Every e-commerce brand runs the same playbook: segment audiences by funnel stage, show dynamic product ads to cart abandoners, offer escalating discounts, watch the ROAS climb. It’s digital marketing’s closest thing to a sure bet.
Here’s the problem: that playbook is quietly bankrupting sophisticated advertisers.
The retargeting strategies that built empires between 2014-2019 have become victim to their own success. What worked brilliantly when you were early has become table stakes-except now you’re competing in the most expensive, oversaturated auction environment Facebook offers.
Let me show you what’s really happening beneath your dashboard metrics, and why the brands winning today are doing the opposite of conventional wisdom.
The Prisoner’s Dilemma of Retargeting
Facebook retargeting worked so well that it created its own problem. When every brand discovered the magic of reaching cart abandoners and site visitors, they all poured budget into the same audiences. This transformed your most valuable users into your most expensive inventory.
Think about the paradox: your retargeting audience represents both your highest-intent customers and your most over-served, ad-fatigued, expensive-to-reach users.
Most brands respond by increasing retargeting budget because attributed ROAS still looks phenomenal. But here’s what they’re missing: they’re confusing correlation with causation. Many of those “retargeting conversions” would have happened anyway. You’re just paying Facebook to take credit for customers already planning to return.
Attribution Is Lying to You
Facebook’s attribution model systematically over-credits retargeting while under-crediting prospecting. This isn’t a conspiracy-it’s a structural limitation of last-touch attribution.
Scenario A: A customer sees your prospecting ad Monday, visits your site, browses for 15 minutes, bookmarks it, then returns Thursday via direct traffic and purchases. Attribution credit: None. That prospecting ad did the heavy lifting, but it’s invisible.
Scenario B: Same journey, except Thursday you hit them with a retargeting ad 30 minutes before they were coming back anyway. Attribution credit: Retargeting. Your ROAS looks amazing.
This is how brands end up with 60-70% of budget in retargeting while wondering why new customer acquisition has stalled. The data tells you to shift more to retargeting, which further compounds the problem.
You’re optimizing based on a lie your attribution model is telling you.
The Customer Experience You’re Not Seeing
Let’s talk about what your “strategic” seven-touch retargeting funnel actually looks like from the customer’s perspective:
- Day 1: They browse your site. Interesting brand.
- Day 2: One retargeting ad. “Oh right, should think about that.”
- Day 3: Another ad. They scroll past it.
- Day 4: Two more ads. Starting to feel pushy.
- Day 5: Three ads showing exact products they viewed, plus your cart abandonment sequence, plus your engagement campaign.
- Day 6: Six impressions. They’re now actively avoiding your ads.
- Day 7: Seven impressions. They’re annoyed at your brand.
Here’s the critical insight most marketers miss: Facebook doesn’t have a cross-campaign frequency cap.
You might be running 4-7 retargeting campaigns simultaneously, each showing “average frequency: 2.3” in reporting. But your customer is seeing 15-25 total impressions per week across all those campaigns combined.
You’re not running a sophisticated funnel. You’re running a brand damage operation while paying premium CPMs for the privilege.
The Discount Death Spiral
Standard practice says to increase discount offers as urgency decreases. Cart abandoners see 10% off. Three-day-old visitors see 15%. Week-old visitors see 20%. This “maximizes conversion rate.”
But examine what you’ve actually built: a customer education program in patience.
Smart shoppers learn your discount cadence fast. Why buy today when a 20% offer is coming in 5-7 days? You’ve trained them to wait, to ignore your initial touches, to expect escalating discounts.
Worse, your precise segmentation means you’re showing 20% off to people who would’ve converted at 10%, or showing any discount to people planning to buy at full price anyway.
Your retargeting campaigns show positive ROAS. But what’s the opportunity cost? What would revenue and margin look like if you hadn’t trained discount-seeking behavior?
Nobody’s measuring the profit you’re leaving on the table.
The iOS 14.5 Shift Nobody Adapted To
The privacy changes of 2021-2023 didn’t just reduce tracking-they fundamentally altered retargeting economics.
Pre-iOS 14.5, retargeting was algorithmic arbitrage. You had data Facebook didn’t (who visited your site), so you could bid more accurately. The algorithm under-priced this advantage, creating extraordinary returns.
Post-iOS 14.5, Facebook lost 30-50% of tracking data but rebuilt their algorithm to compensate through probabilistic modeling and on-platform signals. The algorithm now identifies high-intent users almost as effectively as your retargeting pixel-but at lower CPMs because there’s less competition.
The edge you gained from retargeting audiences has compressed dramatically. Yet most brands still allocate budgets based on 2019 economics, watching efficiency decline quarter after quarter.
The Test You’re Afraid to Run
Want to know the actual incremental value of your retargeting spend?
Take your site visitors from the last 3 days and randomly split them in half. Retarget one group with your full strategy. Exclude the other group from all advertising. Wait 30 days and compare purchase rates.
Most brands refuse this test because they suspect the truth: retargeting lift is often 15-30%, not the 300-800% that attributed ROAS suggests.
If you’re spending $100K monthly on retargeting with an 8x attributed ROAS, but only generating 20% incremental lift, you’re paying $100K for $20K in incremental revenue. That’s 0.2x actual ROAS, not 8x.
The attributed ROAS isn’t wrong-those people really did convert after seeing your ads. But 80% would have converted anyway. You’re paying Facebook to take credit.
The Contrarian Strategy: Retargeting Minimalism
Here’s what almost nobody’s testing, but the smartest performance marketers are quietly implementing:
Radically simplify retargeting.
Instead of seven campaigns with intricate segmentation:
- One retargeting campaign
- 7-day window maximum
- 3 impressions per week cap (monitored across all campaigns, not per campaign)
- No discount escalation
- Budget capped at 20-25% of total spend
Take that liberated budget and reinvest in prospecting with longer attribution windows and holdout testing to measure true incrementality.
This approach feels terrifying. Your immediate attributed ROAS will drop from 8x to maybe 5x. But here’s what actually happens:
- Total revenue stays flat or increases (you’re not cannibalizing organic returns)
- Margin improves dramatically (you’re not over-discounting)
- Customer lifetime value increases (you haven’t trained discount-seeking)
- Prospecting efficiency improves (the algorithm has more signal)
- Brand perception strengthens (customers aren’t ad-fatigued)
The brands testing this are seeing something remarkable: lower attributed ROAS but higher actual profit.
The Real Cost of “Efficiency”
Here’s what ties this together: by over-indexing on retargeting efficiency, you’re systematically underinvesting in growth.
Retargeting is defensive spending. It captures existing demand and protects against competitors. Important functions, but not growth functions.
Prospecting is offensive spending. It creates new demand and expands your addressable market. It looks less efficient in attribution because it does the hard work of customer acquisition.
When you allocate 60-70% of budget to retargeting because it “performs better,” you’re creating a business that looks efficient while starving growth. You’re optimizing into a local maximum while missing the global maximum.
The brands winning the next era of Facebook advertising aren’t the ones with the most sophisticated retargeting funnels. They’re the ones who recognized that yesterday’s edge became today’s table stakes.
Questions You Should Be Asking
If you’re running e-commerce Facebook ads, here are the uncomfortable questions worth confronting:
1. What percentage of your retargeting conversions are actually incremental?
If you haven’t run holdout tests, you’re guessing.
2. What’s your true cross-campaign frequency?
If you’re only looking at individual campaign metrics, you’re blind to customer experience.
3. Have you trained customers to wait for discounts?
If your retargeting involves escalating offers, you probably have.
4. Is your attribution window aligned with your purchase cycle?
If you’re using 7-day click attribution for a product with a 21-day consideration window, you’re undervaluing prospecting.
5. What would happen if you cut retargeting budget by 50%?
If you’re scared to test this, you don’t actually understand your business mechanics.
The Path Forward
Retargeting became so successful that it stopped working the way we think it works. It’s still valuable-but nowhere near as valuable as attribution suggests, and probably half as valuable as three years ago.
The uncomfortable truth is that most e-commerce brands are over-retargeting by a factor of 2-3x while under-prospecting by the same margin. They’re doing this because their attribution model is systematically lying about what’s actually driving growth.
The brands that thrive over the next 24 months will be the ones who recognize this shift early. The ones clinging to 2019 best practices will watch efficiency slowly erode, wondering why the playbook stopped working.
The smartest move in e-commerce Facebook advertising right now isn’t optimizing your retargeting funnel. It’s questioning whether you need most of it at all.
Start with one simple test: cut your retargeting budget by 30% for one month and reallocate it to prospecting. Watch what happens to total revenue, not just attributed ROAS.
You might be surprised to discover that less retargeting equals more profit.
And if that scares you? That’s exactly why you should run the test. The strategies that feel safest are often the ones quietly destroying value while your dashboard tells you everything’s fine.
