Best PromptsMeta Ad Hooks
Generate paid social hooks across distinct psychological angles plus the creative brief and test plan needed to learn something from them.
Create a hook and creative test plan for [PRODUCT] on [Meta / TikTok / both]. Audience: [WHO, AND THE MOMENT THEY ARE IN WHEN THEY SEE THIS] The problem in their words: [HOW THEY DESCRIBE IT, NOT HOW WE DO] What we offer and the one thing that makes it different: [OFFER + DIFFERENTIATOR] Proof we can show: [RESULTS, REVIEWS, DEMOS, BEFORE AND AFTER] Offer: [PRICE, TRIAL, DISCOUNT, GUARANTEE] What has already worked or failed: [PAST WINNERS AND LOSERS, or "no history"] Produce: 1. **12 hooks**, three each across four distinct angles, so the test measures the angle and not the wording: - Problem callout: name the pain in their words in the first three seconds - Result proof: lead with the specific outcome or number - Pattern interrupt: a contrarian or unexpected opening claim - Identity: speak to who they are or want to be Each hook: the spoken or on-screen line (12 words or fewer), plus the visual in one line. 2. **3 full concepts**: for the three strongest hooks, a shot-by-shot brief with beat timings for a 15 to 30 second video, on-screen text per beat, and the call to action. 3. **Primary text and headline** for each concept, written for the feed and not for a landing page. 4. **Test plan**: what varies between cells and what is held constant, how many creatives per ad set, the metric that decides a winner at each stage (hold rate first, then cost per action), and roughly how much spend or how many impressions before the result means anything. Say plainly which comparisons this test cannot settle. Rules: every claim must come from the proof I listed. No invented statistics, no invented testimonials, no implied medical, financial, or income guarantees. Flag any hook likely to trip platform policy and give a compliant alternative.
How to use
Testing twelve rewordings of one angle teaches you nothing, which is why the hooks are grouped into four genuinely different psychological angles: you want to learn which angle the audience responds to, then generate variations inside the winner. The 'what this test cannot settle' line keeps you from reading a creative result as an audience or offer result. Judge hooks on three-second hold rate before cost per action, because a hook that nobody watches cannot convert.
More marketing prompts
Audit my visibility in AI answer engines and give me a fix list. Business: [WHAT YOU SELL, TO WHOM] Category terms people use: [e.g. "invoicing software for freelancers"] Competitors that get named instead of us: [LIST, or "unknown"] Pages to evaluate: """ [PASTE THE FULL TEXT OF 1-5 KEY PAGES: homepage, a product page, your best guide]
AI Search Visibility Audit
Audit whether ChatGPT, Perplexity, and AI Overviews can find, quote, and recommend you, then get a prioritized plan to fix what they cannot.
Audit my existing content and tell me what to refresh, merge, or retire. Page data (title, URL, target keyword, traffic now, traffic at peak, publish date, last update, conversions if known): """ [PASTE AS A TABLE OR LIST] """ Business goal this content serves: [e.g. "trial signups from mid-funnel guides"] For each page, assign one act
Content Refresh Audit
Decide which existing pages to update, merge, or delete based on decay and intent drift, with a specific change list for each keeper.
Write responsive search ad copy for this ad group. Product or service: [WHAT IT IS, WHO IT IS FOR] Ad group theme and keywords: [THE TIGHT KEYWORD SET THIS AD GROUP TARGETS] What the searcher is trying to do: [e.g. "compare options", "buy now", "fix an urgent problem"] Landing page: [WHAT THE PAGE ACTUALLY OFFERS AND WHAT THE PRIMARY ACT
Google Ads Copy
Write responsive search ad headlines and descriptions that match query intent, with pinning guidance and a matching landing page check.