Build a practical prompt pack for my digital marketing work. Base every recommendation on the business, audience, offer, channels, and evidence I provide. Do not invent customer research, performance data, claims, testimonials, compliance approval, or competitor facts. Business and offer: [BUSINESS AND OFFER] Audience and buying context: [AUDIENCE] Goals and current metrics: [GOALS AND METRICS] Channels and available assets: [CHANNELS AND ASSETS] Brand, legal, and operating constraints: [CONSTRAINTS] Produce 10 distinct copy-paste prompts for ChatGPT, Claude, or Gemini: audience-message map, campaign brief, landing-page audit, paid-ad concept matrix, email sequence outline, SEO content brief, social-content calendar, lead-nurture diagnosis, experiment design, and monthly performance analysis. Each prompt must be 100-150 words and contain its own square-bracket inputs so a marketer can reuse it without returning to this conversation. For each prompt, include a one-line use case, the exact requested output, and domain checks appropriate to the task. For example, paid-ad prompts should require angle, audience segment, proof, CTA, and policy-risk review; SEO briefs should distinguish search intent, primary query, supporting questions, internal-link opportunities, and claims requiring sources; reporting prompts should separate correlation from causation and ask for a next test. Order the pack by planning, production, distribution, and measurement. Finish with a simple selection guide that maps common situations to one of the 10 prompts. Before answering, check that every prompt has a measurable outcome or decision, and flag where I need customer research or channel data before using it. Ask up to 3 clarifying questions only if a required input is missing.
Fill in
| Placeholder | What to enter | Example |
|---|---|---|
| [BUSINESS AND OFFER] | Enter what you sell, the pricing or conversion action, and the key customer problem it addresses. | B2B scheduling software for independent clinics; $249 per location monthly; demo request is the primary conversion. |
| [AUDIENCE] | Enter the target segments, roles, needs, objections, buying triggers, and purchasing process. | Operations managers at 3-20 location physical therapy groups; frustrated by no-shows and manual reminders; want proof before switching systems. |
| [GOALS AND METRICS] | Enter current marketing goals, baselines, targets, time period, and tracking limitations. | Generate 45 qualified demos per month; current rate is 27; paid search CPL is $118 and demo-to-opportunity rate is 31%. |
| [CHANNELS AND ASSETS] | Enter active channels, audience sizes, budget, content, creative, website pages, and marketing tools available. | Google Ads, LinkedIn, 8,000-contact email list, webinar recordings, two case studies, $9,000 monthly paid budget, HubSpot. |
| [CONSTRAINTS] | Enter brand voice, prohibited claims, approval rules, regulated-industry requirements, and team capacity. | Plain, practical tone; do not claim guaranteed revenue gains; legal approves customer stories; one marketer and a freelance designer. |
How to use
- Fill the five inputs with current channel data and actual customer language, then generate the prompt pack.
- Choose one prompt for a real campaign rather than trying to use all ten at once.
- Check claims, audience assumptions, tracking events, and policy constraints in the output before publishing.
- Follow up with: “Adapt the paid-ad concept prompt for LinkedIn, using our $3,000 test budget and one case study.”
Variations
Campaign brief
Use this when you need a campaign plan before making assets.
Create a campaign brief from these inputs. Produce audience segment, insight or evidence, offer, message hierarchy, funnel stage, channel roles, asset list, measurement plan, budget assumptions, and a 30-day test plan. Include one primary hypothesis and two alternatives. Do not invent customer insights or performance benchmarks; label unverified assumptions. [OFFER] [AUDIENCE RESEARCH] [CAMPAIGN GOAL] [CHANNELS] [BUDGET AND DATES]
Content brief
Use this when assigning an SEO article or educational asset.
Create an SEO content brief for a writer. Include likely search intent, reader job-to-be-done, primary query, supporting questions, outline, evidence needed, examples to seek, internal links, CTA, and quality bar. Separate factual claims needing sources from opinion or guidance. Do not promise rankings or fabricate keyword volume. [TOPIC] [PRIMARY QUERY] [AUDIENCE] [PRODUCT CONTEXT] [EXISTING PAGES]
Performance review
Use this when a campaign needs a decision based on results.
Analyze the campaign results below and produce an executive summary, funnel table, segment comparison, likely explanations, data limitations, and three prioritized next actions. Identify which findings are descriptive versus causal. Recommend a single next experiment with hypothesis, success metric, duration, and stop rule. Do not infer conversion causes from small samples without flagging uncertainty. [CAMPAIGN GOAL] [PERFORMANCE DATA] [CHANGES MADE] [SEGMENTS] [TRACKING LIMITATIONS]
Tips
- Give AI the conversion event and the baseline for that event; clicks and impressions alone rarely determine what to change.
- Keep a claims ledger for ads and landing pages that links every proof point to a source or approved customer evidence.
- Use one hypothesis per experiment, with a preselected success metric and minimum run conditions, so post-test interpretation is less subjective.
- Ask for channel-specific outputs because a LinkedIn ad, nurture email, and SEO article require different intent and proof structures.
FAQ
Can AI create a complete marketing strategy from a short description?
It can create a working draft, but customer research, channel economics, and capacity determine whether the strategy is usable.
How do I avoid generic AI marketing copy?
Provide customer language, objections, proof, offer details, and channel context. Then require the model to label unsupported claims.
Should I use the same prompt for every channel?
No. Keep your positioning consistent, but specify the audience intent, asset constraints, CTA, and measurement model for each channel.