How to Use Analytics to Improve Ad Copy

Everyone’s trying to use AI to write better ads. But the real unlock isn’t the tools — it’s building a system that turns raw customer data into a machine for scalable ad concepts. At Obvi, we’ve been actively building a system designed to: → Surface deep customer insights → Map them to scalable ad frameworks → Generate ad angles that humans alone couldn’t uncover Not just faster. Fundamentally smarter. Here’s the system so far: Step 1: Fuel the system with raw customer language We collect authentic, emotional data from: - Post-purchase surveys - Product reviews - Customer service tickets - Ad comments - Email replies The more you can educate the machine with real customer feedback, the better. It gives the system deep context to work from. Step 2: Extract personas from the data We use this AI prompt (with uploaded files): "Act as a psychologist and direct response marketer. Identify 5-6 buyer personas based on motivations, pain points, and desires. Use real language from the data." This becomes the emotional blueprint of our system. Step 3: Map product benefits to personas Next, we prompt: *"Match our product features to each persona’s needs. For each: - 3 most compelling benefits - Emotional drivers - Specific pain points solved - Real language patterns to mirror. Now the system can speak to customers in the frame of their wants and needs. Step 4: Analyze top-performing ads We upload screenshots of our best ads and prompt: *"Dissect this ad: - Psychological techniques - Messaging frameworks - Tone, structure, voice - What made it work at a deep level."* This helps the system understand not just what customers want — but how we’ve already succeeded in reaching them. Step 5: Generate new ad creative components For each persona, we prompt AI to create: - 3 headlines (5-7 words) - 3 subheads (10-12 words) - 3 body copy snippets (2-3 sentences) - 3 trust signals - 3 CTAs (matched to tone: authoritative, empathetic, or urgent) Each output is grounded in real customer psychology, not guesswork. Step 6: Humanize and design This is where human judgment takes over. Anything AI generates is conceptual — the final polish and emotional resonance comes from creative teams who know how to craft stories that move people. Early Results: - More diverse, hyper-targeted ad angles at scale - Messaging that resonates with specific personas and pain points - Creative cycles moving with superhuman intelligence Like I said, my goal is not just faster, but smarter ad creation. 🔑This isn’t "using AI to write ads." It’s about building a system that compounds customer learnings in ways people can't — and translating those insights into scalable creative. This system is still evolving.  If you’re working on something similar, I’d love to swap notes. One thing feels clear: The brands that master customer insight systems will dominate the next phase of performance marketing.

I've audited 43 Meta ad accounts this year. Here's the 9-step AI-powered system I use to find predictable wins. (The secret? Your internal data + external viral structures = winning ads) 1. Set up your Motion (Creative Analytics) top-performing ads report → Filter for last 90 days → Minimum $1000 spend per ad → Look for outliers (> 20% improvements) This reveals insights only YOU know about your performance. 2. Build your audience comparison report → Use naming conventions to segment → Example filter: "ad name contains [millennials]" vs "ad name contains [gen-z'" → Compare CPA's across segments Shows which audiences actually convert for your brand. 3. Create your angle comparison report → Example Group 1: "ad name contains [comfort]" → Example Group 2: "ad name contains [style]" → Analyze performance differences Reveals which message your audience responds to. 4. Set up your offer comparison report. → Compare "20% off" vs "BOGO" vs "free shipping" → Same naming convention logic → Track conversion rates by offer type Discover what actually drives your customers to convert. 5. Analyze your unique winning combination → Look for intersection of audience + angle + offer → Find your highest-performing combo → Document the winning formula This is proprietary intelligence your competitors don't have. 6. Create an Apify account and configure the TikTok scraper agent → Set up scraper with category hashtags → Example: #skincare #skincareroutine #glowup #skincaretips → Set date range for recent viral content Cast a wide net to capture what's working organically. 7. Export viral content and analyze patterns → Download CSV with all scraped videos → Sort by view count and engagement rate → Identify top 20 most viral formats Hundreds of proven formats at your fingertips. 8. Connect to n8n workflow (advanced level) → Auto-analyze video transcripts with Claude → Extract winning hooks and structures → Create an automated database of viral elements Saves hours of manual analysis. 9. Combine your brand insights with viral formats → Take your winning audience/angle/offer combo → Apply it to top viral video structures → Create new ads using this hybrid approach You're combining internal intelligence with external proof. The magic happens when you merge what only you know about your brand with what the internet has proven works. Most brands fail because they either ignore their own data or ignore what's working organically. This system gives you the best of both worlds.

Big wins often happen from lots of small moves. Here's a real example from a client of Storybook Marketing, where we are managing the paid demand gen program. In this program, leads are moved through an orchestrated demand gen journey, through paid media and email marketing, ultimately being retargeted towards a demo request conversion at the end. As we took ownership of the program, one key area of opportunity stuck out in the way people were being asked for demos at the end of that journey. Most notably, they were all being delivered a general demo request message, and the results were low as a result. The small, but meaningful, change we quickly implemented was building out a campaign for each audience using language and imagery that was tailored to them. It took a couple of iterations to find what resonated best, but the key learnings were these: 1. The headline mattered so much No matter how 'warm' you think a lead is, the attention your ad will get in the feed is still minimal. You can't expect them to do work to see themselves in the ad copy. So by making it clear that were were speaking to them directly, it dramatically improved the ad performance. 2. Build visuals that they like, not what you like Too often, marketers build campaigns that appeal to marketers, but there are plenty of audiences who negatively respond to this. We saw exactly this within the program. One audience responded well to traditionally designed ads because of the persona, but the other was anything but responsive. By researching how other successful orgs target this persona in their ads, we rebuilt the ad imagery entirely, and the results quickly followed. This campaign all happened inside of 3 months, in a mature program, which is often where you don't expect to see such dramatic results. But dramatic they were - an overall drop in the cost/MQL of 52% and a 5x increase in the form conversion rate. And if there's just one key takeaway, it's this: Take the time periodically to make sure you're not leaving opportunity on the table because you served up too general a message or offering. Of course, it can be challenging to go this deep, and an agency like ours has a luxury of time that in-house marketers rarely do. And there is a limit to how personalized you can reasonably get, but those generalized moments that try to speak to everyone often speak to no one at all. And it can be leaving very wins on the table.

One of the most underappreciated data points Google Ads gives is the "Quality Score." I’ve audited MULTIPLE enterprise B2B SaaS port-cos that neglected it, burning hundreds of thousands of dollars a month in wasted spend. >> What It Is Quality Score << Google's 1-10 rating of the quality and relevance of your keywords and PPC ads. It's made up of: — Expected clickthrough rate — Ad relevance — Landing page experience >> What’s a Good Quality Score << A score of 7-10 is considered good, 5-6 is average, and 1-4 sucks. Anything under a 7 is wasted money. >> Why It's Important << Quality Score directly affects your ad rank and cost per click (CPC). Higher scores = better ad positions and lower costs. Lower scores: crappy ad positions and higher costs. Think about it as Google's way of rewarding relevant, well-crafted ads. Per SEMRush, improving your Quality Score from 5 to 7 can reduce your CPC by up to 28%. And a 10 can decrease CPC compared to the average by 30-50%. >> Where to Check It << 1. Sign in to your Google Ads account. 2. Navigate to the "Keywords" section 3. Add the “Columns” Quality Score, Landing page exp., Ad relevance, Exp. CTR 4. Gasp when you see shitty scores and the words “Below average” >> Ignore It At Your Own Risk << I've seen companies waste millions because of low Quality Scores. One B2B SaaS company was spending $100K/week on ads with scores of 1-3. Google was charging them $42 per click for some keywords. They might as well have been throwing money around like confetti. >> How to Optimize Quality Scores << 1. Ensure your keyword, ad copy, and landing page all align on language & value prop/s. 2. Use your target keywords in your ad copy (seems obvious, but you'd be surprised). 3. Make sure your landing page is relevant to the keyword (make sure the keyword’s in there a few times). 4. Use exact instead of broad-match keywords. 5. Review and optimize underperforming keywords weekly if not more P.S. Fellow marketers, let's have at it. What are your thoughts on Quality Score: good, bad, ugly, and badass?

Your ad metrics tell you exactly what's wrong ↳ (it's not rocket science) I've audited 100+ Facebook ad accounts over the last 12 months. And the same story keeps repeating itself. Brands complaining about poor performance but ignoring the obvious signals right in front of them. Your Facebook metrics are literally telling you what's wrong. You just need to know which combinations to look for. 1. High Hook Rate, Low CVR     Your ad stops the scroll but doesn't convert. Why it's happening: → Hook is completely unrelated to your product → You've baited viewers with something misleading → Product reveal is disappointing after the hook The fix: → Make sure your hook relates to what you're actually selling → Test adding your CTA earlier in the video → Strengthen connection between hook and product ------ 2. Low Hook Rate, High CVR     The few people who watch your ad convert, but not enough people are watching. Why it's happening: → First 3 seconds are boring → Thumbnail doesn't grab attention → No pattern interrupt to stop scrolling The fix: → Test 5+ completely different hooks → Use split screens in your intro → Lead with your strongest claim or pain point ------ 3. High CVR, Low CTR     Your site converts well, but your ad doesn't get clicks. Why it's happening: → Call to action is weak → No reason to click NOW → Ad copy doesn't drive action The fix: → Make your CTA impossible to miss → Add urgency/scarcity elements → Test stronger benefits in your copy ------ 4. High CTR, Low CVR     People click your ads but bounce from your site. Why it's happening: → Major disconnect between ad and landing page → Post-click experience is confusing → Ad sets unrealistic expectations The fix: → Use identical messaging in ad and landing page → Simplify your landing page experience → Match ad imagery with site imagery ------ 5. High CTR, Low Hook Rate     People are clicking without watching your video. Why it's happening: → Misleading thumbnail → Ad copy doesn't match creative → Getting wrong traffic that won't convert The fix: → Create thumbnails that honestly show your offer → Test hooks that align with your copy → Fix any misleading elements — Stop guessing why your Facebook ads aren't working. The diagnosis is right there in your metrics. You just need to know where to look. If you've found this valuable, drop me Evan Carroll a follow and repost to help others in your network with their Facebook ads ♻️

I recently worked on a campaign that wasn't yielding the desired results. Instead of throwing in the towel, I decided to take a data-driven approach. Using Google Analytics, I analyzed the campaign's performance in detail. I looked at metrics like click-through rates, bounce rates, and conversion rates to identify areas for improvement. I discovered that the landing page was confusing and difficult to navigate. Users were getting lost and leaving before taking any action. Based on these insights, I made several changes, including simplifying the page layout and adding more clear calls to action. Additionally, I refined our targeting to reach a more relevant audience. By focusing on users who had shown interest in similar products or services, we were able to increase our conversion rate significantly. These data-driven adjustments resulted in a 42% increase in sales and a 23% improvement in return on investment (ROI). Have you used Google Analytics to turn around an underperforming campaign? Share your experiences and strategies in the comments below.