Attribution & Measurement
View-through conversions: should you count them?
There isn't a single correct answer — a view-through conversion is a purchase credited to an ad someone saw but never clicked, and whether that credit is meaningful depends on your product, your volume of impressions, and how much you trust the assumption behind it. Here's both sides, and a way to decide using your own numbers instead of a rule of thumb.
What a view-through conversion actually is
An ad is served to someone's screen. They don't click it, don't interact with it at all — maybe they scrolled past it in a feed, maybe it played as a pre-roll video they skipped after a few seconds. Some time later, within a defined window, that same person makes a purchase. The platform that served the impression credits itself with the conversion, on the theory that seeing the ad played some role in the eventual purchase. This applies across Meta's feed and Reels placements, and across Google's Display Network and video inventory — the mechanic is the same wherever an ad can be served without requiring a click.
Click-through attribution, by contrast, requires an actual click before the purchase. That's a real, logged interaction — the customer did something. View-through has no equivalent action to point to. The entire claim rests on proximity in time between an impression and a purchase.
The case for counting it
Not every real purchase is preceded by a click, and treating clicks as the only legitimate signal of ad influence ignores a genuine pattern in how people actually shop. Someone scrolls past a product ad, doesn't click because they're mid-commute or mid-meeting, remembers it that evening, and types the brand name directly into a search bar or their browser. No click was ever logged against the original ad, but the ad plausibly did the work of introducing the product. A measurement system that only ever credits clicks will systematically undercount awareness-stage advertising — video, Reels, Display placements — that's doing exactly what it's supposed to do: get someone thinking about a product they'd otherwise never have searched for.
The case against — or at least for skepticism
Impressions are cheap and abundant in a way clicks aren't. A single campaign can serve an ad to the same broad audience dozens of times over a week for a fraction of what a click costs, which means almost any active shopper in your addressable audience has a reasonable chance of having “seen an ad” shortly before buying — whether or not that ad had anything to do with the decision. If someone was already going to search for your brand and buy, an impression that happened to land in front of them beforehand isn't evidence it caused the sale. It's closer to correlation dressed up as causation.
No platform can prove counterfactual incrementality — what would have happened without the ad — from view data alone. Meta and Google both know whether an impression was served and whether a purchase followed. Neither knows what that same person would have done if the impression had never been shown, because there's no unexposed version of that specific person to compare against. The view-through number is a plausible story, not a measured causal effect.
Why this shows up as inflated ROAS, not just a philosophical debate
This isn't only an abstract measurement question — it shows up directly in reported ROAS. Because view-through credit only ever adds purchases to a campaign's attributed total and never subtracts any, it's one of the largest single levers behind a platform reporting more revenue than the campaign plausibly drove. It also contributes to the broader pattern where Meta and Google both claim the same Shopify order and their combined reported revenue overstates what the store actually made — see why Meta and Google both claim the same Shopify sale — and it's a big part of why widening an attribution window setting can raise reported ROAS without the campaign itself changing at all, covered in why changing your Meta attribution window changes your ROAS.
A framework for deciding with your own numbers
Look at the split, not just the total. Break reported conversions into click-through and view-through separately — most platforms let you view this. If view-through is a small share of total attributed conversions, the debate barely matters to your numbers either way. If it's a large share, it's worth scrutinizing before you trust the headline ROAS figure.
Consider your consideration cycle. A cheap, impulse product has less room for a delayed, unclicked purchase to plausibly trace back to a specific impression — the case for view-through credit is weaker. A higher-consideration purchase, where people research before buying, has more genuine room for an impression to do real work days before the purchase — the case is stronger, though still not proof.
Watch what happens when you exclude it. Recalculate ROAS using click-through data only, and compare it against your blended, Shopify-revenue-based efficiency number. If click-only ROAS is already healthy, view-through credit is icing, not the thing holding the campaign's case together. If a campaign only looks profitable once view-through is included, treat that as a reason to test further, not as confirmation the campaign works.
Use it for direction, not for individual campaign verdicts. Tracking whether view-through volume is trending up or down over time can be a reasonable proxy for changing awareness — but making a scale-or-kill decision on a specific campaign based mostly on view-through credit is building a decision on the least verifiable part of the data you have.
Where Attribix fits
Attribix ties reported conversions back to actual Shopify orders, so you can see how much of a campaign's attributed revenue depends on view-through credit versus orders with a real click or server-recorded event behind them. See how that connection works on Shopify attribution and Meta CAPI for Shopify.
Common questions
What is a view-through conversion?
Should I turn off view-through attribution entirely?
Does a high view-through conversion count mean my ads are working?
How is view-through different from click-through attribution?
Want your view-through share checked?
Send an inquiry and we'll break down how much of your reported conversions come from view-through credit versus clicks, on your actual account.