Sales & Conversion
Personas
Ecommerce
Time to ROI
Short-term (< 3 months)
Last year, I was burning through Facebook ad budgets faster than a startup burns through venture capital. My B2C Shopify client was pulling $50+ cost per acquisition numbers that made absolutely no sense for their average order value. Every "expert" was telling us to optimize audiences, test lookalikes, and dive deeper into detailed targeting.
Here's what nobody talks about: Facebook's algorithm fundamentally changed in 2024, and most marketers are still playing by 2022 rules. While everyone's obsessing over audience targeting, the platform has shifted to a creative-first approach that most businesses completely ignore.
After working with this client for three months and testing over 200 creative variations, we managed to drop their CPA from $50 to $18 while actually increasing their conversion volume. The kicker? We did it by throwing conventional Facebook wisdom out the window.
Here's what you'll learn from this breakdown:
Why audience targeting is dead (and what replaced it)
The 3-creative-per-week system that killed our CPA
How to structure campaigns for Facebook's 2024 algorithm
The creative testing framework that works for any budget
Real numbers from our campaign optimization
If you're tired of watching your Facebook spend disappear into the void, this playbook will show you exactly how to work with the platform instead of against it. Check out our ecommerce playbooks for more conversion strategies.
Strategy Shift
The Death of Detailed Targeting
Every Facebook ads course from 2019-2023 taught the same gospel: detailed targeting is king. Create laser-focused audiences based on interests, behaviors, and demographics. Test lookalike audiences. Exclude past customers. Layer interests for "super targeted" audiences.
Here's what the industry typically recommends for reducing CPA:
Create highly specific audience segments (photography enthusiasts who live in urban areas and have high income)
Test multiple lookalike percentages (1%, 2%, 5%, 10%)
Use detailed exclusions to "clean" your audiences
Optimize for specific conversion events (add to cart, initiate checkout)
Run separate campaigns for different funnel stages
This conventional wisdom exists because it used to work. When Facebook had access to detailed third-party data and iOS wasn't blocking tracking, micro-targeting was incredibly effective. You could genuinely find your exact customer and serve them perfect ads.
But here's where it falls short in 2024: Privacy regulations killed detailed targeting. iOS 14.5+ destroyed attribution. Facebook's machine learning now relies on creative signals more than audience data. The platform literally tells you to use broad audiences, yet most marketers ignore this advice.
When you're still trying to out-smart Facebook's algorithm with complex audience strategies, you're fighting a machine that already knows who wants to buy your product better than you do. The real optimization opportunity moved to creative testing and campaign structure.
Consider me as your business complice.
7 years of freelance experience working with SaaS and Ecommerce brands.
My client ran a B2C fashion accessories store with a pretty straightforward problem: their Facebook ads were hemorrhaging money. $50+ cost per acquisition with an average order value around $80. The math wasn't mathing, as the kids say.
This wasn't some massive enterprise operation. They were doing maybe $30K monthly revenue, spending $8K on Facebook ads, and getting roughly 160 customers per month. Not terrible volume, but those unit economics were brutal for a small ecommerce store.
The typical agency approach we tried first (and why it failed):
Like any rational person, I started with conventional wisdom. We spent two months testing different audience combinations:
Lookalike audiences based on purchasers (1%, 2%, 5%)
Interest-based targeting (fashion, accessories, competing brands)
Behavioral targeting (online shoppers, frequent travelers)
Retargeting campaigns with complex exclusion logic
Results? CPA went from $50 to... $48. Barely moved. We were optimizing deck chairs on the Titanic.
The frustrating part was that the ads were getting engagement. People were clicking, visiting the site, even adding items to cart. But something in the funnel wasn't converting these clicks into profitable customers.
That's when I realized we were solving the wrong problem. Everyone talks about audience optimization, but nobody talks about the fact that Facebook's algorithm has fundamentally changed how it delivers ads. The platform now prioritizes creative performance over audience targeting.
This client became my testing ground for a completely different approach—one that focused on creative velocity instead of audience sophistication. The results changed everything I thought I knew about Facebook advertising.
Here's my playbook
What I ended up doing and the results.
Here's the system that dropped our CPA from $50 to $18 in three months, and it's the opposite of everything you've been taught about Facebook ads.
The Creative-First Campaign Structure
Instead of running multiple campaigns with different audiences, we consolidated everything into one campaign with broad targeting and multiple ad sets focused purely on creative testing:
1 Campaign: Broad audience (women 25-55, no additional targeting)
Multiple Ad Sets: Each featuring different creative angles
Testing Cadence: 3 new creatives every single week
The magic wasn't in the structure—it was in how we approached creative development. Instead of trying to create "perfect" ads, we focused on rapid iteration and let Facebook's algorithm tell us what worked.
The 3-Creative-Per-Week System
Every Monday, we launched three new creative variations testing different angles:
Week 1 Example:
Lifestyle-focused (showing product in use)
Problem-solving (addressing specific pain points)
Social proof (customer testimonials/reviews)
Week 2 Example:
Comparison-based (vs. competitor products)
Seasonal relevance (limited-time offers)
Behind-the-scenes (product creation/quality)
Each creative ran for exactly 7 days. If CPA was above $25 after day 3, we killed it. If it hit profitability by day 7, it graduated to our "winner" campaign for scaling.
The Algorithm Feeding Strategy
Here's where it gets interesting. Facebook's machine learning thrives on fresh signals. By constantly feeding it new creative approaches, we were essentially training the algorithm to find people who responded to different value propositions.
A lifestyle creative might attract aspirational buyers. A problem-solving creative might convert people with specific pain points. A social proof creative might persuade skeptical shoppers. Instead of trying to predict which audience wanted which message, we let Facebook figure it out.
The Creative Development Process
Creating 3 new ads weekly sounds overwhelming, but we systematized it:
Monday Morning: Analyze previous week's performance
Monday Afternoon: Brainstorm new angles based on data
Tuesday-Wednesday: Create new creatives (video/image + copy)
Thursday: Launch new ads, pause underperformers
Friday: Document learnings and plan next week
The client provided product footage and images. I handled creative direction and copywriting. We used simple tools—Canva for static images, basic video editing for short clips. Nothing fancy.
Most importantly, we stopped trying to create "perfect" ads and started embracing rapid testing. Some weeks, all three creatives flopped. Other weeks, we'd hit gold with an unexpected angle that outperformed everything.
Creative Velocity
Test 3 new ad variations weekly rather than optimizing audiences endlessly
Broad Targeting
Use minimal audience constraints and let Facebook's algorithm find your buyers
Data-Driven Killing
Pause underperforming creatives within 72 hours to minimize wasted spend
Winner Scaling
Graduate profitable creatives to dedicated scaling campaigns for maximum reach
The transformation happened gradually, then suddenly. Month one, we saw CPA drop from $50 to $38—not dramatic, but moving in the right direction. Month two brought us to $28 CPA, and by month three, we consistently hit $18-22 CPA.
The numbers that mattered:
CPA dropped from $50 to $18 (64% improvement)
Monthly ad spend stayed at $8K but generated 440+ customers vs. 160
ROAS improved from 1.6 to 4.4
Time spent on campaign optimization decreased by 70%
But here's what really surprised us: the creative angles that performed best weren't what we expected. Our highest-converting ad was a simple product demonstration video that we almost didn't run because it felt "too basic." Meanwhile, our most polished, professionally-shot lifestyle content consistently underperformed.
Facebook's algorithm rewarded authenticity and clear value communication over production quality. The platform could identify people who responded to straightforward product demos better than we could identify audiences manually.
The approach also solved our attribution problem. Instead of trying to track complex customer journeys across devices, we focused on creative performance within Facebook's native reporting. If an ad generated profitable conversions, we scaled it. If not, we killed it and moved on.
What I've learned and the mistakes I've made.
Sharing so you don't make them.
Here are the key lessons from three months of creative-first Facebook advertising:
1. Algorithm Partnership Over Audience Outsmarting
Facebook's machine learning is genuinely better at finding buyers than manual targeting. Your job isn't to tell Facebook who to target—it's to give the algorithm compelling creative signals to work with.
2. Creative Fatigue is Real and Fast
Even winning creatives lose effectiveness within 2-3 weeks. Build creative production into your ongoing process, not as a one-time project.
3. Broad Targeting Requires Creative Precision
When you target everyone, your creative has to do the segmentation work. Each ad needs to speak to a specific mindset or need state.
4. Testing Velocity Beats Testing Complexity
Running 3 simple tests weekly outperforms 1 complex test monthly. Facebook rewards fresh content and frequent iteration.
5. Production Quality ≠ Performance Quality
Our best-performing ads were often the simplest ones. Clear value proposition beats cinematic production every time.
6. Data-Driven Creativity is Possible
Use performance data to inform creative direction. If problem-focused ads outperform lifestyle ads, lean into problem-solving angles.
7. Platform-Native Content Wins
Ads that look like native Facebook content (user-generated style) typically outperform obvious advertisements.
This approach works best for businesses with sufficient creative resources and budget for regular testing. It's less suitable for businesses that can only create ads monthly or have extremely limited creative assets.
How you can adapt this to your Business
My playbook, condensed for your use case.
For your SaaS / Startup
For SaaS startups implementing this creative-first approach:
Focus on problem/solution creative angles rather than lifestyle content
Test demo videos, feature comparisons, and customer success stories
Use screen recordings and simple explainer videos for rapid creative production
Target broad business professional audiences and let creative do the segmentation
For your Ecommerce store
For ecommerce stores applying this framework:
Emphasize product demonstration and social proof in creative rotation
Test seasonal angles, lifestyle contexts, and problem-solving approaches
Use customer photos and videos for authentic, high-performing content
Focus on broad demographic targeting rather than interest-based audiences