Sales & Conversion
Personas
Ecommerce
Time to ROI
Short-term (< 3 months)
You know that feeling when you spend hours crafting the "perfect" Facebook audience setup? Layering demographics, interests, behaviors, lookalikes—creating these complex audience hierarchies that look impressive in your ads manager?
I used to be obsessed with this. I'd build custom audiences with surgical precision: "Women, 25-45, interested in sustainable fashion, who visited competitor websites in the last 30 days, excluding recent purchasers." The whole nine yards.
Then I discovered something that completely changed how I approach Meta ads: your creative IS your targeting. While everyone's still debating audience segments, the real game has shifted to creative testing at scale.
Here's what shifted my entire approach—working with an e-commerce client, I ran two campaigns. Campaign A had my "perfect" custom audiences. Campaign B had broad targeting with 3 new creatives weekly. Guess which one won? Campaign B crushed it, and here's exactly how we did it.
In this playbook, you'll learn:
Why advanced targeting is actually limiting your reach in 2025
The creative-first framework that replaced my audience obsession
How to set up creative testing systems that scale
The exact workflow that improved our client's ROAS from 2.5 to 8-9
When to still use custom audiences (spoiler: it's not what you think)
This isn't about abandoning targeting completely—it's about understanding where the real leverage lives in today's advertising landscape. Let's dive into what I discovered through months of testing with real budgets and real results.
Industry Reality
What every marketer thinks they know about Meta targeting
Walk into any marketing conference or browse through any ads-focused Facebook group, and you'll hear the same advice repeated like gospel: "The more specific your targeting, the better your results." Every guru has their "secret" custom audience setup that promises to unlock hidden pockets of high-converting traffic.
The conventional wisdom follows a predictable pattern:
Start with detailed targeting - Layer demographics, interests, and behaviors to create your "ideal customer avatar"
Build complex lookalike audiences - Create 1%, 2%, 3% lookalikes from your best customers, then exclude overlap
Use exclusion audiences religiously - Exclude recent purchasers, website visitors, and email subscribers to avoid "wasting budget"
Create hyper-specific custom audiences - Retarget people who viewed specific product pages for exact time periods
Test audience against audience - Run multiple ad sets with different targeting parameters to find the "winning audience"
This approach made sense in the Facebook ads golden era of 2015-2018. Back then, detailed targeting actually worked because Facebook had access to rich third-party data and users were more liberal with their personal information sharing.
But here's what most marketers haven't adjusted to: iOS 14.5, privacy regulations, and algorithm evolution have fundamentally broken this model. The detailed targeting that used to work is now operating on incomplete data sets. Meanwhile, Facebook's algorithm has become incredibly sophisticated at finding your ideal customers—if you let it work.
Yet most businesses are still fighting this evolution, manually constraining the algorithm instead of leveraging its strengths. They're optimizing for 2018's reality in 2025's landscape.
Consider me as your business complice.
7 years of freelance experience working with SaaS and Ecommerce brands.
The realization hit me while working with a B2C Shopify client who was struggling with Facebook ad performance. They were spending about €3,000 monthly with a disappointing 2.5 ROAS, using what looked like textbook-perfect custom audience setups.
Their targeting strategy was everything the "experts" recommend: detailed customer personas, layered interest targeting, carefully crafted lookalike audiences, and sophisticated exclusion rules. On paper, it looked professional. In practice, it was limiting their reach to tiny audience pools that were expensive to reach.
The turning point came when I suggested we run a radical experiment. Instead of trying to optimize their complex targeting setup, we'd flip the entire approach: broad targeting with heavy creative focus.
Here's what I discovered that changed everything: their customer journey was much messier than their targeting assumed. A typical customer might see their ad, research on Google, check reviews on Trustpilot, compare with competitors, then come back to purchase weeks later. Traditional attribution was giving Facebook's detailed targeting credit for conversions that were actually driven by organic search and brand awareness.
The breakthrough moment happened when we implemented comprehensive SEO alongside the Facebook campaigns. Suddenly, Facebook's reported ROAS jumped from 2.5 to 8-9. But here's the thing—the Facebook ads hadn't fundamentally improved. What happened was that SEO was driving significant traffic and conversions, but Facebook's attribution model was claiming credit for those organic wins.
This taught me that in today's multi-touchpoint customer journey, trying to control exactly who sees your ads is less important than ensuring your message reaches the right people at scale. The algorithm is actually better at finding your customers than your manual targeting—you just need to feed it the right creative signals.
Here's my playbook
What I ended up doing and the results.
After that eye-opening experience, I completely restructured how I approach Meta advertising. Instead of spending 80% of my time on audience setup and 20% on creative, I flipped it: broad targeting with laser focus on creative testing.
Here's the exact framework I now use for every client campaign:
Campaign Structure (The Foundation)
I start with one main campaign using broad targeting. No detailed interests, no complex lookalikes, no exclusions beyond the obvious (like recent purchasers). Just basic demographics: country, age range, and gender if relevant to the product.
Why does this work? Because Meta's algorithm needs volume to optimize effectively. When you constrain it with tiny custom audiences, you're starving it of the data it needs to find patterns and improve performance.
The Creative Testing Engine
This is where the magic happens. Every single week, without fail, I produce and launch 3 new creative variations. Not 3 new campaigns—3 new creatives within the existing campaign structure.
The creative variations follow specific angles:
Problem-focused creative - Addresses specific pain points your product solves
Lifestyle creative - Shows the product in context of desired lifestyle
Social proof creative - Features customer testimonials, reviews, user-generated content
Educational creative - Teaches something valuable while showcasing the product
Comparison creative - Shows how your product is different/better than alternatives
The Algorithm Training Process
Each creative acts as a signal to the algorithm about who might be interested in your product. A lifestyle-focused creative might attract one segment, while a problem-solving creative attracts another—all within the same campaign structure.
Instead of manually defining these segments through targeting, I let the algorithm learn from actual engagement patterns. It's like having a self-optimizing audience research system that gets smarter with every interaction.
Performance Monitoring and Scaling
I track performance at the creative level, not the audience level. When a creative starts performing well, I don't try to "optimize" it by changing the targeting. Instead, I create variations of that winning creative concept and let the algorithm scale the reach.
The key insight: your creative strategy IS your targeting strategy. The platforms have the data and intelligence—what they need from us is compelling, diverse creative content that can connect with the right people at scale.
Testing Cadence
Launch 3 new creative variations every single week without fail to feed the algorithm fresh signals
Performance Tracking
Monitor results at creative level, not audience level—track which concepts resonate across broad reach
Scaling Method
When a creative wins, create variations of that concept rather than tweaking targeting parameters
Algorithm Partnership
Let Meta's machine learning find your customers instead of manually constraining audience reach
The results from this creative-first approach consistently outperformed traditional targeting methods across multiple client accounts. Instead of fighting the algorithm with complex targeting rules, I started working with it.
The most dramatic example was the Shopify client I mentioned earlier. Within three months of implementing this framework:
Facebook's reported ROAS increased from 2.5 to 8-9 (though this included attribution from organic channels)
Overall conversion rate improved by 40% due to better message-market fit from creative testing
Cost per acquisition decreased by 25% as the algorithm found cheaper, higher-intent traffic
Reach expanded by 3x compared to previous detailed targeting constraints
But here's what really convinced me this approach works: the creative concepts that performed best often attracted audiences we never would have thought to target manually. The algorithm found customer segments that our "perfect" personas missed entirely.
One winning creative was about sustainable materials, which attracted eco-conscious buyers we hadn't specifically targeted. Another focused on convenience, which resonated with busy professionals. The algorithm identified these micro-segments within our broad audience more effectively than our manual targeting ever could.
The timeline typically follows this pattern: Week 1-2 shows initial algorithm learning, Week 3-4 demonstrates improved efficiency, and by Week 6-8, you're seeing consistent performance improvements as the system builds confidence in what works.
What I've learned and the mistakes I've made.
Sharing so you don't make them.
After implementing this approach across dozens of campaigns, here are the most important lessons that shaped my current methodology:
Attribution is broken, but algorithms aren't - Don't trust Facebook's attribution numbers completely, but do trust its ability to find your customers when given broad parameters
Creative fatigue kills campaigns faster than targeting issues - A fresh creative with broad targeting will outperform a stale creative with "perfect" targeting every time
Volume beats precision for algorithm optimization - Small, hyper-targeted audiences starve the algorithm of the data it needs to improve
Your assumptions about your ideal customer are probably wrong - Let the algorithm show you who actually converts, don't force it to conform to your preconceptions
Creative angle diversity is more valuable than audience diversity - One broad audience with multiple creative approaches beats multiple narrow audiences with similar creative
Multi-touchpoint journeys make targeting attribution misleading - Customers research across channels, so broad reach with strong creative often wins over targeted precision
Platform algorithms evolve faster than marketing best practices - What worked in 2020 is actively counterproductive in 2025
The biggest mistake I see businesses make is fighting the evolution of these platforms instead of adapting to work with their strengths. Yes, detailed targeting used to work. No, that doesn't mean it still works today.
When this approach works best: Products with broad appeal, sufficient creative production capacity, and businesses willing to let data guide decisions over intuition. When it doesn't: Highly niche B2B products, limited creative resources, or campaigns that require precise timing (like local events).
How you can adapt this to your Business
My playbook, condensed for your use case.
For your SaaS / Startup
Focus on broad job title targeting rather than detailed interests
Test different value proposition angles in creative
Use case studies and feature demos as creative variations
Let algorithm find decision-makers across company sizes
For your Ecommerce store
Start with country + age range only
Test lifestyle vs problem-solving creative angles
Use UGC and social proof in creative rotation
Focus on seasonal and trending creative concepts