Best ChatGPT Prompts for Marketing Strategy

Why Most People Get Garbage Out of ChatGPT for Marketing

Most people use ChatGPT for marketing strategy the wrong way. They type something like "give me a marketing strategy for my business" and get back a five-paragraph essay that could apply to literally any company on earth. Useless.

The problem isn't ChatGPT. The problem is the prompt. Garbage in, garbage out - and vague prompts produce vague strategy.

Here's some context on how widespread this tool has become: 44% of surveyed marketers now use ChatGPT, making it the most common AI model for marketing and content work. And 71% of marketers who use AI tools at all report using ChatGPT specifically. It's table stakes at this point. The question isn't whether to use it - the question is whether you're using it well enough to get a real edge over the people who are doing it lazily.

I've spent years running outbound for agencies and SaaS companies, writing the cold emails myself, making the calls, building the lists. I've worked out which ChatGPT prompts actually produce useful marketing strategy outputs - positioning statements, ICP definitions, messaging frameworks, competitive angles, and channel plans you can actually execute on. That's what this article is about.

These aren't generic. They're built for founders, agency owners, and B2B sales professionals who want strategy that converts.

How to Get Good Output from ChatGPT (Before You Touch the Prompts)

One concept that changes everything: give ChatGPT a role and context before you ask your question. Don't just drop a question in cold. Prime it first.

A strong primer looks like this:

"You are a senior B2B marketing strategist with 15 years of experience helping SaaS companies and agencies generate enterprise pipeline. You speak directly, avoid fluff, and give specific recommendations over generic frameworks. I'm going to give you information about my business and I need your help building our go-to-market strategy."

Then you add context: what you sell, who you sell to, deal size, current channels, what's working, what isn't. The more context you feed in, the more specific the output. Think of it like briefing a consultant - they can't give you good advice without knowing the situation.

There are a few structural principles that separate prompts that produce real strategy from prompts that produce word salad:

• Assign a role. Tell ChatGPT who it is before you tell it what to do. "You are a..." changes the entire register of the response.
• Provide constraints. "Don't recommend channels I can't execute with a two-person team" forces trade-offs instead of wish lists.
• Specify format. "Give me this in a table" or "break this into three bullet points per section" produces structured, usable output instead of an essay.
• Include real data. Actual competitor names, real numbers, genuine pain points. The more specific your input, the more specific and useful the output.
• Ask for pushback. After any output, follow up with "What are the weakest assumptions in this plan?" That's where real insight lives.

One critical limitation to understand upfront: ChatGPT doesn't have access to real-time market data or live analytics. It's excellent for reasoning through strategy conceptually, but when it comes to current market conditions, you need to layer in real-world data from your CRM, your ad accounts, and your prospecting tools alongside it. More on that below.

With that said, here are the prompts that actually work.

Prompts for Defining Your Ideal Customer Profile (ICP)

Most marketing strategies fail because they're aimed at nobody in particular. The ICP prompt forces specificity - and specificity is the difference between a list of 50,000 random companies and a list of 500 companies who actually need what you sell.

Prompt 1 - ICP Builder:

"Based on this description of my business [insert description], help me define my ideal customer profile. Include: company size (employees and revenue), industry, job titles I should target, the business problem they have that my solution solves, what triggers them to look for a solution, and what objections they'll raise. Be specific. Don't give me ranges like 'small to mid-size businesses' - give me specific firmographics."

What you get back is a proper ICP you can actually use to build a prospect list. If you want to go further and pull actual contacts that match it, a B2B lead database lets you filter by title, seniority, industry, and company size once you know exactly who you're targeting.

Prompt 2 - ICP Validation:

"Here is my current ICP: [paste ICP]. What are three signals that would indicate a prospect matches this ICP and is actively experiencing the pain I solve? What are three disqualifiers that would tell me they're a waste of my time?"

This gives you a scoring framework your SDRs or outbound system can actually use - not just a demographic description, but a real qualifier checklist.

Prompt 3 - Buyer Persona Development:

"Using this ICP [paste ICP], build me a detailed buyer persona for the primary decision-maker I'll be selling to. Include: their typical day, their top three professional pressures, what success looks like in their role, how they prefer to consume information, what would cause them to forward my email to a colleague, and what would cause them to delete it immediately. Write it in second person - 'You are...' - so I can use it as a reference card during outreach."

This persona format is something you can paste directly into your cold email prompts and ad copy prompts later in the same session. It becomes the connective tissue across all your marketing output.

Prompt 4 - ICP Prioritization Matrix:

"I have identified three potential ICP segments: [Segment A], [Segment B], [Segment C]. Score each segment across five dimensions: size of the market, urgency of the pain, ease of reaching decision-makers, average deal value, and sales cycle length. Present the output as a comparison table and recommend which segment I should attack first and why."

Most founders try to go after everyone at once. This prompt forces a ranked decision. I've used it with agency owners who were splitting their time between three completely different markets and burning out without results - picking one, clearly, changes everything.

Prompts for Positioning and Messaging

Positioning is the most leveraged work in marketing. One shift in how you describe what you do can double response rates on cold email, ads, and website copy simultaneously. And it's also one of the hardest things to get right when you're too close to your own business.

Prompt 5 - Positioning Against Competitors:

"Here are three competitors in my space: [Competitor A], [Competitor B], [Competitor C]. Here's how they position themselves: [paste their taglines or homepage copy]. Here's what I do differently: [your differentiators]. Help me write a positioning statement that clearly separates me from these competitors and speaks directly to [target persona] who is frustrated with the current options. Give me three variations - one focused on speed, one on ROI, one on simplicity."

Prompt 6 - Message-Market Fit Test:

"Here is my current core value proposition: [paste it]. Here is my ICP: [paste ICP]. Score this value proposition on a scale of 1-10 for: specificity, relevance to ICP pain, believability, and differentiation. Then rewrite it to improve each score. Show me the before and after."

I've used variations of this to rewrite email subject lines, LinkedIn connection request messages, and cold call openers - all from one positioning session in ChatGPT.

Prompt 7 - Objection Mapping:

"Based on this value proposition [paste it] and this ICP [paste ICP], list the top seven objections a prospect would raise when I reach out cold. For each objection, write: the psychological root of the objection (what fear or belief drives it), a one-sentence reframe, and a longer response I could use during a sales call. Prioritize by how frequently each objection will come up."

This is underused. Most people use ChatGPT to write the pitch but forget to stress-test it against the buyer's objections. Running this exercise before you launch an outbound campaign means your follow-up sequences are already built around real resistance, not made-up FAQs.

Prompt 8 - Tagline and Hook Generator:

"Write 20 tagline options for [business description] targeting [ICP]. Vary the angles: some should lead with the outcome, some with the pain avoided, some with the time to result, some with the unique mechanism. Flag the top five you'd recommend and explain why. Do not use jargon, adverbs, or filler phrases like 'cutting-edge' or 'world-class.'"

The instruction to flag and explain the top five is important. Without it, ChatGPT gives you 20 options and leaves you to figure out which ones are actually good.

Prompts for Channel Strategy

ChatGPT is good at thinking through channel mix when you give it enough constraints. The key is forcing it to make trade-offs, not just list every possible channel. Companies report 25-45% productivity improvements using ChatGPT when it's applied to structured decisions like this - but only when the prompt forces a real recommendation instead of a hedge.

Prompt 9 - Channel Prioritization:

"I run a [type of business]. My ICP is [ICP]. My monthly marketing budget is approximately [budget range]. My team consists of [who handles marketing]. I currently use [current channels]. Based on this, recommend the top three channels I should focus on for the next 90 days. For each channel: explain why it fits my ICP, what my first 30 days of execution should look like, and what a realistic success metric is. Do not recommend channels I don't have the resources to execute."

The constraint at the end is important. Without it, ChatGPT will recommend a podcast, a YouTube channel, paid ads, and a webinar series - all at once - which is exactly how marketing teams spread themselves thin and produce nothing.

Prompt 10 - Outbound Channel Plan:

"I want to build a cold outbound system targeting [ICP]. I have [budget] and [team resources]. Design a multi-channel outbound sequence that includes cold email, LinkedIn, and cold calling. Give me the day-by-day touchpoint sequence, the message framework for each channel, and the follow-up logic (what triggers a follow-up vs. a disqualification)."

For outbound, cold email remains one of the highest-ROI channels when done right - I've broken down the full mechanics in my Cold Email GPT Prompts resource if you want the prompts purpose-built for that specifically.

Prompt 11 - Paid vs. Organic Decision Framework:

"My business is [description]. My ICP is [ICP]. My budget is [range]. I have [X months] before I need to show results. Help me decide how to split my attention between paid acquisition (ads) and organic acquisition (SEO, content, LinkedIn). Factor in: time to results, budget required, team skillset needed, and which approach better suits my ICP's behavior. Give me a clear recommendation, not a 'it depends' answer."

Most early-stage founders either go all-in on paid (expensive and fragile) or all-in on organic (slow). The reality depends on factors ChatGPT can reason through pretty well if you give it the constraints above.

Prompt 12 - Partnership and Co-Marketing Opportunity Finder:

"My business is [description] targeting [ICP]. Identify ten non-competing businesses or platforms that already have access to my ICP. For each, suggest: what a partnership or co-marketing arrangement could look like, what value I bring to them, and what a realistic first outreach message would say. Focus on arrangements I can initiate with a single email, not proposals requiring months of negotiation."

Partnerships are one of the most underutilized growth levers for agency owners and SaaS founders. This prompt surfaces options you probably haven't considered and makes them actionable immediately.

Prompts for Email Marketing Strategy

Whether you're running cold outbound, a nurture list, or a product newsletter, ChatGPT can help you architect the logic of the email strategy - not just write individual emails. There's a difference between asking it to "write me an email" and asking it to design the entire sequence architecture.

Prompt 13 - Email Nurture Sequence Architect:

"I have a list of [X] leads who opted in by downloading [lead magnet]. They are [ICP description]. They have not bought yet. Design a 10-email nurture sequence that moves them from 'interested but not ready' to 'ready to book a call.' For each email: give me the subject line, the core message, the call-to-action, and the psychological principle you're applying (e.g., social proof, loss aversion, specificity). Space the sends out over 21 days."

The psychological principle annotation is a prompt engineering trick worth using everywhere. It forces ChatGPT to be intentional about why each email is structured the way it is, which helps you make better editing decisions.

Prompt 14 - Subject Line A/B Test Generator:

"Here is my email topic: [topic]. Here is my ICP: [ICP]. Write 10 subject line options that I can A/B test. Use a variety of approaches: curiosity gap, direct benefit, personalization token, controversial statement, ultra-specific number, and question. For each, score it 1-10 for open rate potential and explain what might limit it."

Once you have verified emails to send to, tools like Smartlead or Instantly let you run proper A/B tests on subject lines at scale. But you need the variants first, and this prompt produces them fast.

Prompt 15 - Re-engagement Campaign Design:

"I have a list of [X] leads who went cold after initially showing interest. Their last point of contact was [timeframe] ago. They originally engaged with [context]. Design a five-email re-engagement campaign that acknowledges the gap, re-establishes relevance, and creates a low-friction next step. Do not use desperation tactics like 'Are you still interested?' Give me something with genuine value at each touch."

Also worth mentioning: before you send to any cold list, run it through an email validator to cut bounce rates and protect your sender reputation. ChatGPT can design the sequence; the infrastructure behind it still needs to be clean.

Prompts for Content and Thought Leadership Strategy

If you're doing inbound at all - blog, LinkedIn, YouTube - ChatGPT can help you build a content strategy that's actually tied to your ICP's buying journey, not just generic "value posts" that get likes from your competitors and nobody else.

Prompt 16 - Content Pillar Builder:

"My ICP is [paste ICP]. The problems they face are [list problems]. My product/service solves these by [how you solve them]. Build me a content pillar strategy with five core topics. For each topic: explain why it resonates with my ICP at each stage of the funnel (awareness, consideration, decision), list five specific content ideas, and tell me which format (LinkedIn post, blog article, video, email) is best suited for each idea."

Prompt 17 - Thought Leadership Angle Finder:

"Here is the conventional wisdom in [your industry]: [paste the standard takes everyone repeats]. Based on my experience and positioning as [describe yourself], what are three contrarian but defensible positions I could take publicly that would attract my ICP and differentiate me from other voices in the space? Back each one up with the core argument I'd need to make."

This one is gold for LinkedIn content. The contrarian angle gets 3-4x the engagement of the "5 tips for..." posts everyone defaults to. The key is "defensible" - you're not trolling, you're staking out a real position that's different from the consensus.

Prompt 18 - SEO Content Brief Generator:

"I want to rank for the keyword '[target keyword].' My ICP is [ICP]. Write a full SEO content brief that includes: the search intent behind this keyword, the ideal article structure (H1, H2s, H3s), the key questions the article must answer, the secondary keywords to incorporate naturally, and the ideal word count. Also identify the three most likely competing pages and what my article needs to do differently to outrank them."

This transforms a keyword into a full article brief in about 60 seconds. A human writer still needs to execute it - and should edit the final output heavily - but the brief structure ChatGPT generates here is genuinely solid.

Prompt 19 - LinkedIn Content Calendar:

"I post on LinkedIn five times per week. My ICP is [ICP]. My core expertise is [expertise]. My goal is to attract inbound leads, not just followers. Build me a four-week LinkedIn content calendar. Vary the post types across: personal story with lesson, contrarian take, tactical how-to, social proof/case study, and direct offer. For each post, give me the hook (first line), the structure, and the CTA. Flag which posts are most likely to go viral and which are most likely to generate DMs from prospects."

If you want tooling to schedule and analyze your LinkedIn performance, Taplio is worth looking at for LinkedIn-specific analytics and scheduling alongside this workflow.

Prompts for Competitive Intelligence and Market Research

ChatGPT's training data makes it decent for synthesizing known competitive landscapes. Pair it with actual market research tools and you've got a real edge.

Prompt 20 - Competitive Battlecard:

"I compete against [Competitor A] and [Competitor B]. Based on publicly available information, create a competitive battlecard for each that includes: their positioning, their most common sales objections about me, and the three strongest counter-arguments I can make. Format it so my sales team can use it during calls."

Prompt 21 - Market Segmentation Analysis:

"My product serves [broad market]. Break this market into five distinct customer segments based on the different problems they have, the different ways they buy, and the different budgets they operate with. For each segment, tell me: who the decision-maker is, what their primary buying trigger is, and what messaging approach would work best."

If you want to go deeper on using AI for market research, I've built out a dedicated set of prompts for that at GPT Market Research Prompts - it covers everything from TAM analysis to customer interview synthesis.

Prompt 22 - Win/Loss Analysis Framework:

"Here are five deals I won recently: [brief descriptions]. Here are five deals I lost: [brief descriptions]. Analyze the patterns. What are the common factors in deals I win? What are the common factors in deals I lose? Based on this, what changes should I make to: my ICP definition, my messaging, my qualification process, and my competitive positioning? Be specific."

This is one of the prompts I return to every quarter. The pattern recognition ChatGPT does on deal data is genuinely useful - it surfaces things you know intuitively but haven't articulated clearly enough to act on.

Prompt 23 - Customer Interview Synthesis:

"Here are transcripts or notes from five customer interviews: [paste notes]. Synthesize the most common pain points, the exact language customers use to describe their problems (pull direct quotes where possible), and the moments they describe as their 'before' (life without the solution) and 'after' (life with it). Then rewrite my current homepage headline and value proposition using their language, not mine."

This is one of the highest-leverage applications of ChatGPT for marketing. Taking real customer language and using it to rewrite your positioning is the fastest way to improve conversion rates across every channel simultaneously.

Prompts for Lead Generation Strategy

Marketing strategy and lead gen strategy are different but connected. ChatGPT can help you build the logic for a lead generation system - but you still need the actual data.

Prompt 24 - Lead Generation System Design:

"I need to generate [X] qualified leads per month for [product/service] targeting [ICP]. Design a lead generation system that covers: how I'll identify prospects, how I'll enrich their contact data, what my top-of-funnel offer will be to capture interest, and how I'll qualify leads before passing them to sales. Be specific about the tools and steps at each stage."

When it comes to actually executing the "identify and enrich" steps ChatGPT describes, that's where tools matter. ScraperCity's email finder is one tool worth plugging in at that stage for finding verified contact emails at scale.

Prompt 25 - Lead Magnet Strategy:

"My ICP is [ICP]. I want to create a lead magnet that attracts them at the top of funnel and captures their email. Suggest ten lead magnet ideas across different formats: checklist, template, calculator, mini-course, case study, swipe file, audit, quiz, webinar, and tool. For each, rate the likely opt-in conversion rate (low/medium/high), the production effort (low/medium/high), and the quality of lead it will attract. Recommend the top two and explain how I'd promote each one."

Most people default to "free ebook" and wonder why their list is full of people who never buy anything. This prompt forces you to think about what format actually attracts buying intent, not just information-seeking behavior.

Prompt 26 - Referral and Warm Outreach Strategy:

"I have [X] existing customers and [X] contacts in my network. Design a referral program and warm outreach strategy to generate [X] new leads from my existing relationships over the next 60 days. Include: the outreach message to existing customers asking for referrals, the incentive structure, the follow-up logic, and how I track results. Make the ask feel natural, not transactional."

For the full GPT-assisted lead gen workflow, check out the GPT Lead Gen Prompts resource - it covers the prospecting side in more depth.

Prompts for Cold Outbound Copywriting

Outbound is where most of my direct experience lives - I've written cold emails for dozens of agencies and SaaS companies, made the calls myself, and tested what actually books meetings. ChatGPT is useful here, but only if you understand what "good" looks like before you prompt it.

Prompt 27 - Cold Email Framework Generator:

"Write a cold email framework for [offer] targeting [specific ICP description]. The email should: be under 100 words, lead with a relevant observation about their business or industry (not a compliment), state the value in one sentence, and end with a low-friction CTA that asks for a yes/no response, not a 30-minute commitment. Write five variations using different opening angles. Do not use: 'I hope this finds you well,' 'I wanted to reach out,' or any version of 'Our company is...'"

Prompt 28 - Cold Call Opener Script:

"Write a cold call opener for [offer] targeting [role/industry]. The opener should: state who I am and why I'm calling in the first ten seconds, immediately connect to a business problem they likely have, and ask a question that gets them talking rather than saying 'I'm not interested.' Give me three versions: one that leads with a question, one that leads with a stat, and one that leads with a brief story. Each should be under 30 seconds to speak aloud."

If you're serious about cold calling, having mobile and direct-dial numbers matters a lot - reaching a general office line is about half as effective as reaching someone's direct number. A mobile finder tool can pull direct numbers once you've identified who you're targeting.

Prompt 29 - LinkedIn Connection Request Template:

"Write ten LinkedIn connection request messages for [ICP] targeting [role]. Each should be under 300 characters (LinkedIn's limit), feel like it came from a real person who looked at their profile, and create a reason to connect beyond 'I'd like to add you to my network.' Use different angles: shared industry observation, specific content they posted, mutual connection reference, and direct mention of something on their company page. Do not pitch in the connection request."

Prompt 30 - Follow-Up Sequence for Non-Responders:

"I sent an initial cold email [paste email] to [ICP] and got no response. Write a five-touch follow-up sequence spaced over 21 days. Each follow-up should: add new value or a new angle rather than just bumping the thread, be progressively shorter, and have a clear break-up email as the final touch. The tone should be persistent but never desperate. Include subject lines for each."

For tools to run these sequences at scale, both Lemlist and Reply.io handle multi-step outbound sequences well and integrate with most CRMs.

Prompts for Ad Copy and Paid Acquisition

Even if paid isn't your primary channel, understanding how to write ads well helps you understand positioning. The discipline of writing a Google ad - where you have 30 characters for a headline - forces a clarity that improves all your marketing copy.

Prompt 31 - Google Ad Copy Generator:

"Write Google Search ad copy for [offer] targeting people searching for [target keyword]. The searcher intent is [describe intent]. Write five headline variations (30 characters max each) and five description variations (90 characters max each). For each headline, specify: the angle (benefit, urgency, question, number-based, or comparison). Flag which combinations you'd recommend testing first."

Prompt 32 - Facebook and LinkedIn Ad Concept Builder:

"I want to run a paid social campaign for [offer] targeting [ICP]. Build five ad concepts that I could test. For each concept: describe the creative (image or video idea), write the primary text (under 125 words), write the headline (under 40 characters), and explain the psychological trigger it's using (curiosity, fear, aspiration, social proof, or specificity). Vary the angles so we're not just testing copy variations of the same concept."

Prompt 33 - Retargeting Ad Sequence:

"I have website visitors who viewed my [specific page] but didn't convert. Design a five-ad retargeting sequence for [platform]. Each ad should move the prospect one step closer to conversion by addressing a different objection or adding a new layer of social proof. Specify: the ad creative concept, the copy, the CTA, and the psychological job each ad is doing in the sequence."

Prompts for Campaign Planning and Execution

Once you have strategy, you need campaigns. ChatGPT is excellent at turning strategy into execution plans when you give it the right frame.

Prompt 34 - 90-Day Campaign Plan:

"Based on this marketing strategy [paste your strategy], build a 90-day campaign plan. Week by week, tell me: what campaign is live, what the primary call-to-action is, what assets need to be built before launch, and how I measure whether it worked. Assume my team has [X] hours per week to execute. Be realistic about what we can do - fewer campaigns done well beats more campaigns done poorly."

Prompt 35 - Launch Sequence for a New Offer:

"I'm launching [describe the offer] to [ICP]. Build a 14-day launch sequence that includes: a pre-launch LinkedIn or email teaser strategy, the launch day announcement, a follow-up nurture sequence for people who showed interest but didn't convert, and a post-launch retrospective checklist. Include the actual message frameworks for each touchpoint, not just descriptions."

Prompt 36 - KPI and Metrics Framework:

"My marketing goal for the next quarter is [goal]. Build a metrics framework that tells me: the top three KPIs I should track, the leading indicators that predict whether I'll hit each KPI, the lagging indicators that confirm whether the strategy worked, and the thresholds that would tell me to change course vs. stay the course. Format it as a dashboard I could review weekly in 15 minutes."

A tool like Monday.com works well for building this kind of campaign tracking view once ChatGPT gives you the framework - the AI gives you the structure, the project management tool gives you the place to actually run it.

Prompt 37 - Event and Webinar Strategy:

"I want to run a webinar targeting [ICP] to generate [X] qualified leads. Design the full strategy: the webinar topic and angle that would generate registrations (not just a product demo), the promotional sequence (email and LinkedIn), the webinar structure and talking points, the post-webinar follow-up sequence for attendees vs. no-shows, and how I close the loop into sales. Be specific about timing for each step."

Prompts for Marketing Analytics and Performance Review

One of the most underused applications of ChatGPT is using it to reason through data you already have. You can paste in your performance numbers and ask it to think through what's happening - not as a replacement for real analytics tools, but as a thinking partner for interpreting what the data means.

Prompt 38 - Campaign Performance Debrief:

"Here are the results from my last 90-day marketing campaign: [paste metrics - impressions, clicks, conversions, cost per lead, close rate, etc.]. Diagnose what worked and what didn't. For each channel, identify: the most likely reason for performance above or below expectations, what I should do more of, and what I should stop or change. Give me a ranked list of the three highest-leverage changes I could make for the next 90 days."

Prompt 39 - Funnel Leak Finder:

"Here is my marketing and sales funnel with conversion rates at each stage: [paste your funnel data]. Identify where the biggest drop-off is happening and diagnose the most likely cause at each stage. For the stage with the worst conversion, give me five specific hypotheses for why it's underperforming and five tests I could run to confirm or rule out each hypothesis."

Prompt 40 - Quarterly Strategy Review:

"Here is my original marketing plan for the quarter: [paste plan]. Here is what actually happened: [paste results]. Analyze the gap between plan and reality. What assumptions proved wrong? What surprised you positively? Based on this, what should my revised strategy be for next quarter? Don't just recommend doing more of what worked - challenge whether the 'wins' were actually strategic or just lucky, and whether the losses were systemic or one-off."

The instruction at the end is key. Without it, ChatGPT tells you to do more of what worked and less of what didn't - which is obvious and not particularly useful. Asking it to challenge the wins forces more rigorous thinking.

Prompts for Agency-Specific Marketing Strategy

If you run an agency - marketing, design, development, PR, whatever - you have a specific set of marketing problems that don't map cleanly to the generic prompts most articles give you. You're not selling a widget. You're selling time, expertise, and outcomes, and your buyers are skeptical because they've been burned before.

Prompt 41 - Agency Positioning Differentiator:

"I run a [type] agency. Here is our current positioning: [paste it]. Here are the three most common ways agencies in my space position themselves: [list them]. Help me find a genuine positioning angle that is: based on something we actually do differently (not just claiming we do), specific enough to repel bad-fit clients, and attractive enough to command a premium price point. Ask me up to ten questions about our work before giving me your recommendation."

The "ask me questions first" instruction is one of the most powerful prompt engineering moves you can use. It turns ChatGPT from an answer machine into a consultant. The questions it asks will often surface angles you hadn't considered.

Prompt 42 - Case Study Extraction:

"Here is information about a client result we achieved: [paste details]. Turn this into a case study that: opens with the client's situation before working with us (specific, not vague), describes the strategy we implemented with enough detail to demonstrate expertise without giving away everything, and closes with specific measurable results. Write it in a way that speaks to [target client type] who has the same problems. Also give me a 100-word summary I can use in cold emails and a three-sentence version for LinkedIn."

Prompt 43 - Agency Service Productization:

"I currently offer [describe your agency services] in a custom/bespoke way. Help me design three productized service packages from these services. For each package: name it, describe what's included, specify what's excluded, set a suggested price range logic (not a specific number), and write the positioning statement that explains who it's for and what outcome they get. The goal is to make our services easier to sell, easier to deliver, and easier for prospects to say yes to quickly."

Advanced Prompt Engineering Techniques for Marketing

Once you've got the basic prompts down, here are the meta-techniques that make everything better.

Chain of Thought Prompting

Add "Think through this step by step before giving me your answer" to any complex strategy prompt. This forces ChatGPT to reason before concluding, which produces significantly better output on anything involving trade-offs or recommendations.

The Skeptic Role Play

After any strategy output, use this: "Now play the role of a skeptical CMO who has seen this kind of strategy fail before. What are the three biggest flaws in this plan, and what would you change?" This surfaces holes before you execute, not after.

The Constraint Tightener

If the output is too broad, add: "Now redo this assuming I have zero budget, two hours per week, and one person executing. What's the minimum viable version of this strategy?" Constraints produce creativity. The minimum viable version of a strategy is often the best starting point.

The Persona Switcher

After you get a piece of copy or messaging, prompt: "Now read this from the perspective of [specific persona - e.g., a skeptical 48-year-old CFO at a 200-person manufacturing company who has been pitched by five agencies this month]. What's their first reaction? What word or phrase immediately loses their trust? What would make them read the next sentence?"

The Version Comparison

Always ask for at least three variations of any output and ask ChatGPT to label what's different about each one. Never accept a single version. The first version is the obvious answer. The third version is where the interesting thinking happens.

Building Your Prospect List to Match the ICP You Just Defined

One thing that trips people up: ChatGPT can help you define who to target, but it can't go build you a list of those people. That's a separate step, and it's where the strategy meets execution.

Once you've run the ICP prompts above and you know your specific firmographics - the company size, industry, job titles, and geography - you need a way to pull actual contacts that match. ScraperCity lets you filter an unlimited B2B database by exactly those parameters: job title, seniority, industry, company size, and location. The ICP definition ChatGPT gives you maps directly to the filter logic in the database.

If you're prospecting locally - say you run an agency targeting restaurants, law firms, or contractors in a specific city - a Google Maps scraper pulls local business contact data that you can then run outreach against. The workflow is: ChatGPT defines who, the tool finds them, your outbound system contacts them.

And before any list goes into an email tool, run it through an email validator to remove bad addresses. Bounce rates above 3-5% start damaging your sender reputation, which kills deliverability for everyone on the list, not just the bad addresses.

Integrating ChatGPT into Your Marketing Workflow (Not Just Using It One-Off)

Most people treat ChatGPT like a search engine - they go in, get an answer, close the tab. That's leaving most of the value on the table. The real gain comes from building it into your process at specific, repeatable stages.

Here's how I'd structure a basic marketing workflow with ChatGPT integrated:

• Monthly strategy session (60 min): Start a new chat with the role-setting primer. Paste your current ICP, your metrics from last month, and your goals for next month. Run the quarterly review prompt and the channel prioritization prompt. Exit with a clear 30-day focus.
• Weekly campaign planning (30 min): Paste this week's goals into a fresh chat. Run the content calendar prompt or the sequence architecture prompt depending on what you're building. Get your framework for the week.
• Asset creation (per asset): Use a fresh chat per asset type. Cold email gets its own session. LinkedIn posts get their own session. Don't mix contexts - it degrades the output quality.
• Post-campaign review (45 min): Paste actual results into the performance debrief prompt. Get the diagnosis. Add the changes to next month's strategy session input.

The discipline of using it at these specific stages - not just whenever you're stuck - is what separates teams that get consistent value from it versus teams that use it sporadically and wonder why the output is inconsistent.

Common Mistakes to Avoid with ChatGPT Marketing Prompts

A few failure modes I see constantly:

Accepting the first output. ChatGPT's first response is always the most predictable answer. It's the average of what good looks like. Iterate. Ask it to challenge its own assumptions. Ask for three more versions that are completely different in approach.

Using it for final copy without a human pass. The strategy and structure ChatGPT produces is genuinely useful. The prose it writes usually sounds like everyone else's AI-generated content - because it was trained on everyone else's content. Use it for thinking and structure, then write the actual words yourself.

Not providing competitive context. ChatGPT doesn't know your specific market unless you tell it. Paste in competitor positioning, actual customer feedback, and real deal notes. The more real-world data you inject, the more specific and useful the output.

Using it to avoid hard decisions. ChatGPT is good at generating options. It's not good at telling you which option is right for your specific situation. The strategic judgment - which channel to prioritize, which ICP to focus on, which offer to lead with - that's still on you. Use it to think more clearly, not to outsource the decision.

Treating it as a substitute for real customer research. The best marketing insights come from talking to actual customers, reading their exact words, and listening to sales call recordings. ChatGPT can help you synthesize that data once you have it, but it can't replace the act of collecting it.

Getting the Most Out of These Prompts

A few things I've learned running these at scale:

• Always paste real data. The more you give ChatGPT - your actual offer, real competitor names, specific numbers - the more specific and useful the output.
• Iterate, don't accept the first draft. After any output, follow up with "What are the weakest assumptions in this plan?" or "What would a skeptical CMO push back on here?" That's where the real insight comes from.
• Use ChatGPT for thinking, not for final copy. The strategy and structure it produces is genuinely useful. The prose it writes usually needs a human pass to not sound like everyone else's AI-generated content.
• Layer in real data sources. ChatGPT knows markets conceptually - it doesn't know who specifically is in your pipeline or what your conversion rates actually are. Use it alongside real data from your CRM and your prospecting tools.
• Build a prompt library. Every time you find a prompt combination that produces great output, save it. Over time you'll have a set of go-to prompts tuned for your specific business, and you won't have to start from scratch each session.

If you want to build an AI-assisted proposal workflow alongside your marketing strategy work, the Proposal AI Templates resource covers that side specifically.

If you want to go beyond prompts and actually implement a full marketing and sales system, I work through this hands-on inside Galadon Gold - drop in and see if it's a fit.

The prompts above will get you unstuck and thinking more clearly about your marketing strategy. The execution is still on you - but at least now you've got the right questions to ask, in the right order, with enough context for ChatGPT to give you something worth acting on.