How We Fixed Our Facebook Tracking, Stole Competitor Secrets with AI, and Why I Almost Quit Content

🔍 FUNNEL VISION:

Gary Vee always talks about being a practitioner.

That’s why he is so successful.

He actually does it on a day-to-day basis.

He's in the trenches.

And when you're actually doing not just teaching, sometimes it's frustrating to keep up with all these content creators and coaches.

I love writing this newsletter, but sometimes the work takes over.

This is a constant battle I have with myself.

To help combat this issue, I'm going to mix in some general personal development content as well as business/marketing.

I think the two go hand-in-hand and I love talking about both.

So if you've ever been battling with this issue yourself, just know it's OK and sometimes you just have to keep pushing.

🥡 THIS WEEK'S FUNNEL FEED

What we tested: We finally fixed our Meta Pixel tracking issues using our own company, Upstack Data, to implement proper server-side tracking and attribution modeling after months of inconsistent data that was killing our ad performance and budget allocation decisions.

What happened: Our ads absolutely exploded. ROAS jumped 40% in the first week as we could finally see which campaigns were actually driving conversions versus the ones that looked good but were attribution ghosts. We're now scaling spend confidently for the first time in months because we trust our data again.

Why it worked: Three critical tracking principles that most brands get wrong:

1. iOS 14.5 reality: Browser-based tracking is dead - server-side tracking captures the conversions that Facebook's pixel misses due to privacy updates

2. Attribution accuracy: We can now properly attribute conversions to the right campaigns, audiences, and creative, eliminating the guesswork that was wasting our budget

3. Scaling confidence: When you trust your data, you can push winning campaigns harder and kill losers faster, dramatically improving overall account performance

How to try it:

FOR $10M+ BRANDS:

• Implement server-side tracking through tools like Upstack Data, Triple Whale, or Northbeam

• Set up first-party data collection to reduce reliance on third-party cookies

• Create attribution models that account for multi-touch customer journeys

• Use conversion APIs to send purchase data directly to Meta's servers for better optimization

STARTING OUT? TRY THIS:

• Start with Google Analytics 4 enhanced e-commerce tracking as your foundation

• Use UTM parameters religiously to track traffic sources and campaign performance

• Set up Facebook Conversions API (free) to supplement your pixel data

• Focus on first-party data collection through email capture and customer surveys to build your own attribution picture

🍱 THIS WEEK'S CONTENT BUFFET

These are some pieces of content that I felt inspired to write about.

If you're starting a new brand, this is the way you should test ads.

So many people come to me when they start off and don't know what to do.

This is the simplest way to get data on whether or not people are interested in your product and what angle are they interested in the most?

From there, you take that data and double down on what people are engaging with the most.

Sahil is probably one of my favorite voices online.

His book is killer.

Good things take time to build.

Social media makes it seem that everybody is an overnight success but they're not.

Typically the ones that seem like they are - are usually faking it.

It's why I try to stay off social media as much as possible.

But I'm human and by myself going down the rabbit hole from time to time.

What I noticed, though is my mental health goes to shit.

It's subtle, so I don't really realize it but it's happening.

🍟 SWIPE OF THE WEEK

Why it's smart:

The US v Them AD never fails.

This is a must for any ad account.

When a market is saturated, you need to show why you’re better.

And you need to do it clearly and concisely.

Steal this idea: 

Use Reddit and other social media channels to see what customers are saying about competitors.

From there, you can then figure out how to position your company as a superior option.

You should also do this when creating a product, but that’s a conversation for another day.

🧪 AI SECRET SAUCE

Tool/Prompt I'm Using: Claude/ChatGPT to systematically create winning ad variations from our top-performing creative, analyzing what makes them work and generating multiple iterations that maintain the core psychological triggers while testing new angles.

Step 1: Performance Analysis & Setup:

• Export your top 5 performing ads from the last 90 days (highest ROAS, not just reach)

• Screenshot each ad and document the key metrics (CTR, conversion rate, ROAS)

• Identify the common elements: hooks, pain points, social proof, offers

• Create a brief for each ad explaining why you think it performed well

Step 2: AI Iteration Prompts:

1. "Analyze this high-performing ad and identify the 3 key psychological triggers that made it successful. Then create 5 variations that maintain these triggers but use different hooks, angles, or social proof elements."

2. "This ad got a 4.2% CTR and 3.8x ROAS. Create 3 variations that target the same pain point but approach it from different emotional angles (fear, aspiration, frustration)."

3. "Take the core message from this winning ad and rewrite it for 3 different customer avatars while keeping the same offer and urgency."

4. "Create 5 iterations of this ad that test different lengths: one 50% shorter, one 50% longer, and three that reorganize the information in different sequences."

Step 3: Systematic Testing Framework:

• Launch iterations in small test groups (10% of winning ad's budget)

• Test one variable at a time (hook vs. body copy vs. CTA)

• Give each iteration 3-5 days minimum before making decisions

• Keep the winning creative running while testing variations

• Document what specific elements drive performance differences

How to use this approach:

• Start with your absolute best performer and create 3-5 iterations before moving to the next

• Use AI to generate variations you'd never think of yourself

• Test seasonal/timely angles of your proven winners

• Create a "variation bank" of proven psychological triggers you can apply to new products

Pro tip: Add this instruction for better results: "Before creating variations, identify the specific customer objection or desire this ad addresses. Then create iterations that address the same core need but through different emotional pathways, social proof types, or urgency mechanisms. Maintain the original's conversion intent while exploring new creative territories.

🧂 UNDERRATED TACTIC

Download your competitors' Facebook Ad Library data and upload it to ChatGPT for pattern analysis.

Go to Facebook Ad Library, search your competitors, screenshot their last 20 ads, then ask AI: "What hooks, pain points, and offers do these ads have in common? What gaps can I exploit?"

Most brands spy on competitors manually. Smart brands let AI find the patterns they'd miss.

🧃 FOUNDER FUEL

🍽️ WHAT'S COOKING

I’m working on a passion project that I truly believe can change the world.

Stay tuned for updates.

✌️ That’s All For This Week

Hope you guys got value out of this email.

If you have any specific questions or want me to cover any topics, just respond to this email and I'll do my best to get to them.

See you next week!