Picture this. You run a Shopify store doing $40K per month in revenue. You spend $8,000 on Meta and Google ads. Your dashboard says ROAS is 3.5x. Life feels good.
But what if $2,300 of that $8,000 isn't doing anything? What if those dollars are buying customers who were going to purchase anyway, or clicking on ads that stopped working weeks ago?
That's not a hypothetical. It's a realistic estimate for a typical store, and it's measurable.
The number nobody talks about
This 29% is our estimate, not a measured universal, but the independent research points the same way. In eBay's Econometrica field experiment, branded-search ads turned out to be roughly 99% non-incremental: turning them off barely moved sales. Stella's 2025 analysis of 225 geo-incrementality tests found branded search returns just $0.70 per dollar (below breakeven) with about 17% of all tests below the 1.0x line. The only way to know your store's number is a holdout experiment on your own data.
For a merchant spending $8,000 per month, an estimated 29% is about $2,320, every month. On that estimate, that's roughly $27,840 a year buying conversions that may never have been incremental.
The platforms don't tell you this. Why would they? Meta's business model depends on you believing every reported conversion was caused by the ad. Google's incentive is identical. They're not lying. Their attribution models are just designed to take credit, not prove causation.
Where the waste hides
Ad waste isn't one problem. It's three problems wearing a trenchcoat.
Creative fatigue: the slow leak
Every ad has a lifecycle. It launches, finds its audience, hits peak performance, then slowly dies. The problem is that most merchants don't monitor this decline. An ad that peaked at 2.5% CTR three weeks ago might be running at 1.1% today, but it's still spending $50 per day.
The math is brutal. If your creative's efficiency dropped 56% but your spend stayed flat, 56% of that daily budget is waste. Over a month, that single fatigued creative could cost you $840 in lost efficiency.
Multiply that across five or ten active creatives, and you start to see how quickly it adds up.
The fix isn't complicated: rotate creatives, test variants, kill underperformers. But you can't fix what you can't see. And most merchants can't see it because Meta's dashboard doesn't flag efficiency decay. It just shows impressions and spend.
If you want to check your own creatives right now, our Creative Fatigue Checker will tell you which ones to scale, fix, or kill.
Brand cannibalization: paying for customers you already have
This one is emotionally hard to accept. Your brand search campaign probably shows the highest ROAS in your entire account: 8x, 10x, maybe higher. It feels like your best campaign. It might actually be your worst investment.
Here's why. When someone searches "your brand name" on Google, they already know who you are. They're not discovering you through the ad. They're navigating to your store, and Google is charging you for the privilege.
Not all brand search spend is wasted. There are legitimate defensive reasons to bid on your own name (preventing competitors from poaching your traffic). But a significant portion of those "conversions" would have happened organically. The question is: how much?
That's what makes this hard. You can't just turn off brand search and see what happens. That's too risky. What you need is a controlled experiment that measures the actual incremental value. We'll get to that.
Channel overlap: when both platforms claim the same sale
This is the most straightforward waste type, and possibly the most expensive.
Here's a real pattern. A merchant has 120 conversions reported by Meta and 95 by Google. Total reported: 215. Actual Shopify orders in the same period: 147.
That's 68 phantom conversions: orders that both platforms claimed but only happened once. The merchant's true blended ROAS is significantly lower than either platform shows individually.
Typical overlap ranges from 15% to 40% of total reported conversions, depending on how much you spend on both platforms and how your audiences intersect. If you want to see your estimated overlap, try our Ad Waste Calculator.
Why your dashboard can't tell you this
Meta Ads Manager and Google Ads both have a fundamental conflict of interest: they report on their own performance. Neither platform has an incentive to say, "Actually, that conversion would have happened without us."
This isn't malicious. It's structural. Each platform uses its own attribution model (Meta uses a 7-day click, 1-day view window by default, Google uses data-driven attribution) and neither coordinates with the other. They're both telling the truth according to their model. The problem is that their models are designed to maximize claimed conversions, not to show you reality.
The only way to know for certain whether an ad caused a sale is to run a controlled experiment. Turn off the ads for a sample of your audience. Measure what happens. If sales don't drop, the ads weren't driving them.
That's called an incrementality test, and until recently, only enterprise brands with $50K+ monthly ad budgets and dedicated data science teams could run them. More on that in our guide to holdout experiments for Shopify merchants.
What you can do today
You don't need to wait for a formal incrementality test to start finding waste. Here are three things you can do right now:
Audit your creative performance. Pull CTR data for every active ad. Compare current CTR to peak CTR. Anything that's declined more than 30% from peak is a candidate for replacement. Our Creative Fatigue Checker automates this math.
Question your brand search ROAS. If your brand campaigns show 8x+ ROAS, be suspicious. That number is high because those customers were already coming to you. Consider a small-scale test: lower your brand bid by 20% for two weeks and see if organic traffic compensates.
Compare platform totals to Shopify. Add up Meta's reported conversions and Google's reported conversions for a given week. Then look at actual Shopify orders. If the platform total is 30%+ higher than Shopify, you have significant overlap. Our True ROAS Calculator can estimate the gap.
The 29% problem is solvable
The good news: wasted ad spend isn't a mystery. The analytics industry has known about these problems for years. Enterprise brands solve them with holdout experiments, media mix models, and dedicated measurement teams.
The bad news: those solutions have been out of reach for Shopify merchants spending $3K–$50K per month. The tools have reportedly cost $24,000–$60,000 per year. The expertise requires a data scientist. The experiments require direct API access to ad platforms.
That's starting to change. Self-serve incrementality testing is becoming possible at a fraction of the enterprise price. The methodology is the same (holdout experiments with statistical significance) but the interface is designed for merchants, not analysts.
Whether you use Ripplux or another tool or a spreadsheet and manual testing, the point is the same: stop trusting platform-reported ROAS as gospel. Your ads might be working. But 29% of them probably aren't. Isn't it worth finding out which 29%?
Keep reading
- Your ROAS Is Lying to You: the full manifesto on the three inflation engines behind the 29% number
- Free Ad Waste Calculator: plug in your spend and see your estimated waste range
- How Ripplux Works: the methodology behind detecting waste and running holdout experiments
- Brand Cannibalization: why your highest-ROAS campaign might be your worst investment
- Creative Fatigue: the slow leak that compounds across your entire creative portfolio
- See the founding offer: while the cohort is open, founding stores lock Pro at $99/mo for life