Ready-to-paste ChatGPT prompts for SEO, PPC, content, email, analytics, and strategy. Each prompt includes the exact text, what it produces, and a pro tip for better results. Tested by our team at ScaleGrowth.Digital across 40+ client engagements.
Last updated: March 2026 · Reading time: 18 min
Every prompt in this collection has been tested on real client work. We didn’t pull these from a generic list. Each one was used on actual campaigns, refined based on output quality, and scored on three criteria: specificity of output, time saved versus doing it manually, and consistency across repeated use. Prompts that produced vague or generic results were cut. What remains are the 50+ prompts our strategists at ScaleGrowth.Digital reach for weekly.
A ChatGPT marketing prompt is a structured instruction given to ChatGPT that produces a specific marketing deliverable, such as ad copy, keyword clusters, email sequences, or competitive analysis summaries.
According to a HubSpot survey (2025), 64% of marketers now use AI tools weekly, but only 22% report being satisfied with the quality of AI-generated output. The difference between the two groups? Prompt quality. A well-structured prompt with context, constraints, and format instructions outperforms a vague ask by 3-5x in usability.
We scored each prompt on a 1-5 scale across three dimensions:
| Dimension | What it measures | Minimum score to include |
|---|---|---|
| Output specificity | Does it produce something you can use immediately? | 4/5 |
| Time saved | How much faster than doing it manually? | 3/5 |
| Consistency | Does it produce similar quality every time? | 4/5 |
These 12 SEO prompts cover keyword research, on-page optimization, content briefs, and technical SEO analysis. They’re the same prompts our AI visibility team uses to accelerate research that used to take hours. The key to SEO prompts is feeding ChatGPT enough context about your domain, audience, and competitive position.
The prompt:
I'm targeting the keyword "[primary keyword]" for a [type of business] targeting [audience]. Generate 30 related keywords grouped by search intent (informational, commercial, transactional, navigational). For each keyword, estimate relative search volume (high/medium/low) and competition level. Format as a table.
What it produces: A categorized keyword list with intent mapping. Saves 30-45 minutes compared to manual Ahrefs exploration.
Pro tip: Add “Exclude branded keywords and keywords with monthly search volume likely under 100” to filter noise.
The prompt:
Here are 50 keywords related to [topic]: [paste list]. Group these into topic clusters where each cluster could be a single page. Name each cluster, identify the primary keyword for each, and list supporting keywords. Suggest a URL slug for each cluster page.
What it produces: Content architecture mapped to keyword clusters. We used this on a SaaS client and reduced planned pages from 47 to 18 by identifying overlap.
Pro tip: Paste actual search volume data alongside keywords for more accurate grouping.
The prompt:
Write 3 meta descriptions for a page about "[topic]" targeting the keyword "[keyword]". Each must be 150-160 characters, include the keyword naturally, contain a clear call-to-action, and avoid clickbait. The brand is [brand name] and the audience is [audience description].
What it produces: Three ready-to-use meta descriptions. CTR improvements of 8-15% are common when replacing generic descriptions with targeted ones (Search Engine Journal, 2024).
Pro tip: Include your current meta description and ask ChatGPT to beat it. Competition against a benchmark produces tighter copy.
The prompt:
Create a content brief for the keyword "[keyword]" with monthly search volume of [MSV]. Include: target word count, search intent, recommended H2 headings (as questions), 5 People Also Ask questions to answer, 3 competitor URLs to analyze, suggested internal links, recommended schema type, and a unique angle that differentiates from existing results.
What it produces: A complete writing brief that a content writer can execute without additional research. For a deeper template, see our content brief template.
Pro tip: Paste the top 3 Google results’ H2 headings into the prompt so ChatGPT can identify gaps.
The prompt:
Generate 5 title tag options for a page targeting "[keyword]". Each must be 50-60 characters, front-load the keyword, and include a compelling modifier (number, year, "free", "guide", "template"). The page is about [brief description]. Current title is "[current title]".
What it produces: Five title variations ranked by likely CTR. Pages with numbers in titles get 36% more clicks according to a Moz study (2023).
Pro tip: Ask it to also generate a “power word” version and a “curiosity gap” version.
The prompt:
Generate 7 FAQ questions and answers for a page about "[topic]". Each answer must be 40-80 words, factually accurate, and written in a direct tone. Format each Q&A pair so it can be directly used in FAQPage schema markup. The audience is [audience] and they care about [specific concerns].
What it produces: Schema-ready FAQ pairs. Pages with FAQPage schema see an average 8% CTR improvement (Ahrefs, 2024).
Pro tip: Check Google’s “People Also Ask” for your target keyword first and include those questions as a starting point.
The prompt:
Here's a list of pages on my website with their URLs and topics: [paste sitemap or page list]. I'm writing a new page about "[topic]". Suggest 5-8 internal links I should include, with the recommended anchor text for each and where in the content it should appear.
What it produces: A linking map with contextual anchor text. Internal links remain one of the most underused SEO tactics.
Pro tip: Also ask it to identify which existing pages should link back to your new page.
The prompt:
Generate JSON-LD schema markup for a [Article/Product/LocalBusiness/FAQPage/HowTo] page with the following details: [paste page details]. Include all required and recommended properties per Schema.org specs. Output valid JSON-LD that can be pasted into the <head> of the page.
What it produces: Copy-paste schema markup. Always validate the output in Google’s Rich Results Test before publishing.
Pro tip: Ask for nested schema (e.g., Article with author as Person with Organization affiliation) for richer entity signals.
The prompt:
I run a [type of business]. My top 3 competitors are [URLs]. Based on their likely content strategy, what topics are they probably covering that I'm not? Suggest 10 content gaps I should fill, with recommended keywords for each and estimated content type (blog post, landing page, tool, or guide).
What it produces: A gap analysis that maps to content planning. Pair this with actual Ahrefs content gap data for best results.
Pro tip: Feed it your existing sitemap so it can identify true gaps rather than topics you’ve already covered.
The prompt:
Classify the search intent for each of these keywords as informational, commercial investigation, transactional, or navigational. For each, recommend the ideal page type (blog post, comparison page, product page, landing page, tool). Keywords: [paste list of 20-30 keywords]
What it produces: Intent mapping that determines what type of page each keyword needs. Mismatched intent is the #1 reason pages rank on page 2 instead of page 1.
Pro tip: Include the current top 3 SERP results for ambiguous keywords so ChatGPT can calibrate.
The prompt:
Here's the content from a page published [date] targeting "[keyword]": [paste content]. The page currently ranks #[position] and gets [X] clicks/month. Analyze what's likely causing it to underperform. Suggest specific changes: what to add, what to remove, what to restructure. Focus on information freshness, content gaps vs. current top results, and E-E-A-T signals.
What it produces: A prioritized refresh plan. Content refreshes drive 106% more organic traffic on average versus publishing net-new content (HubSpot, 2025).
Pro tip: Always paste the content rather than asking ChatGPT to fetch the URL. It can’t browse reliably.
The prompt:
I want to build topical authority on "[broad topic]" for a [type of site]. Create a topical map with 3 pillar pages and 8-12 cluster pages for each pillar. For each page, include: title, primary keyword, content type, and which other pages it should link to. Show the hierarchy visually using indentation.
What it produces: A complete content architecture. We’ve used this approach to plan 6-month content roadmaps for clients at ScaleGrowth.Digital.
Pro tip: Specify your domain authority and current content volume so it calibrates ambition to your starting point.
PPC prompts need tight constraints. Ad platforms have character limits, policy restrictions, and performance metrics that generic prompts miss. These 8 prompts produce ad copy, audience targeting ideas, and campaign structures that align with Google Ads and Meta Ads specs. Google processed 8.5 billion searches per day in 2025 (Statista), making search ads the highest-intent channel available.
The prompt:
Write 15 Google Ads headlines for [product/service] targeting the keyword "[keyword]". Each headline must be 30 characters or less. Include a mix of: benefit-focused (5), urgency-focused (3), social proof-focused (3), and question-based (4). The brand name is [brand]. Do not use ALL CAPS or excessive punctuation.
What it produces: 15 policy-compliant headlines ready for upload. Google Ads allows up to 15 headlines per responsive search ad.
Pro tip: Include your top 3 performing headlines and ask ChatGPT to generate variations that test different angles.
The prompt:
Write 4 Google Ads descriptions for [product/service]. Each must be 90 characters or less. Include: the main benefit, a specific number or proof point, and a call-to-action. Keyword: "[keyword]". Landing page offer: [describe offer]. Avoid superlatives that Google Ads will flag.
What it produces: Four description lines. Pair with the headline prompt above for a complete RSA asset set.
Pro tip: Tell it your industry. Google Ads policies vary by vertical (finance, health, legal have stricter rules).
The prompt:
Write 3 Facebook/Instagram ad copy variations for [product/service] targeting [audience]. Each version should include: a hook (first line that stops scrolling), 2-3 lines of body copy, and a CTA. Version 1: problem-agitation-solution format. Version 2: testimonial/social proof format. Version 3: direct offer format. Keep each under 125 words.
What it produces: Three distinct ad angles. Meta’s own data shows ads with varied creative see 27% lower cost per acquisition.
Pro tip: Specify the audience’s pain point explicitly. “Busy CMOs frustrated by reporting” beats “marketing professionals.”
The prompt:
I'm advertising [product/service] on [Google/Meta/LinkedIn]. My ideal customer is [persona description]. Suggest 10 audience targeting options including: interest-based targets, behavioral targets, custom intent keywords (for Google), lookalike seed ideas, and any demographic overlays. Explain why each would work for this product.
What it produces: A targeting strategy with rationale. Saves the trial-and-error of testing random interests.
Pro tip: Include your customer lifetime value (LTV) so it can suggest targeting strategies appropriate to your budget tolerance.
The prompt:
I'm running Google Ads for [product/service] targeting keywords like [list 5 main keywords]. Generate a negative keyword list of 30+ terms that would trigger irrelevant clicks. Include: free/cheap seekers (if not applicable), job seekers, DIY/tutorial seekers, wrong geographic intent, and competitor brand terms I shouldn't bid on. Format as a list I can paste into Google Ads.
What it produces: A paste-ready negative keyword list. Poor negative keyword coverage wastes 10-20% of ad spend in most accounts (WordStream, 2024).
Pro tip: Run this prompt once per quarter and merge with your search terms report to catch emerging irrelevant queries.
The prompt:
Write landing page copy for [product/service] targeting [keyword]. The page should follow this structure: headline (under 10 words), subheadline (1 sentence value proposition), 3 benefit blocks with icons (headline + 2 sentences each), social proof section (format for 3 testimonials), objection handling section (3 common objections with responses), CTA section with primary and secondary buttons. Target conversion rate benchmark: [industry average]%.
What it produces: Complete landing page wireframe copy. The average landing page converts at 5.89% across industries (Unbounce, 2024).
Pro tip: Paste your current landing page copy and ask ChatGPT to identify the biggest conversion leak before writing the replacement.
The prompt:
I'm running [Google/Meta/LinkedIn] ads for [product/service]. Current performance: CTR [X]%, conversion rate [X]%, CPA $[X]. Suggest 5 A/B test hypotheses I should run next. For each, specify: what to test (creative, copy, targeting, or landing page), the hypothesis in "If we [change], then [metric] will [improve/decrease] because [reason]" format, and expected impact.
What it produces: Structured test ideas with prioritization logic. Most ad accounts run tests randomly rather than systematically.
Pro tip: Include your monthly spend so it recommends tests achievable within your traffic volume for statistical significance.
The prompt:
Design a Google Ads campaign structure for [business] with a monthly budget of $[X]. Include: campaign names, ad group themes, 5 keywords per ad group, match types for each keyword, and suggested daily budget allocation. The goal is [leads/sales/awareness]. Organize by [product line/service/funnel stage].
What it produces: A campaign architecture you can build directly in Google Ads. Saves 2-3 hours of manual planning.
Pro tip: Specify whether you want a SKAG (single keyword ad group) or themed ad group structure and why.
Content prompts fail when they’re too open-ended. “Write a blog post about X” produces mediocre results every time. The prompts below constrain format, audience, tone, and structure so the output is draft-ready rather than brainstorm-grade. Content marketing generates 3x more leads per dollar than paid advertising according to the Content Marketing Institute (2025).
The prompt:
Create a blog post outline for the topic "[topic]" targeting the keyword "[keyword]". The audience is [persona]. Include: a compelling title with the keyword, a meta description (150-160 chars), 6-8 H2 headings as questions, 2-3 H3 subheadings under each H2, a suggested intro approach, and a conclusion with CTA. Target word count: [X] words. The post should be better than what currently ranks because it includes [unique angle].
What it produces: A structured outline that a writer can execute in 60-90 minutes. Without an outline, blog posts take 3-4 hours on average.
Pro tip: Paste the H2s from the current #1 ranking article and ask ChatGPT to identify what’s missing.
The prompt:
Create 10 LinkedIn posts for a [B2B/B2C] brand in the [industry] space. The brand voice is [describe voice]. Mix of formats: 3 text-only thought leadership posts, 3 data-driven posts with a stat hook, 2 "lessons learned" posts, 2 engagement posts (polls or questions). Each post should be 150-200 words. No hashtag spam (max 3 hashtags per post).
What it produces: A week’s worth of social content. LinkedIn posts with specific data points get 37% more engagement (LinkedIn Marketing Solutions, 2025).
Pro tip: Include 2-3 examples of your best-performing past posts so ChatGPT can match your actual voice.
The prompt:
Here's a 2,000-word blog post: [paste content]. Repurpose it into: 5 LinkedIn posts (different angles from the same content), 3 Twitter/X threads (8-10 tweets each), 1 email newsletter intro (150 words), 1 infographic outline, and 1 YouTube video script outline (5 minutes). Each repurposed piece should stand alone and not feel like a copy-paste.
What it produces: A complete repurposing plan from a single piece. One blog post can become 10-15 distribution assets.
Pro tip: Tell it which platform performs best for your brand so it prioritizes that format.
The prompt:
Write a case study from these notes: Client: [industry, size, challenge]. What we did: [strategy/tactics]. Results: [specific metrics]. Timeline: [duration]. Structure it as: Challenge (2 paragraphs), Approach (3-4 paragraphs), Results (data table + narrative), Key Takeaway (1 paragraph). Tone: confident and data-driven. Word count: 800-1,000 words.
What it produces: A publication-ready case study. Case studies convert 70% of B2B buyers during the consideration stage (Demand Gen Report, 2024).
Pro tip: Include a direct quote from the client (even a paraphrased one). It adds credibility ChatGPT can’t manufacture.
The prompt:
Design a content pillar strategy for [brand/topic]. I need 1 pillar page (3,000+ words, comprehensive guide) and 8-10 cluster articles (1,000-1,500 words each). For each piece: title, primary keyword, word count, content type, and how it links back to the pillar. The pillar keyword is "[keyword]" with [X] monthly searches.
What it produces: A hub-and-spoke content architecture. This is the model we use at ScaleGrowth.Digital for every content engagement.
Pro tip: Specify your publishing cadence (e.g., 2 posts/week) so it sequences the cluster articles logically.
The prompt:
Write a product description for [product name]. Specs: [list specs]. Target buyer: [persona]. Write 3 versions: Version 1: SEO-focused (150 words, includes keyword "[keyword]", structured with bullet points). Version 2: Conversion-focused (100 words, leads with primary benefit, includes social proof hook). Version 3: Marketplace listing (follows [Amazon/Shopify] format guidelines).
What it produces: Three channel-specific descriptions. E-commerce sites with unique product descriptions see 35% more organic traffic than those using manufacturer copy (Seer Interactive, 2023).
Pro tip: Include 1-2 customer review quotes and ask ChatGPT to weave their language into the description.
The prompt:
Write a script for a [X]-minute YouTube video about "[topic]". Target audience: [persona]. Structure: Hook (first 15 seconds, must create curiosity), Intro (who you are, what they'll learn, 30 seconds), Main content (3-4 key points, 1-2 minutes each), Recap (30 seconds), CTA (subscribe + [specific action]). Include suggested B-roll notes in brackets. Tone: [conversational/authoritative/educational].
What it produces: A shoot-ready video script. YouTube videos between 7-15 minutes get the highest watch time according to VidIQ (2025).
Pro tip: Ask for “pattern interrupt” moments every 2 minutes to maintain viewer retention.
The prompt:
Generate 20 headline variations for [topic/content piece]. Mix these proven formats: How-to (4), Numbered list (4), Question (4), Negative/mistake (4), Comparison (4). Each headline should be 6-12 words. The keyword is "[keyword]". Rank them by likely click-through rate and explain why your top 3 picks would win.
What it produces: A headline bank for testing. BuzzSumo’s analysis of 100 million headlines (2024) found that list-based headlines outperform other formats by 2x.
Pro tip: Use the CoSchedule Headline Analyzer score as feedback and iterate.
The prompt:
Create a 4-week content calendar for [brand] in [industry]. Publishing cadence: [X blog posts/week, X social posts/day, X emails/week]. For each piece: date, content type, topic, target keyword, funnel stage (TOFU/MOFU/BOFU), distribution channels. Include 2 seasonal/timely hooks relevant to [month/quarter]. Balance between educational (60%), promotional (20%), and engagement (20%) content.
What it produces: A month of planned content. For a more detailed template, download our content calendar template.
Pro tip: Feed it your Google Analytics top pages so it creates content that builds on what’s already working.
The prompt:
I'm competing with [3 competitor URLs] for the keyword "[keyword]". Analyze what content strategy they're likely using based on their publicly visible content. For each competitor: estimated content volume, publishing frequency, content types they favor, topics they cover that I don't, and 3 specific content pieces I should create to beat them.
What it produces: A competitive content gap report. Saves 4-6 hours of manual competitive analysis.
Pro tip: Paste actual competitor page titles and URLs rather than just domain names for more specific output.
The prompt:
Here's a blog post that gets [X] monthly visits but has a bounce rate of [X]% and conversion rate of [X]%: [paste content or URL]. Suggest 5 content upgrades (lead magnets) specific to this topic that would convert readers into email subscribers. For each: what it is, format (PDF, spreadsheet, tool, quiz), estimated effort to create (low/medium/high), and the CTA copy to use.
What it produces: Lead magnet ideas matched to existing content. Content upgrades convert 3-5x better than generic newsletter opt-ins.
Pro tip: Ask for one “quick win” upgrade you can create in under 2 hours.
The prompt:
Write a thought leadership article for [executive name/title] at [company] about [controversial or forward-looking topic in industry]. Take a clear position (not both-sides wishy-washy). Include: a provocative opening statement, 3 supporting arguments with data, 1 counter-argument addressed honestly, and a conclusion that calls for a specific industry change. Tone: authoritative but not arrogant. Word count: 800-1,200 words.
What it produces: An opinion piece with backbone. The best thought leadership takes positions. It doesn’t summarize consensus.
Pro tip: Include the executive’s actual experiences and anecdotes. ChatGPT can structure the argument, but the credibility comes from real stories.
Email prompts need to account for inbox behavior. Subject lines compete with 100+ other emails. Body copy gets scanned, not read. These 7 prompts produce emails that respect those realities. Email marketing delivers $36 for every $1 spent, the highest ROI of any marketing channel (Litmus, 2025).
The prompt:
Write 10 email subject lines for [campaign type: welcome, promotional, newsletter, re-engagement, abandoned cart]. The offer/content is [describe]. Audience: [persona]. Mix formats: curiosity (3), benefit (3), urgency (2), personalization (2). Each must be under 50 characters. No spam trigger words (free, act now, limited time). Preview text suggestion for each.
What it produces: 10 testable subject lines with preview text. Subject lines under 50 characters have 12% higher open rates (Campaign Monitor, 2025).
Pro tip: Include your average open rate so ChatGPT calibrates ambition. If you’re at 45%, it should aim higher than if you’re at 18%.
The prompt:
Write a 5-email welcome sequence for new subscribers of [brand/product]. Email 1: Welcome + deliver lead magnet (send immediately). Email 2: Brand story + what to expect (Day 2). Email 3: Most popular content/product (Day 4). Email 4: Social proof + case study (Day 7). Email 5: Soft sell + CTA to [action] (Day 10). Each email: subject line, preview text, body copy (150-250 words), CTA button text. Brand voice: [description].
What it produces: A complete onboarding sequence. Welcome sequences generate 320% more revenue per email than standard promotional emails (Invesp, 2024).
Pro tip: Include a “reply to this email” CTA in email 1. Replies improve deliverability and build a relationship signal with ESPs.
The prompt:
Write a weekly email newsletter for [brand] in [industry]. This week's content: [list 3-4 topics/links]. Format: personal intro (2-3 sentences from [sender name]), 3 content blocks (headline + 2-sentence summary + link for each), 1 curated industry news item with our take, closing CTA. Total word count: 300-400 words. Tone: like a smart colleague sharing their weekly finds.
What it produces: A newsletter draft in 5 minutes. Newsletters with a consistent personal voice have 25% higher engagement.
Pro tip: Always edit the intro paragraph yourself. The personal voice is what keeps subscribers from unsubscribing.
The prompt:
Write a 3-email re-engagement sequence for subscribers who haven't opened an email in 90+ days. Email 1: "We miss you" angle with a compelling reason to come back (incentive or exclusive content). Email 2: "Last chance" angle with a direct question about whether they want to stay subscribed. Email 3: Breakup email (unsubscribe them automatically unless they click). Keep each under 100 words. Subject lines that cut through a crowded inbox.
What it produces: A win-back sequence. Re-engagement campaigns recover 10-15% of inactive subscribers on average.
Pro tip: Include the subscriber’s first name and the date they last engaged. Specificity gets attention.
The prompt:
Write a cold outreach email to [target role] at [company type/size]. The goal is to [book a meeting/get a reply/share a resource]. Our value proposition: [1 sentence]. Include: personalized opening (reference something specific about their company), 1 sentence on what we do, 1 social proof point, a low-friction CTA (not "book a 30-minute call"). Keep under 100 words total. Write 3 versions with different opening hooks.
What it produces: Three outreach variations. Cold emails under 100 words get 51% more replies (Boomerang, 2023).
Pro tip: The personalization placeholder is where your real work is. ChatGPT can write the structure; you must fill in the genuine research.
The prompt:
Write a 3-email abandoned cart sequence for [e-commerce brand]. Email 1: Reminder (sent 1 hour after abandonment), friendly, no discount. Email 2: Social proof (sent 24 hours), include review snippets and urgency cue. Email 3: Discount offer (sent 48 hours), [X]% off with expiration. Include subject line, body copy (under 100 words each), and CTA button text. Product: [product type and price range].
What it produces: A tested recovery sequence. Cart abandonment emails recover 5-11% of lost revenue (Baymard Institute, 2025).
Pro tip: Don’t offer the discount in email 1. Many shoppers return without an incentive; leading with a discount trains them to wait.
The prompt:
I'm sending [campaign type] emails to a list of [X] subscribers. Current open rate: [X]%, click rate: [X]%, conversion rate: [X]%. Design 3 A/B tests I should run over the next month. For each: what element to test, hypothesis, version A vs version B copy, required sample size for statistical significance, and expected impact. Prioritize by likely impact.
What it produces: A structured testing roadmap. Most email teams test subject lines but ignore send time, preview text, and CTA placement.
Pro tip: Only test one variable at a time. Multi-variable tests need 4x the sample size for valid results.
ChatGPT won’t access your Google Analytics dashboard, but it excels at interpreting data you paste in, generating report structures, and identifying patterns in exported datasets. These 6 prompts turn raw numbers into actionable insights. Companies that use data-driven marketing are 23x more likely to acquire customers (McKinsey, 2024).
The prompt:
Here's our website traffic data for the last 6 months: [paste data table with columns: month, sessions, users, bounce rate, avg session duration, conversions, conversion rate]. Analyze trends, identify anomalies, and suggest 3 hypotheses for any significant changes. Also identify: the strongest month and why, the weakest month and possible causes, and 3 specific actions to improve next quarter.
What it produces: A narrative analysis of raw data. Turns spreadsheet numbers into a story your stakeholders can act on.
Pro tip: Include external context (seasonal events, campaigns launched, site changes) so it can correlate data shifts to causes.
The prompt:
Here's our monthly marketing report data: [paste metrics across all channels]. Write an executive summary for [C-level audience]. Structure: 3-sentence overview of performance, top 3 wins with specific numbers, top 3 concerns with recommended actions, 1-paragraph outlook for next month. Keep it under 300 words. Tone: confident and direct, no jargon.
What it produces: A boardroom-ready summary. Executives read summaries, not 30-page reports.
Pro tip: State the comparison period explicitly (MoM, YoY, vs target) so the summary has context.
The prompt:
Design a marketing KPI dashboard for a [B2B SaaS/e-commerce/lead gen] business. Include: 5 primary KPIs (the ones the CMO checks daily), 10 secondary KPIs by channel (SEO, PPC, email, social, content), suggested visualization type for each (line chart, bar chart, scorecard, pie chart), refresh frequency, and data source for each metric. Format as a table.
What it produces: A dashboard spec that a BI team or tool like Looker Studio can implement in hours.
Pro tip: Include your reporting cadence (daily/weekly/monthly) so it groups metrics appropriately.
The prompt:
Explain [last-click/first-click/linear/time-decay/data-driven] attribution to a marketing team that currently uses last-click only. Include: a simple analogy, how it changes which channels get credit, a specific example with a 5-touchpoint customer journey, and 3 scenarios where switching to this model would change budget allocation decisions. Keep it under 400 words.
What it produces: An internal education piece that justifies changing your attribution model.
Pro tip: Use your actual channel mix in the example. Generic examples don’t convince budget holders.
The prompt:
Write a [X]-question customer survey for [purpose: NPS, product feedback, brand perception, churn analysis]. Include: 3 quantitative questions (rating scale), 3 qualitative questions (open-ended), and 2 demographic/segmentation questions. For each question: the question text, answer format, and what insight it reveals. Keep the survey completable in under 5 minutes. Avoid leading or biased questions.
What it produces: A deployable survey. Surveys over 10 questions see a 40% drop-off rate (SurveyMonkey, 2024).
Pro tip: Put the most important question first. Many respondents don’t finish, so capture the critical insight early.
The prompt:
Here's our marketing funnel data: [paste stage, volume, conversion rate for each step]. Identify: the biggest drop-off point, the likely reasons for that drop-off (3 hypotheses), and 3 specific tactics to improve conversion at that stage. Also calculate: overall funnel conversion rate, cost per acquisition at each stage (total spend: $[X]), and the revenue impact of improving the weakest stage by 20%.
What it produces: A funnel diagnostic with quantified improvement opportunities. Makes the business case for optimization investment.
Pro tip: Include industry benchmarks for each stage so the analysis shows where you’re above or below average.
Strategy prompts require the most context to work well. Give ChatGPT your business model, competitive position, budget constraints, and goals. Without this context, you’ll get textbook strategies that don’t account for your specific situation. These 8 prompts produce frameworks, plans, and analyses grounded in real constraints.
The prompt:
Create a detailed buyer persona for [product/service]. I know the following about our customers: [paste any data: demographics, purchase behavior, survey results, support tickets]. Build the persona with: name, role/title, company size, goals (3), frustrations (3), information sources (where they learn), buying triggers (what makes them act), objections (what makes them hesitate), and preferred communication channels. Make it specific enough that a sales rep could have a conversation with this person.
What it produces: A usable buyer persona. Teams with documented personas generate 2x more qualified leads (ITSMA, 2024).
Pro tip: Feed it 5-10 actual customer support tickets or sales call notes. Real language beats assumptions.
The prompt:
Analyze [my brand] vs [competitor 1] and [competitor 2] across these dimensions: positioning/messaging, pricing model, content strategy, SEO visibility, social media presence, product/service differentiation, and target audience. For each dimension, identify: who's winning and why, where [my brand] has an advantage, and where we have a gap. Summarize in a comparison table with red/yellow/green ratings. End with 3 strategic recommendations.
What it produces: A structured competitive assessment. Pair with Ahrefs and SimilarWeb data for accuracy.
Pro tip: Include links to competitors’ about pages, pricing pages, and recent blog posts for more specific analysis.
The prompt:
Create a go-to-market plan for launching [product/service] to [target market]. Budget: $[X]/month for first 3 months. Include: launch timeline (pre-launch, launch week, post-launch), channel strategy with budget allocation by channel, content calendar for launch month, PR/outreach targets, success metrics for 30/60/90 days, and a risk mitigation plan for the top 3 risks. Be specific with tactics, not just strategy labels.
What it produces: A tactical GTM plan. Most product launches fail because they have strategy but no execution detail.
Pro tip: Specify what you already have (existing audience, email list size, social following) so the plan builds on existing assets.
The prompt:
Create a brand messaging framework for [company] that [description of what the company does]. Core values: [list]. Target audience: [primary persona]. Include: brand positioning statement (1 sentence), elevator pitch (30 seconds), value proposition (for website hero), 3 key messages with supporting proof points, tone of voice guidelines (with do/don't examples), and boilerplate text (50 words for press, 100 words for website).
What it produces: A messaging doc that aligns marketing, sales, and product communications. Consistent brand messaging increases revenue by 10-20% (Lucidpress, 2024).
Pro tip: Include messaging you’ve used that didn’t work so ChatGPT can steer in a different direction.
The prompt:
Create a Q[X] marketing plan for [business type]. Quarterly revenue target: $[X]. Current performance: [paste key metrics]. Available channels: [list active channels]. Team: [X people, roles]. Create a plan with: 3 quarterly objectives (OKR format), monthly milestones, budget allocation by channel with expected ROI, content themes by month, and 5 key experiments to run. Format as an executable plan, not a strategy document.
What it produces: A 90-day execution plan. Quarterly planning is more effective than annual planning because it allows for faster pivots.
Pro tip: Include last quarter’s plan vs actuals so the new plan accounts for what worked and what didn’t.
The prompt:
Conduct a SWOT analysis for [brand] in the [industry] market. Context: [paste company description, recent performance, market position]. For each quadrant, provide 4-5 specific points (not generic). Strengths and Weaknesses should reference internal factors. Opportunities and Threats should reference external market forces. After the SWOT, generate 3 "SO strategies" (using Strengths to capture Opportunities) and 3 "WT strategies" (mitigating Weaknesses to avoid Threats).
What it produces: A SWOT with actionable strategies, not just a 2×2 box of bullet points.
Pro tip: The SO and WT strategies are the valuable output. The SWOT grid alone is a brainstorming exercise; the strategies are the decisions.
The prompt:
I have a monthly marketing budget of $[X] for [business type]. Current channel mix: [list channels with current spend and performance]. Recommend a revised budget allocation based on: highest ROI channels first, minimum viable spend per channel, test budget for 1-2 new channels, and a contingency reserve (10%). Show current vs recommended allocation in a table with expected performance change.
What it produces: A data-informed reallocation recommendation. Most marketing budgets are set historically (“same as last year”) rather than by performance.
Pro tip: Include customer acquisition cost (CAC) and lifetime value (LTV) by channel. Without this, all allocation advice is guesswork.
The prompt:
Here's our current marketing tech stack: [list tools with monthly cost and primary use]. Evaluate: which tools have overlapping functionality, which are underutilized (based on typical feature usage), where we have gaps, and suggest 3 replacements or additions with estimated cost. Total current spend: $[X]/month. Recommend a streamlined stack that covers the same needs for [X]% less.
What it produces: A tech stack audit with consolidation recommendations. The average company uses 91 marketing technology tools (Gartner, 2025), most of which overlap.
Pro tip: Include your team size. A tool designed for enterprise teams isn’t useful for a 3-person marketing department.
“The prompt is the product now. We’ve seen marketers cut content production time by 60% not by using AI more, but by writing better prompts. A 50-word prompt that includes context, constraints, and format instructions outperforms a 10-word ask every single time. The real skill isn’t prompting ChatGPT. It’s knowing what good output looks like before you ask for it.”
Hardik Shah, Founder of ScaleGrowth.Digital
After testing over 200 prompts across client work, five patterns consistently separate prompts that produce usable output from prompts that produce generic filler.
Pattern 1: Context stacking. The more context you give, the better the output. Include your industry, audience, current performance, and competitive position. A prompt with 50 words of context produces 3x better output than one without.
Pattern 2: Format specification. Tell ChatGPT exactly how you want the output formatted. “Give me a table with columns for X, Y, Z” beats “list some ideas.” Format constraints eliminate the need to reformat output after generation.
Pattern 3: Role assignment. Starting with “You are a [specific role] with [specific experience]” focuses the output. “You are a B2B SaaS content strategist who has managed 50+ blog calendars” produces different output than a generic ask.
Pattern 4: Negative constraints. Telling ChatGPT what NOT to include is as useful as telling it what to include. “Do not use jargon,” “avoid generic advice,” or “no motivational fluff” eliminates the padding that makes AI content feel hollow.
Pattern 5: Iterative refinement. Your first prompt is a draft. Follow up with “Make the tone more conversational” or “Now give me a version that’s half the length.” The best marketers treat ChatGPT as a conversation, not a vending machine.
Copy-pasting these prompts without customization will get you 60% of the way there. The remaining 40% comes from filling in the brackets with genuine specifics about your brand, audience, and market position. Here’s how to get the most from this collection.
Build a prompt context block. Create a saved text snippet (use TextExpander, Notion, or a simple Google Doc) with your brand description, target personas, voice guidelines, and competitor names. Paste this context block into every prompt. It takes 30 seconds and dramatically improves output quality.
Save your best outputs as examples. When ChatGPT produces something great, save it. Next time you use a similar prompt, include “Here’s an example of the output quality I expect: [paste example].” This trains the model on your quality bar.
Test with real campaigns. Don’t evaluate prompts in isolation. Use the ad copy prompts on a live campaign and measure CTR. Use the email prompts on a real send and track open rates. Performance data tells you which prompts actually work for your audience.
For more AI-powered SEO prompts, see our dedicated ChatGPT prompts for SEO collection. And if you want to understand how AI is changing search visibility beyond prompts, explore our AI marketing tools guide.
40+ SEO-specific prompts for keyword research, content briefs, technical audits, and competitor analysis.
A curated review of AI tools for content, analytics, ads, and email with real performance benchmarks.
A free Google Sheets content calendar with monthly, weekly, and quarterly planning views.
You can use structured outputs like keyword lists, schema markup, and data tables with minimal editing. Creative outputs like ad copy, blog posts, and email sequences should always be edited for brand voice, factual accuracy, and compliance with platform policies. Plan on editing 20-40% of creative output before publishing.
GPT-4o and GPT-4.5 produce the best marketing output as of March 2026. GPT-4o is faster and cheaper for high-volume tasks like headline generation and keyword clustering. GPT-4.5 produces more detailed and layered creative writing for thought leadership and long-form content. For budget-conscious teams, GPT-4o mini handles structured tasks like schema generation and data formatting well.
Include 2-3 examples of your existing content that represents your ideal voice. Add explicit voice instructions: “Our tone is professional but not stiff, we use contractions, avoid jargon, and prefer short sentences.” You can also create a Custom GPT with your brand guidelines pre-loaded so every conversation starts with your voice constraints built in.
These prompts work with any large language model including Claude, Gemini, and Copilot. The prompt engineering principles (context, constraints, format specification) are universal. You may need to adjust slightly for each model’s strengths: Claude tends to produce longer, more analytical output, while Gemini integrates better with Google Workspace data.
Review and update your prompt library quarterly. AI models improve every few months, and prompts that needed heavy constraints in older models may produce better output with less instruction in newer ones. Also update prompts when your marketing strategy shifts, new channels are added, or your brand voice evolves. Delete prompts that consistently produce output you never use.
Our AI visibility practice helps brands appear where AI recommends. From prompt-optimized content to structured data that AI models prefer to cite, we build the foundations that make your brand the answer.