AI-powered automated campaigns—Performance Max, Advantage+ —promise simplicity. Upload your products, set a target ROAS, and let AI do the heavy lifting.
But here's what actually usually happens: Your growth in performance after testing period looks great, until you dig one layer deeper and you'll find your top 20% of products are generating 8x ROAS while subsidizing the bottom 50% that barely break even. The algorithm doesn't care about your profit margins. It cares about conversions. Any conversions.
Meanwhile, your new product launches get buried under bestsellers from 2022. Seasonal inventory sits untouched while out-of-season items burn the budget. High-margin winners fight for impressions against low-margin losers.
This isn't a Google problem. It's a data problem.
And it's costing you 30-40% in potential efficiency gains—according to recent industry analysis of brands that moved from blended to segmented campaign structures.
Let's talk about what segmentation actually means in 2025.
It's not just splitting campaigns by brand or category, which is still a valid strategy for some campaign types or industries. Real segmentation means organising your product catalog based on strategic performance metrics—ROAS, profit margin, inventory levels, seasonality, price competitiveness—and then building campaign structures that align budget allocation with business priorities.
Here's what advanced segmentation looks like in practice:
The framework that's become industry standard:
Brands using this framework typically see 30-40% efficiency improvements compared to simpler campaigns structures. Why? Because you're finally telling the algorithm what matters—not just letting it figure it out auction by auction.
Most brands know they should segment. The problem is getting the data to do it effectively.
You need:
Oh, and you need all this in one place, updated in real-time, and connected to your campaigns
This is where most brands hit a wall.
Let's look at what fast-fashion retailer did to transform their marketing efficiency:
They implemented segment-specific ROAS targets based on:
The result? They moved beyond their blended 4.2x ROAS and discovered massive optimization opportunities they couldn't see before.
But here's what they needed to make it work:
Not just "this product has 5x ROAS." But:
This is what industry leaders call "multi-dimensional segmentation"—and it's the difference between reactive campaign management and strategic growth.
Product performance isn't static. What was a hero last month might be a villain this month because:
You need continuous data refresh across all these dimensions. Not monthly exports. Not weekly updates. Real-time.
Once you've identified your product buckets, you need to:
Manually, this is a full-time job. Automated, it happens while you sleep.
Understanding your product portfolio isn't a nice-to-have for performance teams—it's a competitive requirement. After analyzing hundreds of campaign structures and product data across countries and industries, we found that 80% of accounts have potential savings of up to 20% while maintaining the same revenue.
The best part? You don't need huge structural changes to your running campaigns. No risk of losing precious campaign historical data.
You don't need to collect every possible data source from day one. You just need to shift your view from campaign-level optimization to product-level optimization.
One look from a different angle than your performance team usually takes—and boom, you might have just found thousands of dollars to reallocate. Simple, fast, and with immediate impact.
This is what we built Product Analytics to solve.
Not as a reporting dashboard. Not as a nice-to-have add-on. But as the foundational layer that makes advanced segmentation actually possible and easy to implement.
Product Analytics connects (November 2025 release)
All in one place. Automatically synced. Product-level granularity.
Within minutes Product Analytics shows you a specific issues:
Remember the Heroes/Sidekicks/Villains/Zombies framework? Villains are the products you need to uncover and fix. Product Analytics highlights them as Underperformers or Losers and alerts you to how serious the problem is.
Once you know which products are wasting or even burning your budget, and which are inactive or not being promoted, it’s time to choose the right Product Strategy.
If you use the Heroes/Sidekicks/Villains/Zombies framework, consider these two popular approaches:
By choosing either of these strategies, you divide products into Heroes, Sidekicks, and Villains.
For Zombies, you can apply the Zombie strategy, followed by a separate approach for new products.
Let's walk through a concrete example:
Fashion E-commerce Brand - $100K/month ad spend
Same brand, same budget. Now running 6 strategically segmented campaigns:
Net result:
Here's what's happening in e-commerce advertising right now:
CPCs are rising. Competition is intensifying. Customer acquisition costs are at all-time highs. The brands winning aren't spending more. They're spending smarter.
And "spending smarter" means:
All of this requires product-level intelligence that most platforms simply don't provide.
That's what Product Analytics was built for.
