Everyone’s celebrating their 5x ROAS. Meanwhile, their business isn’t growing.
Welcome to the biggest con in digital marketing, where platforms claim credit for sales that were happening anyway, and you’re too busy optimising toggles to notice.
The game changed while you were busy optimising
Remember when media buying was about finding the hidden settings Google didn’t want you to know about? The sneaky placement options, the default toggles designed to bleed budgets, those mysteriously expensive mobile app clicks nobody actually wanted?
Those were simpler times. We’d dive into Google Ads like detectives, uncovering wasted spend in search partners, display network placements, and random app inventory. Find the leak, plug it, watch ROAS improve. Job done, right?
That game is over. Not changing, not evolving. Over.
While you’ve been hunting for micro-optimisations inside platforms, the real money drain has moved somewhere else entirely. It’s not in your settings anymore. It’s in the space between platforms, where Google and Meta are both claiming credit for the same bloody sale, and you’re paying for it twice without even realising.
The double-dip economy (and you’re funding it)
Here’s what’s actually happening thousands of times a day. Sarah sees your product on Instagram during her lunch break. She likes it, maybe even saves the post, but doesn’t buy. Later that evening, she Googles your brand name to find your website. Clicks your Shopping ad. Makes the purchase.
Meta celebrates: “We drove that sale! View-through attribution!” Google celebrates: “We drove that sale! Last-click attribution!” You celebrate: “Wow, both platforms are crushing it!”
Your accountant, meanwhile, is wondering why revenue isn’t actually growing despite all this apparent success.
You just paid twice for Sarah. And you’re celebrating it.
Performance Max: The wolf in sheep’s clothing
Let’s talk about Google’s Performance Max, the algorithm everyone’s raving about. Want to know what it really is? A sophisticated retargeting engine wearing a growth costume.
Performance Max doesn’t find new customers. It hunts warm traffic. Your email subscribers googling your brand. Your Meta traffic searching for reviews. Your returning customers who were going to buy anyway. It’s genius, really. Google created a black-box campaign type that hides what search terms it’s targeting, won’t tell you which placements are working, claims credit for brand searches, targets your warm audiences first, then tells you it’s “finding new customers.”
The result? Spectacular ROAS on paper. Zero incremental growth in reality.
The uncomfortable maths nobody wants to do
Let’s say you spent $100,000 last month across Google and Meta. Your dashboards are glowing with success. Meta shows 4x ROAS, claiming $400,000 in revenue. Google shows 3.5x ROAS, claiming $350,000. Total claimed revenue: $750,000.
But here’s the kicker. Your actual Shopify revenue for the month? $250,000.
Where’s the other $500,000? It doesn’t exist. It’s the same $250,000 being counted multiple times across platforms, dressed up as different conversions. But you’re making strategic decisions based on $750,000 worth of phantom performance.
This is why businesses with “amazing ROAS” are going broke. The maths looks beautiful until you realise it’s fiction.
The returning customer scam
Here’s the part that should make you properly angry. Most of your ad spend is going towards people who already bought from you. These returning customers should be coming back through email (free), SMS (nearly free), direct traffic (free), or organic brand search (also free).
Instead, they’re coming back through Google Shopping ads, Meta retargeting, Performance Max, and YouTube ads. All paid channels. You’re essentially paying rent on customers you already own, and the platforms have designed their systems to make this happen.
Think about that for a second. You’re paying to reacquire customers who were already going to buy from you again. It’s like paying someone to walk through your own front door.
Why platforms love the confusion
Google and Meta aren’t really competitors anymore. They’re co-conspirators in the great attribution shell game. They’ve realised something brilliant: as long as both platforms look good in their own dashboards, you’ll keep spending on both. They don’t need to prove incremental value. They just need to look valuable.
Meta’s machine learning is optimised to find people likely to buy. Not to create buyers, but to find them. There’s a massive difference that most marketers miss. Google’s Smart Bidding does exactly the same thing. It doesn’t create demand; it harvests existing demand and charges you for the privilege of accessing customers who were already interested.
Cue the uncomfortable realisation that you’re not really doing marketing anymore, you’re just bidding on inevitable sales.
The new rules of media buying
The old playbook doesn’t work anymore. In-platform metrics have become meaningless vanity numbers that tell you nothing about real business growth. What actually matters? New customer acquisition cost, incremental revenue (sales that wouldn’t have happened without ads), profit after all costs, and whether your customer lifetime value still makes sense against these inflated acquisition costs.
Here’s the brutal test: turn off Performance Max for two weeks. What happens to revenue? If it barely moves, you’ve been paying for sales that would’ve happened anyway. Run geo-experiments where you spend heavily in one city and nothing in another similar city. The difference? That’s your actual incremental impact, not the inflated numbers platforms are showing you.
If you’re not religiously separating new from returning customer performance, you’re flying blind. Most “successful” campaigns are just expensive retargeting to existing customers dressed up as growth. Your analytics setup needs to track first purchase source, actual customer journeys (not last-click fantasies), true new customer rates by channel, and how many returning customers would have come back organically.
And please, stop letting platforms mark their own homework. Use server-side tracking, send them profit data instead of revenue, and tell them which customers are actually new. If you’re using platform conversion data to optimise platform campaigns, you’re essentially asking the fox to guard the henhouse.
What to actually do about this mess
First, build a source of truth outside the platforms. You need unified measurement that doesn’t rely on platform reporting. This means server-side tracking for all conversions, custom attribution modelling based on your actual business (not Google’s black box), clear identification of new versus returning customers, and profit-based reporting that shows real margins, not fantasy revenue.
The only way to know if your ads actually work is to turn them off. Scary? Absolutely. Necessary? Even more so. Run proper holdout tests. Establish a baseline with everything running for two weeks, then turn off suspected redundant campaigns for the next two weeks. Measure the actual impact on real revenue, then make decisions based on incrementality, not platform-reported ROAS.
Stop structuring campaigns the way platforms want you to. Structure them for clarity instead. Separate your high-intent campaigns (brand search, retargeting, customer lists, cart abandoners) from your low-intent campaigns (cold prospecting, broad targeting, new creative formats, unbranded search). Monitor these separately. The first group should have minimal spend because they’re probably cannibalising organic sales. The second group is where actual growth happens.
The platform arbitrage opportunity
While everyone else is chasing fake ROAS, there’s a massive opportunity staring you in the face. Build channels the platforms can’t claim credit for. SEO that drives direct traffic, email that bypasses paid touchpoints, content that builds genuine brand affinity, and community that creates organic advocacy.
Every customer you acquire outside the platform duopoly is worth two inside it, because you’re not paying for them twice. And unlike platform-dependent acquisition, these customers actually belong to you.
Your 30-day cross-channel audit
Week one is about establishing reality. Pull your actual revenue from Shopify or whatever your source of truth is. Compare it to the total claimed revenue across all platforms. Calculate your double-counting rate. Fair warning: it’ll shock you.
Week two is for ruthless segmentation. Identify new versus returning customers properly. Track their actual first touchpoint, not the platform’s claimed attribution. Calculate your true new customer acquisition cost using real data, not platform fantasies.
Week three is when you test incrementality. Turn off Performance Max. Pause the bottom 50% of your Meta campaigns. Document what actually changes in your real revenue. Spoiler: probably not much.
Week four is when you rebuild for reality. Restructure your campaigns based on actual incrementality. Adjust budgets based on true contribution to growth. Set up proper measurement that’ll keep you honest going forward.
The uncomfortable truth about tomorrow
The platforms are getting smarter, and their ability to claim credit for organic sales is improving daily. Performance Max today, AI Mode tomorrow, something even more opaque next year. If you don’t build measurement systems and channels outside their ecosystem now, you’ll wake up one day paying for every single sale, even the ones that would’ve happened if you’d never advertised at all.
Three questions that matter
If you turned off all ads tomorrow, what percentage of sales would actually disappear? Be honest.
How much are you paying to reacquire customers who would come back anyway through organic channels?
What would happen to your business if you moved 50% of your budget to channels you actually control?
If you don’t know the answers, you’re not doing marketing. You’re just feeding the machine. And the machine is getting hungrier every quarter.
The bottom line
The game has changed fundamentally. It’s not about optimising within platforms anymore. It’s about understanding the entire ecosystem, prioritising incrementality over attribution, focusing on profit over ROAS, and most importantly, building a business that doesn’t need to pay for every sale twice.
Because while you’re celebrating your platform metrics and attending another webinar on Performance Max best practices, your competitors are building real growth engines that don’t depend on algorithmic charity. And when the music stops, when CAC finally exceeds LTV for the last time, they’ll still be standing.
Will you?
Ready to see what your ads are really doing? Get a proper cross-channel audit that shows you where your money’s actually going, and more importantly, where it should be going instead.
