Prompt library
Free AI prompts for SEO and marketing
74 ChatGPT prompts, free, no account and nothing to download. Each one is written for a job somebody actually does: cluster a keyword export, read a crawl, write a meta description that survives truncation, work out why ChatGPT names a competitor instead of you. Copy it, run it, then check the result against your live site with the free tool each one points at.
- Prompts
- 74
- Categories
- 7
- Cost
- Free
Newest · Content and briefsTurn a goal into a reusable prompt, with its variables and a test planA prompt generator for ChatGPT or Claude. Give it the job you want done and it returns a reusable prompt, its variables, three test inputs and its failure mode.All 74 prompts
Content and briefsDiagnose a prompt that is nearly working, and get a version two you can comparePaste a prompt that gives inconsistent answers. Get a scored diagnosis, the changes with reasons, a rewritten version and the way to prove it is better.
Content and briefsA system prompt for marketing work that survives more than one chatA system prompt for custom instructions or a project, holding your facts, your banned phrasings and your evidence rules so every chat starts from the same place.
Technical SEODraft an llms.txt from your own pages, without inventing anything about youA prompt that turns your page list and real figures into a valid llms.txt, marks every gap instead of filling it, and says which pages do not belong in it.
Content and briefsHow to write a prompt that works the second time you run itA prompting guide built on what changes the output and what only looks like it does, with a six block skeleton and the test that catches a confident wrong answer.
Content and briefsRewrite an AI draft so it stops reading like one, by putting back what the model left outA rewrite prompt that finds the sentences giving your draft away, and fixes them by adding specifics rather than swapping words for synonyms.
Content and briefsTurn a prompt that worked once into a template you can run every weekA meta prompt that takes a one off request that happened to work and rebuilds it as a reusable template with named inputs, a fixed output shape and stated refusals.
Analysis and reportingBrief a deep research run properly, so twenty minutes of work is not spent answering the wrong questionA scoping brief for deep research modes that fixes the question, the sources and the output format before the run starts, and forbids the summary nobody needed.
AI visibilityUse Perplexity to find out who is currently cited for your buying questionsA prompt built for a cited answer engine: it returns the pages currently quoted for your category questions, and the pattern in why those pages and not yours.
AI visibilityReconstruct the query fan out behind a search, and find which parts of it you answerA prompt that predicts the narrower searches an AI answer runs underneath your query, then marks which of them your site currently answers and which it ignores.
AI visibilityAudit how assistants describe your brand, in a form you can repeatA prompt set for auditing what ChatGPT, Claude and Perplexity say about your brand, recorded so the run you do next month is comparable to this one.
Content and briefsWrite a content brief a writer can execute without asking questionsA prompt that turns one target query into a brief with the angle, the outline, the evidence the writer must gather, and the lines they are not allowed to invent.
Keyword researchMine long tail keywords a low authority site can actually winA prompt that expands a seed into long tail queries, then throws out every one whose results page is already owned by a brand you cannot outrank yet.
Keyword researchClassify a keyword list by search intent and the page type it needsA prompt that labels each query by intent, names the page format that satisfies it, and tells you which of your existing pages is the wrong shape for its keyword.
Analysis and reportingTurn audit findings into SEO recommendations a developer will actionA prompt that converts a raw audit export into ordered recommendations, each with the change to make, who makes it, and what to check afterwards.
AI visibilityAEO and GEO: which is which, and a prompt that does the work either wayAnswer engine optimisation and generative engine optimisation are contested labels, not standards. One prompt that does the work behind both.
AI visibilityAnswer engine optimisation in order: a gated plan for one morning, not another explainerA prompt that turns answer engine optimisation into a gated sequence, where each step has a condition that must be true before the next is worth doing.
Content and briefsDraft an article that does not read as generated, by making the model refuse to start without your specificsA drafting prompt that stops and asks rather than writing, until you supply the real example, the real number and the real objection.
AI visibilityStructure an article so an assistant can lift one passage out of itA prompt that restructures a draft you already have into question shaped headings with a standalone answer under each, and invents no figures.
Analysis and reportingRun a competitor analysis in four passes, one input at a timeA four pass competitor analysis prompt that runs one stage at a time, refuses to move on without the inputs, and states what it cannot see.
Keyword researchThe keyword research prompts worth running, and the order they go inThe five stages of keyword research, which prompt belongs to each, and a router prompt that tells you which stage you are actually at.
Content and briefsTurn a brief into a finished blog post, and find out what the brief was missingA prompt that audits your brief before it writes, drafts only from what the brief contains, and lists every place it had to depart from it.
Content and briefsThe blog post prompt sequence, and the gate that decides whether a draft shipsThe five prompts a blog post actually needs, in the order you run them, plus a publish gate that scores a finished draft against eight tests.
Content and briefsCapture a brand voice from writing you already have, and stop calling it trainingA prompt that reads your best existing writing and returns a reusable style instruction block, with a quotation behind every rule it gives you.
Content and briefsWhere a human has to intervene when a model writes your contentThe workflow around the model rather than the prompt inside it: the three points a person has to intervene, and a review prompt that holds a draft.
Keyword researchDecide what one page should target, and make the model quote the page as evidenceA prompt that takes one URL and a shortlist of queries and returns a single verdict on what that page should target, with a sentence from the page as proof.
Analysis and reportingA Claude prompt for SEO that reads four pages at once and has to quote before it concludesAn XML structured prompt that puts your page and three that outrank it in front of Claude at once, and refuses any finding it cannot quote.
Technical SEOAn SEO audit prompt for Claude, built on the two things it will actually doAn audit prompt that pastes the whole page source in at once and requires a verbatim quote behind every finding. The rest goes in what it could not determine.
Analysis and reportingA competitor analysis example, run end to end, with the inputs and the output shownNot a description of a method. A worked run on one query, showing the exact inputs, the gap ledger that came back, and which gaps got struck out.
Analysis and reportingFind the content gap between you and a competitor from two sitemapsPaste two sitemap URL lists and get the subjects your competitor covers and you do not, read from the paths alone, with no guessing at pages nobody pasted.
Content and briefsTurn a content plan into a calendar with a date and an owner on every pageSequences an existing content plan into dated slots, orders the pages by dependency, and refuses to schedule anything whose evidence does not exist yet.
Content and briefsMap one topic into a content plan where every page answers a different questionTurns one topic into a map of pages, gives each page the single question it answers, and refuses to plan a page that would only rephrase another one.
Analysis and reportingDiagnose why a page is not ranking, and make the model admit what it cannot seeA ChatGPT prompt that lists every cause it cannot rule out from what you pasted before naming one, so you get a diagnosis instead of a confident guess.
Local and ecommerceOptimise an Etsy listing for the search engine that actually ranks itAn Etsy SEO prompt that works to the field limits on your listing form, writes titles for a phone screen, and never invents a tag your item does not match.
Content and briefsA Gemini SEO prompt set that has to show where every claim came fromThree chained Google Gemini prompts for SEO, built around grounded search results, so every claim in the brief arrives with the source that produced it.
Analysis and reportingGemini against ChatGPT for SEO work, settled by running the same task in bothA harness you paste unchanged into both assistants. Every claim lands on its own line, tagged sourced, recalled or uncertain, so you judge on your own task.
AI visibilityGet your business named in ChatGPT answers: the whole mechanism, and which parts are actually leversThere is no ranking API and no way to submit a site to ChatGPT. A prompt that scores every lever you have been sold as documented, inferred or undocumented.
AI visibilityHow LLM SEO works: the pipeline, stage by stage, with the guesses labelledWhat happens between a question typed into an assistant and your name in the answer, traced stage by stage, with each stage marked documented or inferred.
Keyword researchDo keyword research from first principles, with a business and no dataKeyword research without a tool: the three sources of truth you already have, and an interview prompt that names no terms until it knows your business.
Analysis and reportingWhere ChatGPT helps with SEO, where it does nothing, and where it costs youA prompt that sorts your own SEO tasks into three buckets: the ones a model does well, the ones where it only saves typing, and the ones where it hurts.
Keyword researchWork out which half of your keyword question a model can actually answerWhat a language model is genuinely good at in keyword research, what it cannot know, and a prompt that scores your request before it tries to answer it.
Content and briefsJudge a blog post against standards a model can apply, and make it admit what it cannot judgeA prompt that critiques a draft you already have against seven criteria, quotes the evidence for every verdict, and refuses to score what it cannot see.
On pageHow to write a good meta description by hand, and a prompt that argues with itThe craft of writing a meta description by hand, plus a prompt that refuses to write one for you and instead interrogates the line you already wrote.
Local and ecommerceRank your local SEO fixes by effort, for the week you actually haveA triage prompt for local SEO. It sorts the work into three effort tiers, states a mechanism instead of a percentage, and ranks nothing you did not describe.
On pageTurn a sitemap export into internal links you can place this afternoonTurns a sitemap export and your page text into named link placements: the anchor text written out, the sentence to put it after, and a reason for each one.
Analysis and reportingThe five ways ChatGPT damages SEO, and the check that catches each one before you publishChatGPT is not bad for SEO. Five specific things it does are, and each is catchable. A verdict first prompt that fails any draft with an unsupported number.
Analysis and reportingIs ChatGPT good for SEO? Make it mark its own output and find outMake the model mark its own SEO work, tagging every claim as checkable, unverifiable, invented or filler, so the answer comes from your output not a vendor.
Content and briefsTake one keyword to an annotated outline, then to a draft, with a stop in betweenA two stage prompt that builds an annotated outline from the results you paste, stops and waits for approval, then drafts only from the outline you approved.
Content and briefsA copywriter prompt set for landing pages that sell and are still findableFour prompts run in order on a commercial page: audit what you can prove, write claims not rewordings, answer the last objection, then make it findable.
Local and ecommerceBuild local keyword sets without ending up with forty identical suburb pagesBuilds service plus modifier plus area keyword sets from your own words, then rules on which sets would produce the same page twice and must be collapsed.
Local and ecommerceBuild a local SEO plan for one location, without producing a doorway pageA local SEO prompt scoped to one location. It decides first whether the page deserves to exist, then splits website work from Google Business Profile work.
On pageMeta description examples, scored line by line against one rubricFour worked meta description examples, before and after, scored against a five line rubric, plus a prompt that scores your own live descriptions the same way.
On pageWrite meta descriptions that still make sense after Google cuts themA prompt that writes meta descriptions as ad copy, puts the whole promise in the first 120 characters, and refuses to claim anything the page does not say.
On pageOptimise an existing page without letting the model rewrite itReturns a short list of anchored edits to a page you have already published, rather than a rewrite. It quotes the exact text to change and alters no fact.
Local and ecommerceWrite product descriptions at catalogue scale without shipping the same page 300 timesA product description prompt built for hundreds at a time. It refuses any product whose attributes cannot separate it, and reports its own overlap.
AI visibilityRanking a website in ChatGPT, honestly: a site level triage of what a crawler actually receivesBefore this is a content question it is a site question. Triages your URLs into cannot be fetched, fetched but empty, and readable but not quotable.
AI visibilityGetting your business named by ChatGPT, when most of the sources are not yoursThere is no rank inside ChatGPT, and for a small or local business the deciding sources are third party. Build a worklist from the citations in a real answer.
Content and briefsRewrite a finished draft for search without flattening the writingAn editing prompt that rewrites only the passages failing one of four named tests, quotes every original beside its replacement, and states the cost.
Technical SEOGenerate JSON-LD schema from the page, with nothing in it that is not on the pageWrites JSON-LD from your actual page content, uses only properties that exist on schema.org, and omits any value it cannot find rather than inventing one.
Keyword researchTriage the keywords you already rank for, before researching a single new oneA prompt that reads a Search Console query export and sorts every query you already rank for into six action buckets, naming no cause the data cannot show.
Keyword researchExpand one seed keyword into a set worth validatingA prompt that expands a single seed keyword into candidate search terms along named axes, marks which came from your own words, and refuses to invent demand.
Analysis and reportingTell a real ranking movement from noise, before anyone blames an algorithm updateA prompt that walks five gates in order and stops at the first one your data cannot pass, and is forbidden from naming a Google update as the cause.
Technical SEOAn SEO audit example: the input, the scored report, and what it refuses to scoreA worked SEO audit end to end. What you paste in, the scored report that comes back, and why the areas with no evidence are excluded from the total.
Technical SEORun a technical SEO audit from a crawl export, without the generic checklistA prompt that audits a real crawl export instead of reciting best practice. It groups findings by cause and names the column behind every one.
On pageWrite the title tag and the meta description as one pair, not two tasksA prompt that writes the title and the description together and forbids them from repeating each other, so the two lines make two different points.
Technical SEOAudit a whole site by page type, so the fix lands on the templateA prompt that audits a site one page type at a time. Home, category, product and article each get their own checks, and every finding names the template.
AI visibilityWhat ChatGPT search retrieves, as opposed to what the model just remembersChatGPT answers from two places: pages fetched live by its search crawler, and what the model absorbed in training. A prompt sorts your questions by path.
Technical SEOWhat a technical SEO audit covers, and the checklist to hold a supplier toA prompt that turns a quote for an SEO audit into a checklist: what each item is, how you can tell it was really done, and what to ask when you cannot tell.
AI visibilityAnswer engine optimisation: what the term means, and which of your pages qualifyAEO is an industry coinage with no defining document, older than generative assistants. A prompt sorts your pages by which answer surfaces they serve.
AI visibilityGenerative engine optimisation: the paper that named it, and what it actually testedGEO comes from a 2023 arXiv paper, not a standards body. What the paper measured, and a prompt that separates the methods it tested from vendor invention.
Keyword researchWhat keyword research is, explained with something you sell rather than running shoesKeyword research defined in one paragraph, then a prompt that teaches it back to you using your own business as the worked example instead of a generic one.
Content and briefsWork out what actually goes in a blog post before you write a word of itA prompt that interviews you with five questions, then returns a contents sheet naming what belongs in each section and which parts only you can supply.
AI visibilityWhy answer engine optimisation matters, and when it honestly does notThe business case for AEO without the invented statistics, when ignoring it is the correct call, and a prompt that writes you a one page decision memo.
Keyword researchTurn a keyword export into topical clusters that map to real pagesA prompt that clusters a keyword export by search intent rather than shared words, names one page per cluster, and flags the pages that would compete.
What makes this different from a ChatGPT prompts list
Most collections of prompts for ChatGPT are long and untested. This one is 74 prompts, and every one of them has been run against the models it names before it went up. Three differences, and they are the reason a shorter list beats a longer one.
- One job each. Not a category of job. A prompt that clusters a keyword export is a different artefact from one that writes a brief, and a prompt trying to do both does neither well.
- Rules that stop a model inventing figures. Every prompt names what the output must contain and what it must refuse to guess, which is the difference between an answer you can use and one you have to check line by line.
- A free tool to verify the result. Each page points at the checker that tests what the model just told you against your live site. A prompt whose output nobody can check is a prompt you have to trust, and trusting output is how bad SEO advice enters a company.
New to this? Start with how to write a prompt that works twice, which covers the parts of prompt writing that still matter and says plainly which famous techniques no longer do.
What the library covers
7 categories, one per kind of job. Pick the one you are doing today.
How do you use one?
Copy it, fill the brackets, run it, then check the result against your live site.
- Copy the whole block. Every prompt is one paste, not a conversation.
- Replace every bracketed placeholder. The page says what goes in each one and how it changes the output.
- Run it in ChatGPT, Claude or Gemini. The page names the models it is written for.
- Read the output against the rules in the prompt. Those rules are what stop a model inventing figures.
- Check the result on your live site with the free tool the page points at.
When does a new one go up?
Most days. One prompt, one real job, and the free tool to check what it gave you.
A new prompt, most days One working prompt for a real SEO or AI visibility job, what to change in it, and a worked example. No sequences, no offers dressed as newsletters.
Unsubscribe in one click. We never pass your address on.