If you want to analyze competitor ads with ChatGPT, Claude, or any other AI assistant, the quality of your output depends almost entirely on the quality of your prompt. A vague instruction like “analyze my competitor’s ads” produces vague results. A structured prompt that tells the AI exactly what data to look at, what framework to apply, and what output format to use produces something you can actually act on.
To find gaps in competitor ad copy specifically, a prompt that works well looks like this: “Here are five ad headlines from [Competitor]. Identify any customer pain points they are not addressing, benefits they are ignoring, and objections they are leaving unanswered. Format your answer as a bullet list of gaps with one recommended angle for each.” That specificity is what separates useful AI analysis from generic summaries.
The 12 prompts below are organized by job-to-be-done. Each one is ready to copy, includes the data field it requires, and shows an example of the kind of output you should expect. Where relevant, the prompts note which data fields you can pull from competitor ad monitoring with Ad Radar, the live ad intelligence feature built into Adsroid.
How to Use This Prompt Library
These prompts are designed to be pasted directly into ChatGPT, Claude, Gemini, or Perplexity. Before you paste, replace the bracketed placeholders with real ad data. The more specific your input, the more useful the output.
You do not need a dedicated AI tool for competitor ad analysis. What you need is a consistent source of live competitor ad data and a structured way to interpret it. The prompts handle the second part. Getting the data is covered in the notes after each prompt.
AI prompts for competitor ad analysis work best when fed with real, current data rather than your assumptions about what competitors are saying. Collect first, analyze second.
Group 1: Messaging Teardown
Messaging teardowns help you understand the narrative a competitor is building, what emotions they are appealing to, and how consistent their copy is across headlines and descriptions.
Prompt 1: Core Message Extraction
Job: Identify what a competitor is fundamentally promising to the customer.
Data needed: Ad headlines and descriptions for one competitor across at least five ads.
The prompt:
“Below are [X] ads from [Competitor Name]. Read all headlines and descriptions carefully. Identify: (1) the single core promise they are making to the customer, (2) the emotional triggers they are using most frequently, (3) the type of customer they appear to be targeting based on the language used. Present each finding in one clear paragraph.”
Example output: The AI identifies that the competitor consistently promises speed of delivery, uses urgency triggers like “today” and “now” in 80% of headlines, and appears to target small business owners based on phrases like “your team” and “without the complexity.”
Ad Radar data field: Ad headlines and descriptions from Search Ads monitoring. In Adsroid, Ad Radar captures the full headline and description text for every detected ad, along with the keyword that triggered it and the detection timestamp.
Prompt 2: Headline Pattern Analysis
Job: Spot structural patterns in how a competitor writes headlines.
Data needed: At least eight to ten headlines from the same competitor.
The prompt:
“Here are [X] headlines from [Competitor Name]’s Google Search Ads. Analyze the structural patterns: Do they lead with numbers, questions, benefits, or brand name? What sentence length do they favor? Do they include the keyword early or late? List the three most common headline formulas with an example of each.”
Example output: The AI finds that 60% of headlines follow a “[Number] + [Noun] + [Outcome]” formula (e.g., “5 Tools That Cut Reporting Time”), that the brand name appears in the final headline slot in most ads, and that keywords appear in position one more often than not.
Ad Radar data field: Ad headlines. Because Ad Radar logs every ad it detects with a timestamp, you can compare headline patterns over time and see whether a competitor has changed their formula.
Group 2: Offer Comparison
Offer analysis tells you what a competitor is using to convert clicks. Free trials, discounts, guarantees, and urgency mechanics are all visible in ad copy if you know what to look for.
Prompt 3: Offer and Incentive Inventory
Job: Catalog every conversion incentive a competitor is advertising.
Data needed: Ad descriptions from multiple competitors running on the same keyword set.
The prompt:
“Below are ad descriptions from these competitors: [Competitor A], [Competitor B], [Competitor C]. For each competitor, list every explicit offer or incentive mentioned (free trial, discount, guarantee, bonus, money-back policy, etc.). Then rank the competitors from most aggressive to least aggressive offer strategy. Flag any offer that none of your competitors are making.”
Example output: Competitor A offers a 14-day free trial and a no-credit-card-required signup. Competitor B leads with a 20% discount and a satisfaction guarantee. Competitor C mentions none. The AI flags that no competitor is advertising a results guarantee, which could be a differentiator.
Ad Radar data field: Ad descriptions and display URLs. The description field is where offers typically appear, though some competitors embed them in headlines.
Prompt 4: Landing Page Offer vs. Ad Offer Consistency Check
Job: Identify whether a competitor’s ad promises match their landing page.
Data needed: Ad copy plus the content of the destination URL.
The prompt:
“Here is the ad copy for [Competitor Name]: [paste ad]. Here is the content from their landing page: [paste relevant page text]. Does the landing page deliver on every promise made in the ad? List any promises in the ad that are not substantiated on the landing page. Note any friction points a visitor would encounter.”
Example output: The ad promises “setup in under 5 minutes” but the landing page requires a sales call before any trial access. The AI flags this as a likely trust-eroding inconsistency and suggests you can position your own product’s self-serve signup as a direct contrast.
Ad Radar data field: Final URL captured alongside the ad. Ad Radar records both the display URL and the final destination URL for each detected search ad.
Group 3: Positioning Gap Analysis
Positioning gaps are the angles your competitors are leaving unclaimed. These are the most valuable findings from any competitive analysis because they show you where you can win without fighting head-on.
Prompt 5: Unclaimed Benefit Finder
Job: Find the customer benefits that nobody in your category is advertising.
Data needed: Ad headlines and descriptions from three to five competitors on the same keyword.
The prompt:
“Here are ads from the top competitors for the keyword ‘[your keyword]’: [paste all ads]. Based on this copy, list every benefit that is being claimed by at least two competitors. Then list every benefit that a customer might care about that no competitor is mentioning. For the unclaimed benefits, rate each one as High, Medium, or Low priority based on how likely a buyer in this category is to care about it.”
Example output: All competitors mention speed and ease of use. Nobody mentions data privacy, local support, or integration with legacy software. The AI rates data privacy as High priority given the industry context, and legacy integration as Medium.
Ad Radar data field: Headlines and descriptions across all detected competitors for a monitored keyword. Because Ad Radar is keyword-based, you can pull all ads appearing for a specific search term and analyze them together.
Prompt 6: Audience Segment Gap
Job: Identify customer segments that competitors are ignoring in their messaging.
Data needed: A collection of competitor ads for one product category.
The prompt:
“Read these competitor ads for [product category]: [paste ads]. What type of customer does the language in these ads appear to target? What customer segments are clearly not being spoken to based on the vocabulary, examples, or pain points used? List three to five underserved segments with a recommended message angle for each.”
Example output: All ads appear to target mid-sized B2B teams. Freelancers, enterprise procurement teams, and non-English-speaking markets are not addressed in any ad. The AI suggests a freelancer-focused headline variant as the highest-volume underserved segment.
Group 4: Brand-Attack Check
Brand-attack monitoring is the practice of watching whether competitors bid on your brand name and what they say when they do. This is one of the most commercially sensitive areas of PPC intelligence. If you want background on filing a formal complaint when competitors misuse your trademark in ad text, see how to file a Google Ads trademark complaint with the right evidence.
Prompt 7: Competitor Brand-Bid Ad Analysis
Job: Understand how a competitor positions themselves when targeting your brand keyword.
Data needed: Ad copy from a competitor appearing on your brand keyword search.
The prompt:
“A competitor is running this ad when someone searches for [Your Brand Name]: [paste ad]. Analyze: (1) Are they making a direct comparison or simply appearing without mentioning your brand? (2) What specific weaknesses in your brand are they implying? (3) What counter-messaging could neutralize this in your own brand-keyword ad? Write three alternative headlines you could use to defend this keyword.”
Example output: The competitor implies higher pricing and slower onboarding without naming the brand directly. The AI suggests counter-headlines like “[Brand]: Pricing That Scales With You” and “[Brand]: Live Setup Support Included” to address the implied objections.
Ad Radar data field: Ad Radar includes brand protection monitoring. You can enter your own brand name as a monitored keyword and receive an alert whenever a competitor ad appears on that term. The alert includes the ad copy, the competitor’s display URL, and the detection timestamp, which is exactly what this prompt needs as input.
Prompt 8: Comparative Claim Detector
Job: Identify whether a competitor is making claims that explicitly or implicitly compare themselves to named alternatives.
Data needed: A batch of competitor ads across multiple keywords.
The prompt:
“Review these ads from [Competitor Name]: [paste ads]. Flag any headline or description that contains a direct comparison to another product or brand, an implied superiority claim (‘the only,’ ‘unlike others,’ ‘better than’), or a stat that invites comparison. For each flagged line, explain what the implied comparison is and whether it could be challenged.”
Example output: The AI flags three descriptions containing “unlike traditional software” and one headline using “the only platform that” followed by an unverified claim. It notes these could be challenged in ad feedback if they are factually unsupported.
Group 5: Meta Audience Inference
Meta ads do not show targeting data publicly, but the creative, copy, and format choices in a competitor’s Meta ad tell you a lot about who they think their customer is. The prompts in this group help you read between the lines.
Prompt 9: Meta Ad Audience Reverse-Engineering
Job: Infer the intended audience from a Meta ad’s creative and copy choices.
Data needed: Meta ad copy, headline, and a description of the creative visual.
The prompt:
“Here is a Meta ad from [Competitor Name]. Headline: [paste]. Body copy: [paste]. Creative description: [describe the image or video]. Based on these elements, infer: (1) the demographic most likely being targeted (age range, role, lifestyle), (2) the psychographic profile (what they value, what frustrates them), (3) what stage of the buying journey this ad is designed for (awareness, consideration, conversion). Explain your reasoning.”
Example output: The lifestyle imagery and casual language suggest a 25-40 demographic, the mention of “your team” implies a manager or founder, and the lack of a price point or trial offer suggests this is a top-of-funnel awareness ad rather than a conversion play.
Ad Radar data field: Meta Ads data from Ad Radar, which connects to the Meta Ad Library API. For each detected Meta ad, Ad Radar captures the ad copy, headline, destination URL, and detection timestamp. You will need to describe the creative yourself or use a screenshot, since image interpretation requires a multimodal AI model like GPT-4o or Gemini.
Prompt 10: Meta Ad Format and Creative Strategy Audit
Job: Understand how a competitor structures their Meta creative across multiple active ads.
Data needed: A list of active Meta ads from one competitor, including format details.
The prompt:
“Here are [X] active Meta ads from [Competitor Name]: [paste all copy and describe formats]. Identify: (1) the dominant creative format they favor (video, static image, carousel), (2) the hook strategy in the first line of body copy, (3) their call-to-action pattern. Based on this, what does their Meta ad strategy appear to prioritize: brand awareness, lead generation, or direct conversion?”
Example output: The competitor uses primarily static image ads with a problem-first hook (“Tired of chasing invoices?”), short body copy, and “Get Started Free” CTAs consistently. The AI concludes this is a bottom-of-funnel direct response strategy aimed at people already aware of the product category.
For more on the full range of monitoring and AI features inside Adsroid, the features page gives a complete breakdown of what is tracked and how.
Group 6: Display Creative Themes
Display ads are often overlooked in competitor research because they are harder to collect systematically. But the visual and copy themes a competitor uses in display campaigns reveal their retargeting strategy, brand storytelling priorities, and product emphasis in ways that search ads do not.
Prompt 11: Display Ad Creative Theme Mapping
Job: Identify the recurring visual and copy themes in a competitor’s display campaign.
Data needed: Ad copy, headlines, and creative descriptions from multiple display ads from one competitor domain.
The prompt:
“Here are [X] display ads from [Competitor Name], collected across different placements: [paste copy and describe creatives]. Identify: (1) the recurring visual themes (color palette, imagery style, human vs. product-focused), (2) the messaging themes that appear across multiple ads, (3) any evidence of sequential retargeting (do some ads appear designed for repeat viewers vs. cold audiences?). Summarize the display strategy in one paragraph.”
Example output: The competitor uses blue and white consistently, favors product screenshots over lifestyle imagery, and has three distinct message variants that appear to follow a sequence: awareness (problem statement), consideration (feature focus), and retargeting (discount + urgency). This suggests a structured display funnel rather than a single always-on creative.
Ad Radar data field: Display Ads data from Ad Radar. Display monitoring in Ad Radar is domain-based. You enter a competitor’s domain and Ad Radar monitors display ads running from that domain, capturing the ad copy, headline, destination URL, and detection timestamp. This is the raw material this prompt needs.
Prompt 12: Display vs. Search Messaging Consistency Check
Job: Find out whether a competitor uses different narratives in display versus search, and what that tells you about their funnel strategy.
Data needed: Search ad copy and display ad copy from the same competitor.
The prompt:
“Here are search ads from [Competitor Name]: [paste]. Here are their display ads: [paste]. Compare the messaging in both channels. Where does the messaging align? Where does it diverge? What does the divergence suggest about how they think about each channel’s role in the funnel? Identify any message in their display ads that you do not see in their search ads, and vice versa.”
Example output: The search ads focus entirely on features and pricing signals. The display ads use customer story language and outcome-focused copy without mentioning pricing. The AI infers that the competitor uses display for brand-building and trust with cold audiences, while search handles intent-driven conversion. A message about customer success that appears only in display could be an underused angle for search too.
Ad Radar data field: Both Search Ads data (keyword-based, capturing headlines, descriptions, and URLs) and Display Ads data (domain-based, capturing copy, headline, and destination URL). Having both in one place makes cross-channel comparison straightforward.
Getting the Data These Prompts Need
Every prompt in this library is only as good as the data you feed it. There are several ways to collect competitor ad data.
Manual collection works at small scale. You can search your target keywords on Google, screenshot the ads you see, and manually type the copy into your AI prompt. For Meta, the Meta Ad Library allows you to search any advertiser and browse their active ads. This approach is free but time-consuming and inconsistent. You will only capture ads that happen to be showing when you check, not the full picture of what a competitor is running.
Ad Radar, built into Adsroid, automates this collection. For search ads, you define the keywords you want to monitor and Ad Radar scans Google and Bing SERPs continuously. For display ads, you enter a competitor domain and Ad Radar monitors their display activity. For Meta, it connects to the Meta Ad Library API and pulls active ads automatically. Every detected ad is logged with its copy, URL, and timestamp.
The practical benefit for these prompts is that you get a consistent, timestamped record of competitor ad activity without manual checking. When a competitor changes their headline or launches a new offer, Ad Radar sends an alert, and you can drop the new copy straight into one of the prompts above to understand what changed and why it might matter.
If you want to understand how to set up systematic alerts for ad changes more broadly, the guide on setting up Google Ads and Meta Ads alerts covers the logic behind alert thresholds and routing in useful detail.
Common Mistakes When Using AI for Competitor Ad Analysis
- Feeding the AI stale data. Competitor ads change frequently. An ad you collected three months ago may no longer be running. AI analysis of outdated copy produces outdated conclusions.
- Using too few ads. Analyzing two or three ads from a competitor and drawing conclusions about their strategy is unreliable. Collect as many ads as you can before running a pattern or positioning prompt.
- Skipping the “why” question. These prompts are designed to produce insights, not just descriptions. If the AI tells you what a competitor is saying, always follow up with a prompt asking why they might be saying it and what it implies for your own strategy.
- Ignoring the destination URL. Ad copy is only half the story. The landing page a competitor sends traffic to often reveals more about their offer and funnel than the ad itself.
- Treating AI output as final analysis. These prompts are a starting point. The AI will sometimes draw plausible-sounding but incorrect inferences. Use the output as a hypothesis to validate, not a conclusion to act on without review.
The goal of AI-assisted competitor ad analysis is not to copy what competitors are doing. It is to understand what they are doing well enough to find the angles they have left uncovered.
Putting It Together: A Simple Workflow
- Define the keywords or competitor domains you want to monitor.
- Collect live ad data from those keywords and domains (manually or through Ad Radar).
- Group the data by competitor.
- Run Prompts 1 and 2 to understand each competitor’s messaging and structure.
- Run Prompts 3 and 4 to map their offers and check consistency.
- Run Prompts 5 and 6 to find positioning gaps.
- Run Prompts 7 and 8 if brand-bidding is a concern.
- Run Prompts 9 and 10 on Meta ad data.
- Run Prompts 11 and 12 on display data if available.
- Compile the gaps and angles into a brief for your copywriter or use them directly to test new ad variants.
This workflow can be run monthly as a standing competitive review or triggered whenever Ad Radar sends an alert that a competitor has changed their copy. The prompts stay the same. Only the data changes.
If you want to start collecting the live competitor ad data these prompts need, Ad Radar inside Adsroid handles the monitoring across Google Search, Bing, Meta, and Display from a single setup.
Frequently Asked Questions
What AI prompts can I use to analyze competitor ads with ChatGPT?
The most effective approach is to structure your prompt around a specific job: messaging teardown, offer comparison, positioning gap, or audience inference. Feed ChatGPT the actual ad copy and ask it to identify patterns, gaps, or implied strategies rather than simply describe the ads. The 12 prompts in this article are designed for exactly this purpose and can be pasted directly into ChatGPT with your competitor data substituted in.
What is a good prompt to find gaps in competitor ad copy?
A reliable prompt for gap analysis is: “Here are ads from the top competitors for [keyword]. List every benefit being claimed by at least two competitors. Then list every benefit a customer in this category might care about that no competitor is mentioning. Rate each unclaimed benefit by purchase relevance.” This forces the AI to look at what is absent rather than just summarizing what is present.
Can I use Claude or Gemini instead of ChatGPT for these prompts?
Yes. All 12 prompts in this library work with Claude, Gemini, and Perplexity without modification. The output quality varies slightly between models, but the prompt structure is model-agnostic. For prompts involving creative descriptions from display or Meta ads, a multimodal model like GPT-4o or Gemini can interpret images directly if you upload a screenshot rather than describing the creative in text.
How do I collect competitor ad data to use with these prompts?
You have two main options. Manual collection involves searching your target keywords on Google, browsing the Meta Ad Library, and copying ad copy by hand. This is free but slow and incomplete. Automated collection tools like Ad Radar inside Adsroid monitor competitor ads continuously across Google Search, Bing, Meta, and Display and log every detected ad with its copy, URL, and timestamp, which you can then paste directly into your AI prompts.
How often should I run competitor ad analysis?
For most advertisers, a monthly review is sufficient to spot strategic shifts. However, if you are in a competitive category with frequent promotions or seasonal campaigns, a weekly review is more useful. Setting up automated alerts for ad copy changes means you do not have to run the full analysis on a schedule. You run it when something actually changes.
Can AI tell me what audiences my competitors are targeting on Meta?
Not directly, since Meta does not publish targeting data. But AI can infer the likely intended audience from the language, imagery, hooks, and CTA choices in a competitor’s Meta ad copy. Prompt 9 in this article is designed for exactly this inference. The accuracy depends on how much creative and copy detail you provide as input.
Is it possible to use these prompts for Google Display ad analysis?
Yes. Prompts 11 and 12 are specifically designed for display ad analysis. The main challenge with display is data collection, since display ads are harder to find manually than search ads. Domain-based monitoring tools like Ad Radar capture display ads running from a competitor’s domain, including ad copy, headline, and destination URL, which gives you the input these prompts need.
What should I do with the gaps my AI analysis finds?
Use them as hypotheses for new ad variants. If the analysis finds that no competitor is addressing data privacy concerns, write a headline that does and A/B test it against your current control. Not every gap represents an opportunity, since some are unclaimed for good reason. But the gaps that align with real customer concerns and your product’s actual strengths are worth testing immediately.
