7-Day Click vs 1-Day View: The Meta Attribution Setting Most Ecom Brands Get Wrong

A customer that costs you $19 to acquire on one setting costs you $58 to acquire on another. Same ads. Same spend. Same actual buyers walking through the same checkout. The only thing that changed was a dropdown in Ads Manager.

That's not a rounding error. That's the difference between a number you can build a business on and a number that's quietly lying to you. And most brands I look at are running the flattering one without realising what it's costing them in bad decisions.

So here's my verdict up front, and I'll back it with the mechanics: for the overwhelming majority of ecommerce brands, you want 7-day click, 1-day view turned off entirely. Not because the reported numbers look nicer. They'll look worse. You want it because it's the only setting that tells you something true and trains the algorithm on the thing you actually care about.

Let me make the case.

What the attribution window actually does (it's two jobs, not one)

Most founders think the attribution window is a reporting choice. Pick how generously you want to count conversions, read your ROAS, move on. That's half of it, and it's the less important half.

The attribution window does two jobs. The first is reporting: it decides which sales Meta takes credit for. The second, and this is the one nobody talks about, is optimisation: it tells Meta's algorithm what kind of conversion to go and find more of.

That second job is everything. Meta has said it plainly in its own documentation. With a 7-day click setting, the system learns from conversions that happened within 7 days of a click, and then goes and shows your ads to the people most likely to click and convert in that window. The window you pick isn't just a lens you read results through. It's the instruction set you hand the machine.

So when you bolt 1-day view onto your attribution, you're not just changing your report. You're telling the algorithm that a view counts. And once a view counts, the algorithm will happily go chase cheap impressions, because an impression is a much easier thing to buy than a click.

Why view-through inflates everything

Here's the problem with view-through attribution in one sentence: it claims credit for sales it probably didn't cause.

A view-through conversion means someone was served your ad, didn't click it, and then bought within a day. Meta counts that as a win for the ad. But think about what actually happened there. The person scrolled past your ad. Maybe they registered it, maybe they didn't even look up from the feed. Then later they bought, possibly because of a completely different touchpoint, an email, a mate's recommendation, a Google search, a different ad on a different platform.

I think of it like a billboard on the motorway. If someone drives past your billboard and buys your product that afternoon, did the billboard sell it? Maybe. Maybe not. You genuinely can't say. A click is different. A click is someone reaching out and putting their hand up. When somebody clicks your ad and then buys, you can be confident your ad was in the room when the decision got made.

This is why the two numbers in my opening are so far apart. Switch on 1-day view and the reported cost to acquire a customer might read ~$19. It looks brilliant. You feel like a genius. But a big chunk of those "conversions" are sales that were going to happen anyway, that view-through has simply scooped up and put on Meta's tab. Strip the view-through credit out and run pure 7-day click, and that same activity might show a true cost of ~$58. The $19 was never real. It was Meta marking its own homework and giving itself an A.

The kicker: it doesn't just inflate the report. Because view-through also drives optimisation, you've now pointed the algorithm at chasing impression-volume from people who were already going to buy. You're paying to advertise to your own warm demand and calling it acquisition.

What 7-day click does that's worth the uglier numbers

Switch to 7-day click only and the first thing that happens is your reported ROAS drops and your reported CPA jumps. I want to be honest about that because it's where most people lose their nerve and switch back. Don't. The number got worse because the number got honest.

What you've actually done is hand the algorithm a cleaner instruction. Every dollar you spend is now training the system on one outcome: people who click and then buy. That's the most incremental signal you can optimise toward, because a click is a real, deliberate action that you can be confident your ad caused.

Over time, this compounds in your favour. I've watched accounts where, after switching to click-only attribution and holding it for a few months, the click-through rate on the same creative drifted up, the cost per click eased, and the genuine cost per acquisition settled lower than where the "flattering" view-through setting had it landing on a real-money basis. As a rough shape of it, a brand might sit at a reported ~$40 CPA on a mixed view-and-click setup, switch to 7-day click, watch the reported number leap up in week one, and then over a quarter see the true click-based CPA grind down toward ~$31 as the algorithm gets better at finding actual clickers. Those are illustrative numbers, not a promise, but the direction is the pattern I keep seeing. You traded a comforting lie for a metric that actually improves because you're optimising for the right thing.

The volatility nobody warns you about: 1-day click and the learning threshold

There's a third option people reach for when they're spooked by view-through but want "tighter" attribution, and that's 1-day click. I'd steer most brands away from it, and the reason is mechanical, not philosophical.

Meta's algorithm needs roughly 50 conversion events per ad set to get out of the learning phase and stabilise. Here's the catch most people miss: with a 7-day click window, those 50 events can accumulate across a full week. With a 1-day click window, you effectively need to be generating that volume on a daily basis for the system to have enough fresh same-day signal to lean on. That's a brutally higher bar.

Practically, unless you're spending serious money, say north of ~$20,000 a day, 1-day click starves the algorithm of data. And starved algorithms produce wild, jumpy results. I've seen 1-day click accounts swing from a 5x return one day, to 0.1x the next, to a 1x, then back up, with nothing actually changing in the account. That kind of day-to-day whiplash makes it almost impossible to read whether anything's working, and it tempts founders into knee-jerk changes that make it worse.

Seven-day click smooths that out. Because the window is wider, the data is denser, the ad set clears the learning threshold more easily, and the daily numbers hold steadier. You can actually make decisions off a 7-day click report. You mostly can't off a jumpy 1-day click one.

So the spread isn't just "honest versus inflated." It's also "stable versus volatile." Seven-day click wins on both.

The narrow cases where adding view actually makes sense

Now, I said overwhelming majority for a reason, because there's a real exception and I'd be doing you a disservice to pretend it's never on the table.

If you sell a high-consideration, high-AOV product, the kind of thing someone spends ~$400 on and mulls over for a fortnight before buying, the customer journey genuinely doesn't fit neatly inside a click-and-buy window. Those buyers see your ad, don't click, go away, research, compare, sit on it, and come back through a different door entirely. For that kind of product, an honest argument exists that some view-through credit reflects something real about how the ad contributed to a slow, considered decision.

Meta's also rolled out newer engaged-view models that count a view only when someone actually watched a meaningful chunk of the ad and acted within a day, which is a tighter, more defensible version of the old view-through. For a long-cycle, high-ticket brand, that's worth testing.

But notice how narrow this is. High AOV. Long deliberation. A product people research before buying. If that's genuinely you, test view-through with your eyes open and watch what it does to your true blended numbers. If you're a sub-$100 AOV impulse-ish product, which is most ecom, this exception isn't yours. Turn view off.

And here's the rule that holds even for the exceptions: as you scale into bigger, warmer audiences with repeat buyers in the mix, that's exactly when view-through starts claiming the most credit it didn't earn, because those warm buyers were always going to come back. So even brands that justify view early should be stripping it out as they grow.

What I'd actually do

If it were my account, here's the short version. Run 7-day click, 1-day view off, as your default and your source of truth. Expect the reported numbers to look worse on the switch, and hold your nerve through it, because you're now optimising for real clicks and real buyers. Keep an eye on your blended MER as the honest sanity check that sits above any single platform's self-reported figure. And only entertain view-through if you're genuinely a high-AOV, long-consideration brand, and even then, treat it as a test you're watching closely, not a setting you set and forget.

The one thing I'd push back on hardest is panicking at honest numbers. A higher CPA that's true beats a lower one that's fiction every single time, because you can only make good decisions on the true one.

Go and pull your own account up this week. Look at it on 1-day view, then look at it on 7-day click only, and sit the two side by side. The size of that gap is the size of the story you've been telling yourself. I'd love to hear what you find when you run it.