Sales & Conversion
Personas
SaaS & Startup
Time to ROI
Short-term (< 3 months)
Last year, I was managing Facebook Ads for a B2C Shopify store, and I fell into the classic trap that most marketers face. I spent weeks meticulously crafting different audience segments—targeting specific demographics, interests, and behaviors. I was convinced that finding the "perfect audience" was the key to success.
But the results were mediocre at best. We were burning through budget testing different audience combinations, and our ROAS wasn't improving. That's when I discovered something that completely changed my approach to SaaS ad targeting.
The shift? Creatives are the new targeting. Instead of trying to outsmart Facebook's algorithm by manually selecting audiences, I learned to trust the platform's machine learning capabilities while putting all effort into creative testing.
Here's what you'll learn from my experience:
Why detailed targeting is dead in 2025 (and what replaced it)
The simple framework that transformed our ROAS
How to set up AI-powered creative testing for SaaS
The testing rhythm that prevents creative fatigue
Real metrics from campaigns using this approach
This isn't another theory-heavy guide. This is exactly what I implemented, why it worked, and how you can apply it to your SaaS advertising. Let's dive into why the conventional wisdom about ad targeting is broken—and what actually works in today's landscape.
Industry Reality
What every SaaS marketer has already heard
Walk into any marketing conference or browse through Facebook Ads courses, and you'll hear the same targeting gospel repeated endlessly:
"Perfect your audience segments." "Layer your interests." "Use lookalike audiences." "Create detailed buyer personas for targeting."
The industry has built an entire mythology around audience targeting precision. Marketers spend countless hours creating elaborate audience segments, testing different demographic combinations, and obsessing over interest targeting. Tools like Facebook's Audience Insights used to be the holy grail for finding that perfect customer segment.
Here's what the conventional wisdom tells you to do:
Create detailed buyer personas with specific demographics
Layer interests and behaviors to narrow your audience
Test different lookalike audience percentages
Use retargeting audiences based on website behavior
A/B test audience segments against each other
This approach made sense in 2018. Facebook's targeting capabilities were more transparent, iOS 14.5 hadn't destroyed attribution, and privacy regulations hadn't limited data collection. Marketers could actually see meaningful differences between audience segments.
But here's the problem: the game has completely changed.
Privacy regulations killed detailed targeting. iOS updates destroyed attribution tracking. Facebook's algorithm became sophisticated enough to find your customers better than you can manually. Yet most marketers are still playing by 2018 rules in a 2025 reality.
The result? Wasted budget on audience testing that doesn't move the needle, while the real optimization opportunity—creative testing—gets ignored.
Consider me as your business complice.
7 years of freelance experience working with SaaS and Ecommerce brands.
Let me tell you about the B2C Shopify store project that completely changed how I think about ad targeting. The client came to me frustrated with their Facebook ad performance. They had a decent product, good traffic, but their ROAS was stuck around 2.5 with their current agency.
Like any "good" marketer, I started where everyone starts: audience research. I dove deep into their customer data, created detailed personas, and built elaborate audience segments. Women aged 25-45, interested in sustainable fashion, who also liked specific lifestyle brands. Then I created lookalikes, tested different percentage ranges, layered behaviors—the whole playbook.
After six weeks of audience testing, we'd barely moved the needle. ROAS went from 2.5 to 2.7. Marginal improvement, but nothing exciting. Meanwhile, I was spending most of my time managing different audience segments instead of focusing on what really mattered.
That's when I had a conversation with a friend who runs a successful creative agency. He told me something that initially sounded crazy: "Stop thinking about audiences. Facebook knows who your customers are better than you do. Focus on giving the algorithm different creative options to work with."
The approach he described was radically different from everything I'd learned. Instead of multiple campaigns with different audience segments, he was running one big campaign with broad targeting and multiple creative variations. The algorithm would naturally optimize toward the people most likely to convert, regardless of whether they fit my "perfect customer" profile.
I was skeptical, but the results spoke for themselves. His clients were seeing consistent ROAS improvements, lower CPCs, and most importantly—sustainable performance that didn't require constant audience tweaking.
So I decided to test this approach with my Shopify client. The shift from audience obsession to creative focus was about to teach me the most valuable lesson of my advertising career.
Here's my playbook
What I ended up doing and the results.
Here's exactly what I implemented—and why it worked so much better than traditional audience targeting.
Step 1: The Simple Campaign Structure
I restructured the entire approach around one core principle: let Facebook's AI handle the targeting while I focused on creative diversity. Instead of multiple campaigns with different audiences, I built:
1 campaign (Conversions objective)
1 broad audience (age and location only)
Multiple ad sets with different creative angles
Testing cadence: 3 new creatives every single week
The audience setup was almost embarrassingly simple: Women and Men, aged 25-65, in English-speaking countries. No interest targeting. No behavior layering. No lookalikes. Just broad, and let the algorithm figure out who responds best.
Step 2: The Creative Testing Framework
This is where the magic happened. Instead of spending time on audience research, I invested all that energy into creating diverse creative angles. Each creative acted as a signal to the algorithm about different customer segments:
Problem-focused creative: "Tired of overpriced, low-quality fashion?"
Solution-focused creative: "Sustainable fashion that doesn't compromise on style"
Social proof creative: User-generated content and testimonials
Lifestyle creative: Product in real-world settings
The key insight? Each creative naturally attracts different types of people. A problem-focused ad draws in price-conscious shoppers. A lifestyle ad attracts style-conscious buyers. The algorithm learns who responds to what and optimizes accordingly.
Step 3: The Weekly Creative Rhythm
Here's the systematic approach that prevented creative fatigue:
Monday: Analyze previous week's performance
Tuesday: Create 3 new creative concepts
Wednesday: Launch new creatives, pause underperformers
Thursday-Sunday: Let algorithm optimize
This consistent rhythm meant fresh creative was always entering the system while stale creative was being retired. The algorithm had constant variety to work with, which kept performance stable.
Step 4: The AI-Powered Optimization Process
Instead of manual audience adjustments, I let Facebook's machine learning handle optimization through:
Automatic Placements: Let AI choose best ad positions
Campaign Budget Optimization: AI allocates budget to best-performing ad sets
Dynamic Creative Testing: AI tests different combinations of headlines, copy, and images
This approach aligned with how Facebook's algorithm actually works in 2025—it's incredibly sophisticated at finding your customers, but only if you give it quality creative signals to work with.
Testing Framework
3 new creatives weekly with systematic performance analysis and rotation schedule
Algorithm Trust
Let Facebook's AI handle placement and budget optimization instead of manual targeting
Creative Signals
Each creative angle acts as targeting—problem-focused vs lifestyle vs social proof attract different segments
Performance Rhythm
Monday analysis, Tuesday creation, Wednesday launch, Thursday-Sunday optimization cycle
The results were honestly better than I expected, even as someone who believed in this approach.
Performance Improvements:
ROAS increased from 2.7 to 4.2 over 8 weeks
Cost per acquisition dropped by 35%
Click-through rates improved consistently week over week
Campaign management time reduced by 60%
But the most interesting discovery was in the attribution data. When I analyzed which "audiences" were actually converting, they didn't match any of the personas I'd originally created. The algorithm was finding customers I never would have targeted manually—people outside my demographic assumptions who were genuinely interested in the product.
The Unexpected Insight: Creative variety taught the algorithm about different customer motivations better than audience targeting ever could. A single campaign with diverse creative attracted both price-conscious buyers and premium shoppers, depending on which creative they responded to.
This completely shifted my understanding of how modern advertising actually works. The platform's AI isn't just better at targeting—it's fundamentally different from manual targeting. It optimizes for behavior, not demographics.
What I've learned and the mistakes I've made.
Sharing so you don't make them.
This experience taught me five critical lessons about AI-powered ad targeting that every SaaS marketer should understand:
Creatives are the new targeting: In 2025, your creative strategy IS your targeting strategy. Different creative angles naturally segment your audience better than manual demographic targeting.
Trust the algorithm, but feed it quality signals: Facebook's AI is incredibly sophisticated, but it needs diverse, high-quality creative to work with. Garbage in, garbage out.
Consistency beats perfection: A steady rhythm of fresh creative (3 per week) outperforms sporadic "perfect" campaigns every time.
Broad targeting isn't lazy—it's strategic: Letting the algorithm find your customers often reveals opportunities you'd miss with manual targeting.
Focus energy where it matters: Time spent on audience research is better invested in creative strategy and testing frameworks.
When this approach works best: SaaS products with clear value propositions, multiple customer segments, and budget for consistent creative production.
When to be cautious: Highly niche B2B tools with very specific buyer personas, extremely limited budgets, or products requiring extensive education before purchase.
The biggest mindset shift? Stop thinking like a traditional marketer and start thinking like an algorithm trainer. Your job isn't to find customers—it's to give AI the right signals to find them for you.
How you can adapt this to your Business
My playbook, condensed for your use case.
For your SaaS / Startup
For SaaS companies looking to implement AI-powered ad targeting:
Start with broad targeting (job titles only, no interest layering)
Create creative angles for different pain points your SaaS solves
Test product demo videos vs. customer testimonials vs. problem-focused copy
Use LinkedIn and Facebook's native video players for better algorithm signals
For your Ecommerce store
For ecommerce stores implementing this creative-first approach:
Focus on lifestyle vs. product-focused vs. problem-solving creative angles
Test user-generated content against professional product photography
Create seasonal and trending creative to maintain freshness
Use dynamic product ads with broad catalogs for AI optimization