- ADOTAT Newsletter - Adtech, Marketing and Media
- Posts
- The Funnel Trap: When Programmatic Optimization Goes Too Far
The Funnel Trap: When Programmatic Optimization Goes Too Far
Part 1: The Optimization Delusion

Learn AI in 5 minutes a day
This is the easiest way for a busy person wanting to learn AI in as little time as possible:
Sign up for The Rundown AI newsletter
They send you 5-minute email updates on the latest AI news and how to use it
You learn how to become 2x more productive by leveraging AI
📦 Part 1: The Optimization Delusion
The machines are working perfectly... at sabotaging your brand.
Hey folks—quick note: I’m out this entire week for the holiday of Pesach (yes, I was named after it). That doesn’t mean the machines get to run wild. I already produced this four-part series in advance—because if you let the algorithm take over while you’re out of office, you’ll come back to your brand selling crypto knives on cat meme websites.
Each section of this series has a free teaser for everyone, and a deeper dive for paid members—because just like your ad strategy, not everything should be free.
🚨 Let’s talk about the big lie: Optimization
We’ve all been sold the idea that more data + more automation = better results. But sometimes, when you let the AI cook, what you get isn’t a gourmet meal—it’s spam casserole.
🗺️ Brief History of Optimization Metrics
Let’s take a little trip down Memory Lane—the one paved with dashboards and false confidence:
CTR (Click-Through Rate): The OG metric. You clicked? Great! Except it turns out most clicks come from serial ad clickers, rage clickers, or your own QA guy testing creative.
CPA (Cost Per Action): Seemed smarter—tie spend to actual actions. But what kind of actions? If you're optimizing for free trial signups, don’t act shocked when you attract freebie hunters who ghost after day one.
ROAS (Return on Ad Spend): Now we’re cooking, right? Except... not really. ROAS tells you how much money you immediately made. It says nothing about the long-term health of your customer base. It’s like judging a restaurant by how fast it turns tables, not whether anyone comes back.
🤖 Rise of the Black Box: Smart Bidding, Dumb Strategy
“Let the algorithm decide,” they said. “It knows better,” they whispered. And you believed them.
Enter: Smart Bidding. Google’s black-box bidding engine claims to optimize for your goals based on hundreds of contextual signals—device, time of day, user intent, moon phase, maybe your ex’s Spotify playlist.
But here’s the thing:
You don’t know what it’s really doing.
And if you’re not careful, it starts optimizing for the easiest conversions—the low-hanging, half-rotten fruit.
Let’s say you tell it to go after clicks.
Guess what? You’ll get clicks. Lots of them.
From who?
People who love to click on everything.
(And buy absolutely nothing.)
You tell it to go after cheap sales.
You’ll get them.
From bargain shoppers who churn faster than you can say “cart abandonment.”
💣 The Hidden Cost of "Winning" Too Quickly
Here’s where it gets dangerous. When you only look at the now, you miss what really matters:
📉 Lifetime Value Tanks: That cheap customer? They never come back.
💔 No Loyalty, No Brand Love: You're just a coupon dispenser with a logo.
🧨 Innovation Dies: You stop investing in real growth.
🏃♂️ Consumers Run to Competitors: Because the one who builds the relationship wins.
It’s like picking fruit before it’s ripe. Sure, it looks good today, but bite into it and you’re left with bitter pulp and regrets.