How to Spy Like a Strategist (Not a Copycat) (AI Creative Strategy, Edition #003)

👋 Hey friends,

When was the last time you audited your competitors’ Meta ads – and came away with something actually useful? Most people browse the Meta Ad Library… and stop at screenshots.

That’s surface-level swiping. What you want is creative intel – the kind that tells you:

• What angles they’re pushing

• Which offers they’re testing

• Where they’re vulnerable

This issue breaks down how to do competitor research like a strategist – so you don’t copy them… you out-position them.

🔍 Big Idea

Your competitors are running your next winning angle – you just need to deconstruct it.

Instead of saving ads blindly, learn to reverse-engineer the psychology, patterns, and gaps behind them.

🧱 Framework

🧠 “The 4-Lens Competitor Ad Breakdown”

🧠 Pro Tip: Use a spreadsheet to log each ad and tag it by angle type, offer type, and hook structure. Patterns emerge fast.

🧠 Real-Life Application

Do This 30-Min Sprint:

1. Pick 3 direct competitors

2. Save their current Meta ads

3. Break each one down using the 4-Lens Framework

4. Identify 1 “angle gap” per brand

5. Build your next test around the gap

Use Vibe9’s Video 2 Text tool to transcribe any video ad in seconds. Just paste the video URL, and you’ll get a full text version – perfect for breaking down hooks, spotting message patterns, and feeding into AI prompts for deeper analysis.

🧰 Bonus: AI Prompt to Use Today

This post teaches how to mine real customer language from reviews and comments to uncover emotionally resonant pain points – and turn them into scroll-stopping ad hooks.

🛠️ How to Use It
💡 Use this after grabbing a video transcript from Vibe9’s Video 2 Text tool – it turns competitor content into a creative strategy playbook.

💣 Value Bomb

🧠 Don’t chase trends – track consistency. If they’ve run a version of the same angle for 60+ days? That’s the one worth decoding and remixing.

🛠 Tactical Tip

Use Meta Ad Library with filters:

• Sort by impressions (highest first)

• Look for ads with multiple versions – often means testing + scale

• Use archive tools to check if the same ad has run in multiple waves

🎤💻 Last 7 days content summary

• Prompting is the new copywriting.

• How to Create 10x More Ads With 0 Extra Production? You don’t need more content. You need a remix system.

🎤 My take on it

The real value of competitor research isn’t in finding ads to copy – it’s in uncovering how the market thinks so you can say something it hasn’t heard yet. Too often, brands analyze ads through a surface-level lens: what platform it ran on, what the CTA was, or whether it used UGC. But creative strategists look deeper – they treat every ad as a signal of positioning, belief systems, and marketing psychology. The question isn’t “Is this ad good?” – it’s “What does this ad believe about the customer?”

Because behind every message is a strategic assumption: that customers care more about social proof than features, or that urgency will beat education. Your job is to evaluate those assumptions against your own research and either confirm, challenge, or invert them. This is how you carve out new ground. True competitor research reveals the invisible patterns: which offers are saturating, which pain points are overused, which tones feel templated, and most importantly – which angles have been overexposed to the point of ad fatigue. When you start to notice sameness across your category (everyone leading with “no harsh chemicals” or “trusted by 10,000+ customers”), that’s not a cue to join the chorus – it’s your moment to introduce contrast.

Great research isn’t reactive – it’s perspective-building. It helps you understand how customers are being conditioned to see your category, so you can deliberately reshape that perception. This is where the strategist steps up: not to follow what’s trending, but to decide what gets to trend next. It’s not just about what ads are running, but which ones are being repeated, which messages are scaling, and which formats brands haven’t dared to test yet. You’re not collecting data – you’re constructing a creative worldview:

• What promises are gaining traction?

• What stories feel played out?

• What customer objections are being dodged instead of tackled head-on?

Competitor research, when done well, gives you a map of the mental terrain your customers are navigating – and shows you where there's room to build a new path. The outcome isn’t a swipe file. It’s a differentiated idea that hits harder because it was born in context. When you can see the patterns, you can break them on purpose. That’s what makes research strategic. And that’s why the best Meta ads don’t just look different – they think different. Not because the brand is louder, but because the strategist behind it is sharper.

👋 Coming Up

Next Tuesday: Offer & Brand Research – how to map your offer strategy directly to your creative messaging for max ROAS.

Reply if you want a competitor breakdown template – I’ve got a good one.

PS

Deep research might be the difference between "nice ads" and real performance.

I’ve been playing with a bunch of AI tools lately for creative research and here’s the TL;DR: 90% of them are just repackaging surface-level stuff. If you’re serious about creative strategy, deep research still wins.

Here is my video breakdown, and if you don't want to watch, it I put a short text summary below:

I used the SidekickTool.com brand (not affiliated) as a test case and ran the same research prompt across a bunch of AI tools to see what would actually give me something I could use for messaging/angle building.

💡Big takeaway: Most tools just give you output, not actual research. Here’s what I found:

❌ What didn’t work:

• Gwen, Claude, DeepSeek, You.com, Copilot: mostly missed the mark. Either didn’t understand the product well, or spat out short summaries without real insight. Fine if you want a paragraph. Useless if you want raw material for high-performing ads.

• DeepSeek took 58s, but the depth wasn’t there.

• Copilot had a few okay lines, but not much to build on.

✅ Decent but not great:

• Abacus.ai & GetLiner: these gave more angles and insights. With some refinement, you could build something out of it.

• Grok: solid mid-tier. You’ll get useful stuff, but you’ll still need to prompt and filter a lot.

• Perplexity PRO – legit pulls insights from forums like Reddit and gives you a thread to pull on. Also has follow-up functionality which makes digging deeper easier.

🔥 What actually impressed me:

• Gemini 1.5 Pro & ChatGPT 4: these are next-level. Tons of raw insights, customer language, pain points, belief shifts, objections… all structured in a way that feels like you're halfway to writing copy.

My prompt was pretty dialed in (gave it structure, guardrails, even books/authors to pull from), and I think that made a big difference. But even so, Gemini and ChatGPT still outperformed by a mile.

👉 If you’re building a research-backed creative strategy doc, start with one of those two. 

Export the insights, load into a CustomGPT, and you’re set up to scale messaging like a machine.