Home · Business · Marketing & Content · Email & CRM campaigns

YES

As of 13 August 2026, AI can explain your UK email campaign results.

This still needs a person who signs their name to it.

Can you do it?

5 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededchat-fluent

Who has to check ita colleague

What the alternative costsPolymer is a no-code AI analytics and dashboard tool that creates insights from spreadsheets, so it is a purpose-built alternative for this work.

If this goes wrong: you mistake a correlation for a cause and repeat a campaign that underperformed with the wrong audience, message or timing.

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, chat-fluent skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Open your email platform's campaign report and export the results, including deliveries, bounces, opens, clicks, unsubscribes, complaints, conversions and revenue where available.
    2. Gather the campaign brief, target audience, exclusions, send date and time, subject line, preview text, offer, call to action and any changes from the previous campaign.
    3. Remove names, email addresses, tracking identifiers and other unnecessary personal data from the export, while retaining metric names, values, denominators, segments and comparison periods.
    4. Paste the prompt and the cleaned export into a chatbot, then add any previous campaign or internal benchmark that you are allowed to use.
    5. Ask the model to recalculate the percentages and totals, and to label every conclusion as observed, hypothesised or unproven.
    6. Compare the model's figures with the original platform report and check each proposed cause against the campaign brief, send log, links, landing page analytics and conversion records.
    7. Ask a colleague who knows the campaign to challenge the ranked explanations, then record the follow-up test you will run and the result that would change your conclusion.

    Prompt

    Explain the results of this UK email campaign using only the data and context I provide. Do not invent benchmarks, causes, customer motives or missing figures. Separate: 1) directly observed results, 2) reasonable hypotheses, and 3) claims that cannot be established from this campaign alone. Check all percentages and totals against the supplied figures, and flag any inconsistent or missing data.

Campaign objective: [objective]
Audience and exclusions: [audience description]
Send date and time, including time zone: [date and time]
Email subject line and preview text: [subject and preview text]
Main offer or call to action: [offer and CTA]
Relevant changes from previous campaigns: [changes]
Comparison campaign or benchmark, if available: [comparison]

Paste the campaign export below. Include the metric name, value, denominator where relevant, segment, and comparison period. Remove names, email addresses and other personal data before pasting.

[ campaign data ]

Return:
- a short plain-English summary
- a table of the key metrics and what each does and does not show
- the three most plausible explanations, ranked by how well the data supports them
- any segment differences worth investigating
- data quality problems or caveats
- three specific follow-up tests or checks
- actions that are supported by the evidence, clearly separated from suggestions that need more evidence
Do not recommend sending another campaign until you state what should be tested and what result would support or reject each hypothesis.

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

What it gets wrong

  • AI cannot establish whether a subject line, offer, audience, timing or landing page caused the result from one campaign alone.
  • It cannot know unrecorded factors such as deliverability problems, stock availability, sales follow-up or a competing message unless you provide them.
  • It cannot decide which trade-off matters most when open rate, click rate, revenue and unsubscribes point in different directions.
  • It can produce a confident explanation from incomplete or mismatched exports, so the source report still has to be checked.
  • It cannot replace a controlled test or a marketer's knowledge of your customers, brand and commercial constraints.

Even on a YES, the friction has a name: judgement under ambiguity, context depth and verification cost.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs2
Verification1
Liability2
Effort delta2
Total9 / 10

The methodology and its thresholds are published in full.

FAQ

Can ChatGPT analyse my email campaign results?
Yes. Give it a cleaned export and the campaign context, and it can calculate comparisons, spot patterns and separate observations from possible explanations. It cannot prove the cause of a result from a single campaign.
What data do I need to give AI to explain an email campaign?
Provide the campaign metrics with their denominators, audience and exclusions, send time, subject line, offer, call to action, conversion data and a comparison campaign if you have one. Remove names, email addresses and tracking identifiers before sharing the export.
Can AI tell me why my email open rate was low?
It can rank plausible explanations such as audience mix, timing, subject line or deliverability if the relevant evidence is present. It cannot tell you which explanation is true without further checks or testing.
Is it safe to upload my email campaign data to AI?
Use aggregated campaign data where possible and remove personal data that is not needed for the analysis. Check your employer's policy and the AI provider's data terms before uploading customer or prospect information.

Nearby answers

All Email & CRM campaigns rulings

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.