strongest content for the stated goal, separating subject line, offer, call to action, message angle and audience effects where the data allows.
   4. Flag results that could be explained by audience mix, send timing, deliverability, seasonality, list size or other confounders.
   5. Do not claim that one element caused the result unless the data supports that conclusion. Label observations, hypotheses and recommendations separately.
   6. Recommend a small number of follow-up tests, with one change per test and a clear primary metric.
   7. Show the calculations or source rows behind every important conclusion so I can compare them with the original email platform reports.
   8. End with a concise decision: what to reuse, what not to reuse and what still needs testing. Do not invent missing figures or campaign details.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.