Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Yes—ChatGPT and Claude can help plan campaigns, draft sequences, generate subject lines, and audit email copy, but the quality depends on the prompt you give them. The most reliable prompts specify a role, goal, audience, offer, approved source facts, constraints, output format, and checks for unsupported claims. This practical kit gives you 23 copy-ready prompts, from campaign strategy through performance review, plus a workflow for adapting them to either model.
Replace every bracketed field with your information. Put customer research, product details, legal rules, and approved claims inside clearly labeled blocks. Treat the output as a draft: a person must verify facts, consent, privacy, accessibility, and the final send settings.
As an Amazon Associate I earn from qualifying purchases.
What makes an email-marketing prompt work in ChatGPT or Claude?
A useful prompt removes guesswork. State the job the model is doing, who the email is for, what the business is offering, which facts it may use, what it must avoid, and exactly how the answer should be formatted. Clear, specific instructions consistently outperform clever wording.
Use this seven-part prompt header
- Role: the perspective the model should use, such as email strategist or compliance editor.
- Goal: the business outcome and the reader action you want.
- Context: audience, funnel stage, product, channel, geography, and timing.
- Source facts: approved information, enclosed in labeled delimiters such as
<source_facts>...</source_facts>. - Constraints: word count, tone, claims, consent rules, accessibility requirements, and exclusions.
- Output format: headings, a table, a fixed number of variants, or JSON fields.
- Checks: instructions to flag missing inputs, unsupported claims, ambiguity, or compliance risks.
For a task that depends on missing information, tell the model to ask questions before drafting. If voice or formatting is important, include one approved example and label it as an example rather than a fact source. Ask for the result and a concise list of checks; do not ask for hidden chain-of-thought.
#1 Best Overall
How should source material be delimited?
Separate facts from instructions so the model can tell what it is allowed to assert. For example:
<source_facts>
Product: [name]
Price: [amount and currency]
Approved claims: [claim 1]; [claim 2]
Proof: [case study or review text]
</source_facts>Add a rule such as “Use only the source facts for factual claims. If a detail is absent, write missing input and do not guess.” This is especially important for prices, deadlines, performance promises, testimonials, health statements, and legal language.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Which prompts help with strategy and planning?
Use the first five prompts before writing copy. They turn scattered business information into a brief, audience model, offer angle, calendar, and safe segmentation plan.
1. Build a campaign brief
This prompt forces one objective and one primary call to action instead of a crowded email.
Act as an email strategist. Using the business, audience, offer, funnel stage, and source facts below, create a one-page campaign brief with objective, single CTA, message hierarchy, risks, and success criteria. Flag missing information before making recommendations.
<business_context>
[business, audience, offer, funnel stage]
</business_context>
<source_facts>
[approved facts]
</source_facts>2. Extract audience insight without inventing it
Requiring a quote for every insight keeps the model from presenting guesses as established audience evidence.
Using only the customer evidence below, list the audience’s jobs, pains, objections, desired outcomes, vocabulary, and proof needs. Quote the source text for each insight and label any inference. If the supplied evidence does not support an item, write “not established.”
<customer_research>
[paste interviews, survey responses, or support themes]
</customer_research>3. Position competing offers
Ask for a trade-off and a claim-safe value proposition rather than a generic “best” statement.
Rank #2
Compare these [offers] for [segment]. Recommend one email angle, explain the trade-off, and write a one-sentence value proposition without inventing benefits. Use only the supplied facts and identify any missing proof.
<offers>
[offer details]
</offers>4. Create an email content calendar
A structured calendar balances education, proof, promotion, and retention across a defined period.
Create a [30-day] email calendar for [audience]. For each send, provide purpose, topic, CTA, funnel stage, and source facts used. Balance education, proof, promotion, and retention. Do not repeat a topic or invent a deadline.
<business_and_facts>
[context and approved facts]
</business_and_facts>5. Design useful, non-sensitive segments
Give the model behavioral or declared attributes and explicitly prohibit sensitive-trait inference.
Design three segments from the supplied customer attributes. For each, give inclusion rules, exclusion rules, message angle, and one personalization field with a fallback when blank. Do not infer sensitive traits or use protected characteristics.
<customer_attributes>
[available fields and definitions]
</customer_attributes>Which prompts produce email copy?
These prompts generate the assets marketers use most often. Keep the source-fact and unsupported-claim rules in every production request, not only in the first conversation.
6. Generate subject lines
Character limits and labeled risk checks make a list easier to review and test.
Write 20 subject lines for [campaign]. Keep each under [character limit]. Vary curiosity, benefit, urgency, and clarity. Mark any line that could imply an unsupported claim, false scarcity, guaranteed result, or unverified personalization.
<campaign_facts>
[offer, audience, approved claims, real deadline if any]
</campaign_facts>7. Write preview text
Preview text should add information rather than repeat the subject line.
Write [10] preview-text options that complement these subject lines, avoid repetition, and stay within [character limit]. Keep every factual statement within the approved facts.
Subject lines:
[paste subject lines]
Approved facts:
[paste facts]8. Draft a welcome email
Specify the next step and one CTA so the first message does not become a product brochure.
Draft a welcome email for [new subscriber]. Use this brand voice, explain the next step, include one CTA, and separate supplied facts from suggested wording. Do not add benefits, guarantees, or personal details that are not provided.
<brand_voice>
[voice guidance and approved example]
</source_facts>
[product and onboarding facts]
</source_facts>9. Outline a welcome sequence
Sequence-level planning prevents five emails from doing the same job.
Outline a [5]-email welcome sequence with send timing, objective, key message, proof, CTA, and stop conditions. Ask for missing inputs first. Use one consistent promise and do not invent a deadline or testimonial.
Audience: [audience]
Offer or product: [details]
Approved facts: [facts]10. Turn education into a nurture sequence
Assigning a distinct job to each email creates progression instead of repetition.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteTurn this educational material into a [4]-email nurture sequence. Give each email a distinct job, subject line, key teaching point, CTA, and the source passage it uses. Carry one consistent promise through the series and flag unsupported conclusions.
<educational_material>
[paste material]
</educational_material>11. Write a promotional email with a fact-check table
The table makes it obvious which claims still need human verification.
Write a promotional email for [offer] to [segment]. Include headline, body, CTA, and concise objection handling. Include a fact-check table with columns: claim, source, and verification needed. Use a deadline only if supplied.
<offer_facts>
[price, terms, availability, proof, deadline if real]
</offer_facts>12. Create product-launch variants
Announcement, consideration, and last-call messages need different jobs and different levels of urgency.
Create launch-email variants for announcement, consideration, and last call. For each, provide subject line, preview text, body, CTA, and the approved facts used. Reuse only supplied product facts and label assumptions. Do not create scarcity or a last-call date unless provided.
<launch_facts>
[launch date, product details, availability, proof]
</launch_facts>13. Recover an abandoned browse or cart event
A helpful reminder should not pressure people or imply scarcity that the data does not confirm.
Draft a recovery email for [browse or cart event]. Use a helpful tone, one CTA, and clear information about what happens next. Do not use pressure, guilt, or scarcity unless the source data confirms it. Include a fallback for unavailable items.
Event: [what the customer did and when]
Product facts: [facts]
14. Build a re-engagement series
Offer escalating value and an easy preference or unsubscribe path instead of guilt-based copy.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #4
Write three re-engagement emails with escalating value. Include a preference-update or unsubscribe-friendly option, a distinct purpose for each send, and stop conditions. Avoid guilt-based language and use only the supplied reasons to return.
<subscriber_context>
[last activity, available preferences, content or offer options]
</subscriber_context>15. Clarify a transactional email
Transactional messages require precision: preserve every operational detail while improving readability.
Rewrite this service or order email for clarity, scannability, accessibility, and brand voice. Preserve every operational fact, date, amount, link destination, and required instruction. List any ambiguity separately; do not alter policy or legal meaning.
<original_email>
[paste email]
</original_email>Which prompts improve personalization and optimization?
Use these prompts after you have a defensible draft. They help you personalize safely, vary tone without changing claims, handle objections, test one variable at a time, edit for accessibility, and check compliance.
16. Map personalization fields safely
Map these available fields to safe personalization opportunities. For each field, give an example, a fallback when blank, and a risk of sounding intrusive. Exclude sensitive data and explain when generic copy is safer.
<available_fields>
[field names, sources, and consent status]
</available_fields>17. Convert tone without changing meaning
Rewrite this email in [three brand-approved tones]. Keep the meaning, offer, claims, price, and CTA unchanged. Show a short rationale for each version and flag any wording that could alter factual meaning.
<approved_tones>
[tone definitions]
</approved_tones>
<email>
[paste draft]
</email>18. Handle objections from evidence
List the five strongest objections supported by the supplied customer evidence, then write one concise response to each without adding evidence. Quote the supporting passage and label an objection as unconfirmed when the source does not establish it.
<research>
[paste customer evidence]
</research>19. Create an A/B test plan
Require a hypothesis and a decision rule so testing produces a learning, not just a winner.
Create an A/B test matrix for subject line, opening, CTA, and send timing. State one hypothesis, primary metric, guardrail metric, audience split, test duration, and decision rule per test. Change one major variable at a time and note any sample-size or tracking input that is missing.
Campaign goal: [goal]
Available metrics: [metrics and definitions]
20. Edit copy for clarity and accessibility
Edit this draft for clarity, scannability, accessibility, and brand voice. Return tracked-style changes plus a clean version. Preserve factual meaning, links, prices, dates, consent language, and CTA. Flag text that needs a human decision.
<draft>
[paste email]
</draft>
<brand_rules>
[reading level, voice, accessibility and legal rules]
</brand_rules>21. Run a compliance check
This is a screening aid, not legal advice. Supply the actual rules that apply to your business and recipients.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallReview this email against the supplied legal, consent, privacy, and brand rules. Quote the exact risky sentence, explain the issue in plain language, and propose a compliant rewrite. Do not give legal advice beyond the supplied rules. Identify missing consent, data-use, unsubscribe, disclosure, or evidence inputs.
<rules>
[paste applicable rules]
</rules>
<email>
[paste email]
</email>22. Review campaign performance
Using this campaign report, separate observed results from hypotheses. Identify the three largest opportunities, calculate only from supplied numbers, and recommend the next test. Show formulas for any calculated rate and state which data is missing.
<campaign_report>
[paste sends, deliveries, opens, clicks, conversions, revenue, and definitions]
</campaign_report>23. Repurpose an email across channels
Turn this email into a short SMS, social post, landing-page section, and internal brief. Preserve the offer, audience, claims, and CTA. List channel-specific changes, character limits applied, and any detail that cannot be carried over accurately.
<source_email>
[paste approved email]
</source_email>How do ChatGPT and Claude differ for this workflow?
Both can follow a well-structured brief, transform source material, and revise drafts. The practical differences are usually workflow and plan dependent, so check current product documentation before choosing a model for a specific campaign.
| Comparison axis | What to evaluate | Prompting practice |
|---|---|---|
| Instruction clarity | Whether the model follows role, constraints, and prohibited-claim rules | Use direct verbs, numbered steps, and an explicit “do not guess” rule. |
| Context-window needs | How much research, product documentation, or prior copy you need to provide | Split very large inputs into labeled sections and ask for a source map. |
| Formatting control | Reliability of tables, fixed counts, JSON, or tracked-style edits | Show the exact fields, columns, or schema and request a validation check. |
| Long-source handling | Whether important details remain available throughout a long task | Ask the model to list source sections used and identify omissions. |
| Tone consistency | Ability to maintain voice across a sequence | Include an approved example, a tone list, and words to avoid. |
| Source and citation handling | Whether claims can be traced to supplied material | Require a quote or source label for every non-obvious claim. |
| Revision workflow | How easily you can request targeted changes without rewriting everything | Give numbered feedback, then ask for only the affected sections. |
| Cost and plan availability | Current limits, model access, and business-data controls | Verify current terms and privacy settings before uploading customer data. |
What is a reliable review process before sending AI-assisted email?
Use a staged review rather than trusting a polished first draft.
- Check inputs: confirm audience, consent status, offer terms, price, dates, availability, and the single CTA.
- Check claims: compare every factual sentence with an approved source. Remove invented statistics, testimonials, guarantees, urgency, and personalization.
- Check fit: read as the recipient. Ensure the message matches the funnel stage and does not reveal sensitive information.
- Check accessibility: use descriptive link text, logical headings, readable contrast, plain language, and meaningful alternative text for images.
- Check mechanics: test links, personalization fallbacks, unsubscribe behavior, rendering, tracking, and mobile layout in your email marketing platform.
- Check experiments: record the hypothesis, primary metric, guardrail, audience, and decision rule before launching an A/B test.
- Revise narrowly: tell the model exactly what failed—such as “remove unsupported urgency from paragraphs two and three”—and request only that change.
What should you never give an AI model in an email prompt?
Do not paste passwords, private keys, payment-card data, raw health records, or customer information that your organization is not allowed to process in that service. Minimize personal data, remove identifiers where possible, and follow your company’s retention, access, and vendor-review rules. A prompt cannot create consent or make an unlawful use of data acceptable.
Also avoid asking a model to conceal sponsorship, fabricate reviews, impersonate a person, bypass opt-out requests, or make a legal determination. Ask it to flag questions for a qualified reviewer instead.
Frequently Asked Questions
Can ChatGPT or Claude send my email campaign automatically?
They can draft copy and, in some setups, pass structured content to other software, but the model itself is not your consent database or sending authority. Keep segmentation, suppression lists, approvals, scheduling, and delivery in your email marketing platform, and require a human approval step.
How can I keep AI from making up testimonials or statistics?
Put approved proof in a labeled source block and add a hard rule that absent evidence must be marked “missing input.” Require a source label or quotation for every statistic, testimonial, benchmark, or customer result, then verify each item against the original record.
Should I use one giant prompt for an entire five-email sequence?
Plan the sequence first, then draft one email at a time with the same source facts and promise. Smaller, staged requests make it easier to catch repetition, contradictions, unsupported claims, and changing CTAs.
What metrics should I use to judge whether a prompt improved email copy?
Track the metric tied to the campaign goal, such as qualified clicks or completed purchases, alongside guardrails such as unsubscribe, complaint, bounce, and conversion quality. Compare against a defined baseline and document the audience, send conditions, and test variable so a result is interpretable.
Recommended Free Tools
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
