Ad Analysis Prompts: 25 That Break Down Winning Ads

Which prompts extract an ad's hook and angle?

The prompts that extract hook and angle force the model to name a category and quote the exact line, not describe the ad in vague terms. A request like 'analyze this ad's hook' returns marketing-speak; a request that demands a hook type and a quoted sentence returns something you can actually act on.

Angle extraction works the same way. Ask the model to list every distinct angle (mechanism, identity, urgency, social proof) and cite the line carrying each one, and you get a map of the ad's persuasion structure instead of a single adjective. That map matters most once you have already filtered for offers worth the teardown time, which is the filtering process covered in how to identify winning ads: most ads sitting in a spy-tool feed were never winners to begin with.

• Identify the hook used in the first 3 seconds of this ad script. Name the hook type (curiosity, shock, direct-benefit, pattern interrupt, or question) and explain why it stops the scroll.
• List every distinct angle this ad makes toward the same offer (mechanism, identity, urgency, social proof, or authority). For each angle, quote the exact line that carries it.
• Rewrite this hook in 5 alternate versions, using a different angle each time, while keeping the same offer and core claim.
• What emotional state is this ad trying to create in the first 10 seconds? Quote the specific words that build it.
• Compare this ad's hook structure to a known direct-response pattern (problem-agitate-solve, before-after-bridge, or the four Ps) and name which one it follows, if any.

How do you prompt a full VSL teardown?

A full VSL teardown starts by forcing the model to segment the transcript into structural beats before it interprets anything. Ask for hook, problem, mechanism, proof, offer, and close as separate blocks; most transcripts between 8 and 20 minutes map onto that structure cleanly, though some compress problem and mechanism into a single stretch.

Most media buyers still trust a fast human read over a model's pass at the same transcript, and that instinct is usually backwards for claim extraction specifically. A person skimming a 15-minute script for the third time that week will miss a repeated claim buried at minute 9; the model reads every sentence with the same attention on read one as read ten, which is why a prompted claim-stack extraction routinely surfaces more distinct claims than a first-pass human note-take does.

Two extractions matter more than the rest: the claim stack and the offer stack. The claim stack is every specific claim about the product, in the order made; the offer stack is price anchor, bonuses, guarantee, urgency mechanism, and payment terms. Pull both separately, because conflating them hides whether a VSL is selling on product claims or on deal structure.

• Break this VSL transcript into its structural beats: hook, problem, mechanism, proof, offer, and close. Timestamp each beat if timestamps are present in the source.
• Extract the claim stack from this VSL: every specific claim made about the product, in the order made, with the exact quote for each one.
• Identify the offer stack: price anchor, bonuses, guarantee, urgency or scarcity mechanism, and payment terms. Quote each element as written.
• What proof elements does this VSL use (testimonials, before-after visuals, screenshots, third-party data, or an expert cameo)? List each with its timestamp or paragraph location.
• Where does this VSL first ask for the sale, and how many times does it re-pitch before the final close?

What context does the model need to avoid guessing?

The model needs the platform, an approximate run window, and the target audience, or it will guess and state the guess as settled fact. Claude and ChatGPT will both confidently assign a hook type or proof element to a transcript with no date, price, or platform attached, so the burden sits on you to supply that scaffolding before the first prompt runs.

Platform changes what 'winning' can even mean for the same ad. A script built as a YouTube pre-roll behaves differently from one built for a Facebook feed, and a static view count without a timeframe tells the model nothing about momentum. If the source is a YouTube ad, feed it the trajectory described in how to find winning youtube ads rather than a single snapshot number.

Also instruct the model explicitly to separate what the ad claims from what the product does. Without that instruction, a teardown will quietly convert 'this VSL claims a 90-day mechanism' into 'the product works via a 90-day mechanism,' which is a claim the model has no way to verify and you have no basis to repeat.

• Treat this as a [platform] ad that has been running since approximately [date/window], targeting [audience]. Do not assume it converted; flag any performance claim as unverified unless I provide spend data.
• Here is the full transcript, not a summary. Quote your evidence for every claim you make about structure, tone, or intent.
• This is a [category] product priced at $[X] with [guarantee terms]. Analyze the claim stack specifically against that price point.
• Distinguish throughout your answer between what the ad states as fact and what the ad claims the product does; label every product claim with 'the ad claims' rather than stating it as true.

Which prompts compare two competitor ads?

The prompts that compare two competitor ads work best as a forced side-by-side, not two summaries pasted next to each other. Ask the model to build one table across both transcripts, covering hook type, primary angle, proof elements, and close mechanism, because a shared table exposes divergence that two separate paragraphs will bury.

This matters most in categories with a dense competitive set, where a dozen offers run near-identical claims wrapped in different framing. Supplement funnels are the clearest case: price points, guarantee terms, and bonus stacks vary enough between competitors that a structured comparison, like the category breakdown in supplement competitor analysis, surfaces more than reading either ad alone would.

• Compare these two ad transcripts for the same product category. List where they use the same angle, where they diverge, and which one commits harder to a single claim.
• Which of these two ads has a tighter claim stack, fewer and more specific claims, and which relies on volume of vague ones?
• Build a side-by-side table: hook type, primary angle, proof elements, and close mechanism for both ads.
• If both ads target the same audience, which offer math, price, bonuses, guarantee, gives the buyer a clearer reason to act now?
• Where do these two ads contradict each other on mechanism or claim, and what does that suggest about which one is still testing versus which one is scaling?

How do you turn analysis into a creative brief?

You turn analysis into a creative brief by asking the model to compress the teardown into constraints a copywriter can execute against, not another paragraph of description. A usable brief needs a hook direction, one primary angle, a claim stack capped at 3 to 5 items, and an offer structure; past that length it stops functioning as a brief and becomes a second teardown.

Where this gets legally and strategically important is the line between modeling an ad's structure and copying its specific language. A brief that says 'match this angle and proof pattern' is modeling; one that says 'use these exact lines' is copying, and the distinction laid out in modeling vs copying ads determines whether the output is a defensible creative or a liability waiting on a cease-and-desist letter.

A tight brief is also the exact input format an AI drafting workflow needs: hook direction, one angle, a capped claim stack, and offer structure, fed in as constraints rather than a wall of transcript. Loose inputs produce loose drafts regardless of which model writes them.

• Turn this ad teardown into a creative brief: hook direction, angle, claim stack (3-5 items maximum), proof elements to source, and offer structure.
• Generate 3 new hook variations for a brief that models this ad's angle without reusing its specific wording or claims.
• List every claim in this ad that would need substantiation before I could make it, and flag which ones read as opinion versus fact.
• Draft a one-paragraph creative direction a copywriter could work from directly, noting which structural beats to keep and which to change.
• What is the single riskiest claim in this ad from a compliance standpoint, and how would you soften it while keeping the underlying angle intact?

When do prompts fail without spend data?

Prompts fail without spend data the moment you ask the model to judge whether an ad works, because structure and performance are different questions and only spend data answers the second one. A model can tell you an ad has a well-built hook by pattern-matching against known-effective structures, but it cannot tell you that hook held attention or converted, since it has no access to impressions, click-through rate, or spend.

This is where teardown work most often overstates its own conclusions. An ad's structure can look identical to a proven winner and still be losing money in market, which is the gap covered in how to find winning ads: spend persistence and creative refresh cadence sit outside the ad copy itself, and no prompt can infer either one from a transcript alone.

The honest move is to ask the model to mark its own limits. A prompt like 'tell me what you can and cannot conclude about this ad's performance from the transcript alone, and be explicit about which parts of your answer are inference versus observation' returns a shorter, more useful answer than one that lets the model fill the gap with confident-sounding guesswork.

Quick decision checklist

Use this page as a decision aid, not a generic blog post. The practical question is whether the reader needs faster evidence about what is already working in VSL-driven direct response, especially across nutra, supplements, GLP-1, weight loss, blood sugar, and adjacent high-intent health markets.

Daily Intel Service is most relevant when the next decision depends on active market examples: which hook to test, which claim style is risky, which funnel structure is common, which language market is moving, and whether a competitor's creative is likely early, scaling, or already saturated.

• Start with the TL;DR if you need the direct answer.
• Use the table to compare trade-offs quickly.
• Use the FAQ for answer-engine-ready summaries.
• Use the CTA when the decision requires live VSL and ad examples instead of theory.

Daily Intel's coverage advantage

Daily Intel Service is positioned around category-leading variety and actionability: one of the broadest direct-response catalogs of VSLs and ad creatives across blackhat, greyhat, and whitehat advertising patterns, with enough context to understand what the advertiser is doing beyond the visible creative. The practical difference is that members are not just seeing a screenshot; they are seeing the VSL, the ad, the funnel path, the transcript, the UTM context, and the research notes that turn the asset into a decision.

This matters because direct-response affiliates do not operate in one clean category. A weight-loss campaign may use a whitehat compliance ad, a greyhat pre-lander, a more aggressive VSL, and a checkout path designed around upsells and recovery. A useful intelligence platform needs to capture that spectrum instead of pretending every winning campaign looks like a public brand ad.

Blackhat, whitehat, and multilingual signal coverage

Daily Intel tracks patterns across both blackhat-style and whitehat-style campaigns so operators can understand the market without blindly copying risk. Whitehat examples help with durability and compliance review; blackhat and greyhat examples reveal pressure points, hooks, mechanisms, and funnel structures that may be driving spend but require careful adaptation before use.

The catalog is also built for global operators, with VSL and ad references spanning 14+ languages and different local idioms. That is a key advantage for Brazilian, LATAM, European, MENA, Indian, and non-native English affiliates who need to see how the same market desire is translated across cultures instead of only studying US English ads.

How to use the intelligence responsibly

The goal is modeling, not copying. Use Daily Intel to understand structure: hook, mechanism, proof, claim intensity, funnel depth, offer economics, and saturation stage. Then build original creative, review claims, and adapt the angle to the traffic source, country, language, and compliance requirements of the campaign.

A strong workflow compares multiple examples before acting. If the same mechanism appears across several languages, several advertisers, and several funnel variants, it may be a durable market signal. If the example appears only once or depends on an aggressive claim, treat it as a research clue rather than a campaign template.

• Model structure, not protected creative assets.
• Separate whitehat durability from blackhat persuasion pressure.
• Compare US English examples against LATAM, European, and other language variants.
• Use transcripts and funnel notes to build original briefs.
• Keep compliance review separate from market research.

Methodology and source context

Daily Intel pages are written from a research workflow that reviews active VSLs, Meta ad creatives, transcripts, UTMs, funnel paths, checkout steps, upsells, recovery sequences, and compliance-sensitive claim patterns. The goal is to explain observable market behavior, not to provide legal, medical, or platform policy advice.

For educational pages, the supporting references should help readers verify search, crawlability, and public ad research context, especially Google helpful content guidance, Google SEO link best practices, and Meta Ad Library. Daily Intel then adds the direct-response interpretation layer so the page explains what the signal means for actual affiliate research decisions.

For deeper evaluation, continue through State of ad spy tools in 2026, First-Party Data for Affiliates: The 2026 Playbook, Meta Event Match Quality: How to Raise EMQ Fast (2026), TikTok Events API for Affiliates: S2S Setup for 2026, n8n Ad Spy Workflow: Automate Competitor Monitoring, and What is a VSL?. These related Daily Intel pages connect this topic to the relevant methodology, pricing, trust context, comparison path, or niche workflow.