Why 60% of Your 'Meta Sales' Aren't Real: A Plain-English Guide to View-Through Inflation

Picture two founders looking at the exact same Meta ad account on the exact same Monday morning.

The first one opens Ads Manager, sees a 4.1x ROAS sitting at the top of the column, and feels good. The ads are working. Time to push more budget in. The second one sees that same 4.1x, then opens Shopify in another tab, looks at what actually got deposited last week, and goes quiet. Because the bank account is telling a very different story to the dashboard.

Same account. Same number on the screen. Two completely different decisions about to get made.

The gap between those two founders is almost always one thing: view-through conversions. And until you understand how much of your reported Meta revenue is being claimed rather than caused, you're flying on an instrument that's reading high.

What a view-through conversion actually is

Here's the plain-English version. A click conversion is someone who saw your ad, clicked it, and bought. Fair enough. Meta did its job. You can argue about the window, but the causal chain is at least visible.

A view-through conversion is someone who was shown your ad, didn't click, scrolled past, and then bought within a set window anyway. By default Meta counts that as one of its sales. The person never touched the ad. They might have been heading to your site already. They might have seen it on a podcast, in their inbox, from a mate. Meta saw an impression go past their eyeballs, then saw a purchase, and drew a line between the two.

Sometimes that line is real. A lot of the time it isn't.

The problem is that view-through sales get bundled into the same headline ROAS as your hard-won click sales. So the number at the top of the column is a blend of "we caused this" and "we happened to be in the room when this happened". And nobody tells you the split unless you go looking.

How big the inflation actually gets

This is the part that catches people out. View-through isn't a rounding error. On some accounts I've looked at, it's the majority of the reported number.

I've seen accounts where 50 to 60% of the claimed Meta sales were view-through. Think about what that means. If the dashboard says you did A$100k in Meta revenue and 55% of it is view-through, only about A$45k of that traces back to someone actually clicking. The rest is Meta putting its hand up for sales your Shopify revenue would likely have booked regardless.

To put this into perspective: a founder reading 4.1x in the dashboard might genuinely be running closer to 2.0x once you strip the view sales that weren't incremental. That's not a small correction. That's the difference between an account you scale and an account you fix.

And here's the tell that confirms it. When view-through is doing the heavy lifting, you can pour another A$10k a month into Meta, watch the reported ROAS hold steady, and watch your actual Shopify revenue stay almost flat. The dashboard keeps claiming sales. The bank doesn't see new ones. That's the signature of inflation, and it's the single most common reason a "profitable" account quietly isn't.

Why this got worse, not better

You'd think tracking would have improved over the years. For attribution honesty, the opposite happened.

After the iOS privacy changes, Meta lost a chunk of the clean click signal it used to have. So it leaned harder on modelling and on the conversions it could still claim, which includes view-through. The platform has every incentive to attribute generously. More attributed sales make the ads look more effective, which keeps you spending. I'm not saying that's a conspiracy. I'm saying you should never expect the house to mark its own homework conservatively.

Which is why the only number I fully trust is the one I can reconcile against money that actually landed.

The four-step reconciliation I'd run this week

You don't need a A$2k-a-month measurement tool to do this. You need Ads Manager, your Shopify analytics, and about twenty minutes. Here's the method.

1. Find out how much of your number is view-through. In Ads Manager, go to the columns dropdown and choose "compare attribution settings". Add a 1-day-view column next to your 7-day-click column. Now you can see, side by side, how many purchases are coming from clicks versus views. If the view portion is small, breathe easy. If it's a big slice of the total, keep reading.

2. Set your own threshold and hold to it. My rule of thumb: if view-through is more than about 20 to 30% of your conversions, that's too much to leave in the headline number. Below that, I'll usually let it sit. Above it, I want to see the account run on click-only attribution so the optimisation and the reporting are both honest about what's actually being driven.

3. Reconcile the platform total against demand sales. Take what Meta claims it drove last week and hold it up against your Shopify demand revenue for the same window, minus anything you know came from other channels and email. Watch what happens to that gap as you change spend. If you lift Meta budget 20% and demand sales barely move, the extra spend is buying you claimed sales, not real ones. The bank account is the referee here, not the dashboard.

4. Strip view and decide. Once you've seen the split, make the call. For most 6-7 figure brands where view-through is inflated, I'd run prospecting on click-only and judge it on the click number. You'll feel poorer for a day. You'll also be making budget decisions on a number that means something, which is worth more than a flattering one that doesn't.

That's the whole exercise. It's not clever. It's just the discipline most people skip because the dashboard number is right there and it feels official.

The newer setting worth knowing about

There's a more recent option that's genuinely useful here, and it's worth understanding because it changes the conversation a little.

Meta rolled out an "incremental attribution" setting (it landed around April 2025). Instead of counting every click and view inside a fixed window, it uses a model trained on a huge bank of lift studies to predict whether a given conversion was actually caused by the ad. Conversions it reckons would have happened anyway don't count. You set it at the ad-set level in the conversion section, and you can still view your old 7-day-click and 1-day-view numbers side by side for comparison.

When you turn it on, the incremental column reads lower than your standard one. Sometimes dramatically. I've seen an ad set show a reported 15x ROAS sitting next to a 2.8x incremental figure on the same spend. The incremental number is narrow on purpose. It's only counting the sales Meta believes it genuinely made happen.

Here's the part that's easy to miss in the pitch. Meta's own director of signal growth says that optimising toward incremental attribution drove a roughly 24% lift in incremental conversions versus the standard model in their testing. That's their number, on their setting, so treat it as a claim rather than a promise. But the logic holds up: if you tell the algorithm to chase sales it actually causes, it gets better at finding them, instead of getting better at standing next to people who were going to buy anyway.

My take? I wouldn't flip it on blind across the whole account. I'd treat it the way I'd treat any meaningful change. Turn it on for one campaign, compare the incremental column against your reconciled Shopify number for a couple of weeks, and see whether the two start to agree. If they do, you've found a reporting view you can trust. Roll it out from there.

What I really like about the setting is quieter than the ROAS lift. It makes view-through inflation visible inside the platform itself, in a column you can read any time, instead of something you only catch when you go reconcile by hand.

What this changes about how you read the account

Once you've done this once, you can't unsee it. The headline ROAS stops being "the result" and becomes "the optimistic version of the result". You start asking a better question of every channel, not just Meta: if I switched this off, how much of the revenue would still turn up?

That question is the whole game. Everything that would've happened anyway is something you're paying for but not causing. The brands that scale cleanly are the ones who know that split and budget against the honest number. The ones who stall are usually the ones who scaled a dashboard figure that was 40 to 60% claimed rather than caused, and couldn't work out why the bank balance wouldn't follow.

You don't need to get this perfect. You just need to stop trusting a number you've never checked against the money.

If you'd find it useful to see exactly how much of your reported Meta number is real, that reconciliation between platform ROAS and the revenue Shopify actually booked is one of the first things we map in a Signal/Noise Audit. No spreadsheet of your own required. We'll show you where the claimed sales end and the caused ones begin, and you can decide what to do with the gap.

Either way, run the compare-attribution step this week. Worst case, your view-through is low and you've confirmed the account's honest. Best case, you've just found the reason the spend wasn't translating, and you can finally fix it.