100 ChatGPT Cheat Codes for Marketing
I use ChatGPT often enough that I eventually got tired of typing some of the same instructions over and over. Sometimes I want it to challenge an idea instead of helping me justify it. Sometimes I want it to look at something like a first-time customer, find patterns in reviews, tell me what information is missing, or cut the fluff out of something I wrote.
So I started giving those instructions names. I call them cheat codes.
These are not official ChatGPT commands. Typing /truthmode does not unlock a secret setting inside ChatGPT. It is simply shorthand for a set of instructions I have defined. Once ChatGPT knows what I mean by the code, I can use it again without rewriting the full explanation every time.
You can create your own codes too. In fact, that is really the point of this list. Pay attention to the instructions you keep repeating, name the ones that are useful, and eventually move the more complicated ones into a Skill File, Project instruction, or another reusable system.
I grouped these by the marketing job you are trying to accomplish.
I Need to Figure Out What Is Wrong With My Marketing
Before asking ChatGPT for more ads, posts, promotions or campaign ideas, I often want help understanding the problem.
1. /truthmode
Don’t automatically support my idea. Look for weak assumptions, missing evidence, risks, contradictions and reasons my conclusion may be wrong.
Marketing example: Before launching a new paid workshop, give ChatGPT the audience, price, topic and expected demand, then use /truthmode to see whether the idea actually holds up.
2. /rootcause
Look underneath the visible problem instead of assuming the first explanation is the right one. Identify several possible causes and what evidence would help distinguish between them.
Marketing example: Sales are down. Is it traffic, conversion, price, competition, product availability, seasonality or something else?
3. /whatchanged
Compare what is happening now with the period before the problem appeared. Look for changes in pricing, customers, traffic, offers, competitors, staffing, products, channels or external conditions.
Marketing example: Instead of immediately blaming the economy for slower sales, ask what actually changed before the slowdown started.
4. /assumptioncheck
Identify the assumptions behind the current marketing conclusion and separate the ones supported by evidence from the ones everyone has simply started repeating.
Marketing example: Before deciding customers want more discounts, check whether you have any evidence that price is actually the problem.
5. /missinginfo
Tell me what information you would need before answering this well. Do not quietly fill missing gaps with assumptions.
Marketing example: Before diagnosing declining website leads, ask what traffic, conversion, source, audience and website-change information would help.
6. /evidencefirst
Review what we actually know before suggesting what to do. Separate data, customer feedback, observations, assumptions and guesses.
Marketing example: Look at slow-day sales patterns before asking ChatGPT for ten Tuesday promotions.
7. /hypothesismode
Generate several plausible explanations that can be investigated rather than settling on one answer too quickly. For each one, explain what evidence would support or weaken it.
Marketing example: If people are entering a store but buying very little, investigate pricing, merchandising, inventory, customer intent and staff recommendations. I go deeper into that in People Are Coming Into the Store, but They Aren’t Buying Much.
8. /truthgate
Separate what is known, what is inferred, what is assumed and what remains unknown. This is useful when an answer sounds more certain than the available evidence deserves.
Marketing example: Use it when prospects disappear after seeing a quote and everyone assumes the price must be too high.
9. /correlationcheck
Challenge claims that one marketing activity caused an outcome simply because the two happened at the same time.
Marketing example: Sales increased after a social campaign, but did the campaign cause the increase or was something else happening at the same time?
10. /premortem
Assume the marketing plan failed and work backward. What most likely went wrong, and what could you do now to reduce those risks?
Marketing example: Run a pre-mortem before investing in a new loyalty program.
11. /redteam
Attack the plan. Look for weaknesses that someone who likes the idea may be overlooking.
Marketing example: Red-team a new subscription offer before spending money building the checkout flow.
12. /blueteam
Take the weaknesses already identified and strengthen the plan around them.
Marketing example: After /redteam uncovers problems with an event offer, use /blueteam to redesign the parts that are fixable.
13. /socratic
Ask me questions before giving an answer. Keep asking until you have enough context to give useful advice.
Marketing example: Before recommending a new acquisition channel, have ChatGPT ask how customers currently find the business, what has been tried and what a qualified lead is worth.
14. /constraints
Work within the limits I give you instead of recommending things that are not realistic.
Marketing example: Find ways to increase repeat visits with no additional ad budget and no discounting.
15. /diagnosefirst
Do not recommend tactics yet. Help me understand the problem, missing information and likely causes first.
Marketing example: Use it before asking for another campaign because last month’s campaign underperformed.
I Need to Understand My Customers Better
Customer reviews, questions, calls, emails and purchase behavior often contain more useful marketing information than businesses realize.
16. /customerview
Look at this entirely from the customer’s side. What do they know, see, worry about, need and have to decide?
Marketing example: Review a booking flow from the perspective of someone using the business for the first time.
17. /firsttimevisitor
Assume the reader knows nothing about the company. Find anything that depends on insider knowledge.
Marketing example: Review a homepage and flag terminology that makes sense to employees but not to prospects.
18. /reviewminer
Analyze customer reviews and group them into recurring themes. Separate repeated patterns from unusual one-off comments.
Marketing example: If the same complaint keeps appearing, use the analysis alongside Your Reviews Keep Mentioning the Same Complaint.
19. /languagefinder
Pull out the words and phrases customers use themselves when describing problems, priorities, frustrations and benefits.
Marketing example: Analyze 100 reviews and find the phrases customers repeatedly use to describe why they chose the business.
20. /valuefinder
Identify what customers seem to value most and compare that with what the business emphasizes in its marketing.
Marketing example: If customers constantly praise convenience but the website barely mentions it, read Customers Praise Something You Barely Mention in Your Marketing.
21. /painminer
Group customer frustrations into themes and distinguish between minor annoyances, purchase barriers and problems likely to drive customers away.
Marketing example: Analyze support tickets before deciding which customer-experience problem deserves priority.
22. /questionminer
Find the questions customers keep asking and what those questions may reveal about missing information.
Marketing example: If the same questions appear before purchase, use Customers Keep Asking the Same Questions Before They Buy to decide what should be answered earlier.
23. /objectionhunter
Identify the reasons a qualified customer might hesitate, delay or say no. Understand the objection before trying to overcome it.
Marketing example: Analyze a high-priced service offer and identify what a serious prospect might still need to know.
24. /segmentthis
Find meaningful groups within the customer base based on behavior, needs or value rather than creating segments simply because demographic information is available.
Marketing example: Separate one-time customers, repeat buyers, lapsed customers, strong referrers and high-margin customers.
25. /bestcustomer
Identify what makes a customer valuable beyond total revenue. Include margin, repeat purchase, referrals, retention, support burden and fit.
Marketing example: Use it before deciding which customers should receive a loyalty offer.
26. /retentionlens
Look at what happens after the first purchase. Find the communication, follow-up, second-purchase opportunities and places where the relationship appears to stop.
Marketing example: This fits the problem in People Buy From You Once, but You Don’t Know What Happens Next.
27. /referralgap
Investigate why happy customers are not recommending the business. Look at timing, ease, awareness, incentives and whether customers know who would make a good referral.
Marketing example: Use it when reviews are excellent but referrals are rare.
28. /personaconstraints
Give a hypothetical customer real constraints instead of a fictional biography. Include budget, time, knowledge, priorities and barriers.
Marketing example: Evaluate a $99-a-month marketing platform from the point of view of a business owner with no marketing employee and only two hours a week to manage it.
29. /syntheticgroup
Explore possible reactions from several clearly defined hypothetical customer types. Use the results to generate questions, not as a substitute for real customer research.
Marketing example: Stress-test three messages before conducting actual customer interviews.
30. /momentofdoubt
Identify where in the customer journey someone is most likely to hesitate. Then ask what information, proof or reassurance is missing at that moment.
Marketing example: Review the point where a prospect encounters a deposit requirement during checkout.
I Need Better Marketing Content or Messaging
These codes are useful when I already know I need content but want it to be clearer, more useful and less generic.
31. /audience
Define exactly who this is for before creating anything. Include what the audience already knows, what they care about and what they are trying to decide.
Marketing example: Write an email for first-time home sellers rather than “homeowners.”
32. /voicematch
Study examples of writing I actually like before creating something new. Identify what makes the voice recognizable and reproduce the patterns without copying wording.
Marketing example: Give ChatGPT three newsletters you wrote yourself before asking for a new draft.
33. /aisounding
Find the sentences and structures that make this sound AI-written. Do not rewrite everything yet; tell me what feels artificial and why.
Marketing example: Flag repetitive conclusions, generic transitions, over-polished phrasing and constant three-part lists.
34. /clarity
Make this easier to understand without oversimplifying it. Find buried points, vague language and sentences trying to do too much.
Marketing example: Simplify a consulting page so a prospect can tell what the service actually does within a few seconds.
35. /cutfluff
Remove anything that does not help the reader understand, decide or act.
Marketing example: Cut a 900-word event email down to what attendees actually need.
36. /plainenglish
Replace jargon and internal terminology with language the customer would naturally understand.
Marketing example: Explain marketing attribution to a local business owner without assuming analytics knowledge.
37. /specificity
Find vague claims and tell me what concrete detail, example, number or proof would make each one meaningful.
Marketing example: Replace “exceptional customer service” with evidence of what the business actually does.
38. /proofcheck
Flag every claim that needs evidence or support.
Marketing example: Review a page using phrases such as “leading,” “trusted,” “best” and “proven” and ask what backs those claims up.
39. /showdontclaim
Replace broad marketing claims with evidence, examples, process details or demonstrations.
Marketing example: Instead of saying the company is responsive, explain how quickly customers can actually expect a response.
40. /onepromise
Find the single main promise this marketing asset should make and remove competing promises that weaken it.
Marketing example: Tighten a workshop page promising more customers, better content, automation, analytics and productivity all at once.
41. /onecta
Choose the one action that matters most and identify secondary calls to action that may be distracting from it.
Marketing example: Decide whether a page should primarily ask someone to book, call, download or subscribe.
42. /hooklab
Create fundamentally different openings rather than rewriting the same opening 20 ways.
Marketing example: Approach an article through lost money, customer frustration, surprising data, a story or a common misconception.
43. /anglefinder
Find different ways to approach the same marketing subject based on different customer concerns or business outcomes.
Marketing example: Retention could be approached through profitability, loyalty, convenience, customer experience or wasted acquisition costs.
44. /contentgap
Compare what I have already published with what customers still need help understanding.
Marketing example: Review a service-business blog and identify useful customer questions that do not have a good answer yet.
45. /dontwriteyet
Do not start drafting. Work through the audience, objective, evidence, message and structure first.
Marketing example: Use it before asking ChatGPT to write an article about a marketing problem you have not fully understood.
46. /outlinefirst
Build the logic and structure before writing the body.
Marketing example: Outline a workshop or long-form article before drafting 2,000 words.
47. /repurpose
Find several useful standalone pieces inside work I already created.
Marketing example: Turn a 60-minute workshop into a blog article, newsletter topic, short video, LinkedIn post, checklist and FAQ.
48. /tone
Describe how I want the writing to feel and what I do not want it to sound like.
Marketing example: “Knowledgeable and conversational, but not corporate, overly polished or motivational.”
I Need More Leads, Sales or Conversions
These codes help with the point where marketing is supposed to turn into an inquiry, booking, purchase or repeat customer.
49. /offerdoctor
Diagnose the offer before rewriting the advertising around it. Look at audience fit, value, proof, price, packaging, risk and clarity.
Marketing example: Use it when people show interest but few people actually buy.
50. /pricediagnosis
Do not assume price is the problem. Investigate perceived value, proof, customer fit, packaging, timing and how the price is being presented.
Marketing example: This is the thinking behind People Ask for the Price and Then Go Silent.
51. /discountcheck
Look beyond the short-term sales increase. Examine margin, customer quality, purchase timing, repeat behavior and whether customers are becoming dependent on deals.
Marketing example: Use it with Your Promotions Bring People In—but They Only Buy When There’s a Deal.
52. /profitlens
Evaluate marketing based on profit and customer value, not only revenue or number of sales.
Marketing example: A best-selling product may not be the most profitable product once labor, margin and repeat purchase are considered.
53. /leadquality
Ask whether getting more leads would actually solve the problem. Look at fit, intent, source, qualification and conversion.
Marketing example: This is useful when website traffic is growing but the inquiries are poor.
54. /leadresponse
Map what happens between a new inquiry and the first useful response. Find delays, missing information and unclear ownership.
Marketing example: Use it with It Takes Too Long to Respond to New Leads.
55. /funnelwalk
Walk through the customer journey one step at a time and describe what the customer sees, knows, does and needs.
Marketing example: Follow ad click → landing page → form → confirmation → sales call.
56. /dropoff
Find exactly where customers stop progressing and compare what happens before and after that point.
Marketing example: This works with Customers Start Booking, Then Disappear Before Finishing.
57. /frictionfinder
Look for unnecessary effort, repeated information, confusing forms, account creation, waits or extra steps.
Marketing example: Audit a booking process for things customers have to do that do not help them complete the purchase.
58. /confusioncheck
Find places where the customer has to stop and figure out what something means or which option they should choose.
Marketing example: Review a pricing page with too many package names and unclear differences.
59. /choiceoverload
Look for situations where too many products, packages or services are making the decision harder.
Marketing example: Use it alongside Your Business Offers So Many Services That Customers Don’t Know What to Choose.
60. /trustcheck
Identify what a skeptical buyer would need to believe before taking the next step.
Marketing example: Review a high-priced service page for missing reviews, credentials, process information or proof.
61. /journeygap
Find missing communication between stages of the customer experience.
Marketing example: Check what happens between booking and arriving for the appointment, or between purchase and the customer’s first use.
62. /messagestress
Find ways a marketing promise could be misunderstood, exaggerated or interpreted differently than the business intended.
Marketing example: Review what customers might reasonably assume from “results in 30 days.”
I Have Marketing Data but Don’t Know What to Do With It
This is where ChatGPT can become much more useful than a report summarizer.
63. /sowhat
For every finding, explain why it matters and what decision it could affect.
Marketing example: “CTR increased 18%” is an observation. /sowhat asks what that changes.
64. /decisionmode
Turn information into a decision. Show the choices, evidence, tradeoffs, unknowns and what could change the recommendation.
Marketing example: Decide whether to increase, reduce or hold spending on a channel.
65. /measurementcheck
Ask whether the metric being reported actually answers the business question.
Marketing example: Social engagement may not be the right success metric if the real objective is qualified appointments.
66. /8020
Look for the relatively small number of customers, products, channels or activities creating a disproportionate share of the result.
Marketing example: Determine whether a small group of customers generates most of the company’s revenue.
67. /effortvalue
Compare the amount of time and resources an activity requires with what it appears to contribute.
Marketing example: Find social tasks that take hours every week but generate almost no useful business outcome.
68. /priorityfilter
Apply specific criteria to a long list of possible actions.
Marketing example: Score 20 marketing ideas based on impact, evidence, effort, cost and urgency.
69. /keepfixstoptest
Sort current marketing into four groups: what should continue, what needs improvement, what should stop and what deserves a test.
Marketing example: Use it during a quarterly review instead of leaving with 30 unrelated recommendations.
70. /decisiontree
Map the action that makes sense under different conditions.
Marketing example: If traffic is low, do X. If traffic is healthy but conversion is weak, do Y. If acquisition is strong but repeat business is poor, do Z.
71. /trendcheck
Ask whether a change appears persistent enough to act on or could simply be normal variation.
Marketing example: Determine whether one unusually weak week deserves a strategy change.
72. /segmentdata
Break the overall average into meaningful groups and look for differences.
Marketing example: Compare performance by audience, location, product, device, source or customer type.
73. /outliercheck
Find unusually high or low observations and determine whether they are errors, rare events or something worth investigating.
Marketing example: One location has double the conversion rate of every other location. Find out why before treating the average as normal.
74. /datatodecision
End the analysis with the decisions the data actually supports rather than another summary.
Marketing example: This connects directly with You Have More Data Than Ever and Still Don’t Know What to Do Next.
I Need to Plan, Test or Learn From My Marketing
Running more marketing activity does not automatically mean the business is learning anything from it.
75. /learningagenda
Turn marketing activity into a question you are deliberately trying to answer.
Marketing example: Before testing a new offer, decide whether you are trying to learn about price sensitivity, message relevance, audience fit or something else.
76. /hypothesis
Turn an assumption into something you can actually test. Include what you expect to happen and why.
Marketing example: “We believe adding customer proof above the form will increase completed inquiries because prospects currently lack reassurance.”
77. /whatdidwelearn
Compare what actually happened with what you expected. Pay particular attention to surprises.
Marketing example: After a promotion ends, identify which assumptions were supported and which were not.
78. /nexttest
Use what you learned to decide what experiment should come next.
Marketing example: If changing the message improved conversion, decide whether the next question should involve audience, proof, offer or traffic quality.
79. /controlvariables
Identify what needs to stay the same during a test so the result can be interpreted.
Marketing example: Avoid changing the audience, creative, landing page and offer at the same time if you want to learn what caused the result.
80. /successcriteria
Define success before the marketing activity begins rather than deciding afterward whether the result “felt good.”
Marketing example: Set the minimum improvement that would justify rolling out a test.
81. /failurecriteria
Decide what result would cause you to stop, revise or reject the idea.
Marketing example: Define the point at which a paid acquisition test no longer makes financial sense.
82. /90daymode
Turn what you have learned into a practical plan for the next 90 days.
Marketing example: Decide what to protect, improve, test, reduce and measure using AI to Build Your Next 90-Day Marketing Growth Plan.
83. /scenariomode
Plan for several plausible situations rather than relying on one forecast.
Marketing example: What changes if leads fall 20%, stay flat or increase sharply?
84. /futurecast
Explore future scenarios and the assumptions behind them without pretending ChatGPT can predict exactly what will happen.
Marketing example: Explore how a local business might adjust marketing if demand weakens, stays stable or grows.
The thinking in this section is closely related to You Keep Trying New Marketing Ideas, but Don’t Know What You’re Learning.
I Need to Understand Competitors or Positioning
Competitor research is useful when it helps you understand the market. It is much less useful when it becomes “they are doing this, so we should too.”
85. /competitorcontext
Compare competitors across offers, messaging, proof, channels, reviews, customer experience and positioning.
Marketing example: Look at what several competitors emphasize rather than copying one ad you happened to notice.
86. /copyordifferentiate
Ask whether a competitor tactic actually fits your audience, customer journey, resources and business model.
Marketing example: Before joining TikTok because a competitor is there, determine whether your own customers use it during the buying process.
87. /positioningcheck
Identify what the customer should remember about the business and whether current marketing consistently supports that idea.
Marketing example: Compare the homepage, ads, reviews and sales deck and see whether they communicate the same reason to choose the business.
88. /categoryexpectations
Identify what customers already expect from businesses in the category before deciding how to stand out.
Marketing example: A dentist still has to meet basic expectations around trust, cleanliness and professionalism before trying to differentiate on something else.
89. /whitespace
Look for customer needs, questions, segments or marketing angles that competitors appear to be under-serving.
Marketing example: Find useful content questions none of the local competitors answer well.
90. /channelcheck
Investigate whether a competitor being visible on a channel means the channel actually matters to your customer journey.
Marketing example: This is the thinking behind A Competitor Keeps Showing Up in Places You Don’t.
I Need to Stop Recreating the Same Marketing Work
This is where cheat codes start becoming something larger: reusable systems.
91. /repeatable
Ask whether this task happens often enough to document instead of starting over every time.
Marketing example: If someone writes essentially the same follow-up email five times a week, the process probably deserves to be captured.
92. /sopmode
Turn an established process into a reusable procedure with inputs, steps, decisions, exceptions, outputs and review points.
Marketing example: Convert a recorded walkthrough of publishing the monthly newsletter into an SOP.
93. /handoffready
Check whether another person could actually perform the process using the documentation you created.
Marketing example: Find the undocumented knowledge that still lives only in the marketing manager’s head.
94. /humancheck
Identify the points where a person still needs to review, approve, interpret or make a judgment call.
Marketing example: AI can categorize customer inquiries, but a complaint involving a refund may still require human review.
95. /leverage
Look for ways the work can be reused, delegated, templated, automated, repackaged or turned into an asset.
Marketing example: Turn one workshop into an article, resource, paid working session and team-training module.
96. /sourceoftruth
Tell ChatGPT which document, file or dataset should govern the work when information conflicts.
Marketing example: Use the approved pricing document as the source of truth when drafting sales materials.
97. /contextstack
Gather the business, audience, objective, history, examples, constraints and source material ChatGPT should know before it starts.
Marketing example: Give ChatGPT the brand guide, customer reviews, past campaigns and current offer before asking it to plan a new campaign.
This section connects directly with Your Team Keeps Recreating the Same Work From Scratch and 11 AI Skill Files Every Small Business Should Create.
I Need ChatGPT to Be More Reliable
These are simple guardrails, but I use versions of them often.
98. /noinventing
Do not fill missing information with made-up details. Tell me what is missing instead.
Marketing example: Draft a case study from the information provided without inventing customer quotes or results.
99. /factcheck
Identify every claim, number, name, statistic, date or external statement that should be verified before publishing.
Marketing example: Run it on an article before using statistics about customer behavior or industry trends.
100. /datecheck
Find anything likely to have changed since the information was created.
Marketing example: Use it before republishing an article about ChatGPT, software pricing, social-platform features, regulations or other fast-changing subjects.
Stack the Cheat Codes
You do not have to use these one at a time. Some of the most useful requests combine several codes because the problem requires more than one kind of thinking.
If a landing page is not converting, I might use:
/firsttimevisitor /confusioncheck /objectionhunter /trustcheck /onecta
Review this landing page without rewriting it yet. Show me where a potential customer could become confused, hesitate or decide not to continue.
Before launching a promotion:
/truthmode /discountcheck /profitlens /premortem
Here is the promotion, pricing, margin and normal customer behavior. Tell me what could make this a bad idea and what I should verify before we launch it.
For customer reviews:
/reviewminer /languagefinder /valuefinder /painminer
Analyze these reviews. Separate recurring patterns from isolated comments and show me what customers repeatedly value, dislike and talk about in their own words.
For a marketing report:
/sowhat /measurementcheck /datatodecision /keepfixstoptest
Don’t give me another summary. Tell me which findings matter, which metrics actually answer the business question and what the data supports doing next.
That is where the codes become more useful to me. I am not simply telling ChatGPT what output to produce. I am telling it how I want it to approach the problem.
Make Your Own Cheat Codes
Start with the instructions you already repeat. If you constantly ask ChatGPT not to agree with you too quickly, create your version of /truthmode. If you regularly analyze reviews, define /reviewminer once and make it work the way you want. If every marketing report ends with the same questions, build those questions into /datatodecision.
A short instruction may stay a cheat code. Once the process becomes more complicated, it may deserve a Skill File with examples, standards, decision rules and more detailed instructions. That is the next step I describe in 11 AI Skill Files Every Small Business Should Create.
The names themselves do not matter very much. /truthmode works for me because I know exactly what I mean by it. You could call yours something completely different.
What matters is noticing the ways you repeatedly want AI to help you think, research, analyze and create, then capturing those instructions instead of rebuilding them from scratch every time.