Marketing Attribution Windows Explained: Why Cross-Channel ROAS Comparison Is Misleading

Marketing Attribution Windows Explained: Why Cross-Channel ROAS Comparison Is Misleading

Every major ad platform attributes conversions using a different default window - and comparing ROAS numbers across platforms without accounting for those window differences produces systematically wrong conclusions. Google Ads defaults to 30-day click attribution. Meta defaults to 7-day click plus 1-day view. Snapchat includes a 28-day view-through window by default. Pinterest uses 30-day click plus 1-day engagement. A side-by-side comparison of platform-reported ROAS across these four platforms is not a fair comparison - it is four platforms measuring different things and calling them the same metric.

Understanding attribution windows is the prerequisite skill for cross-channel budget decisions. Without it, you risk defunding channels that are contributing significant revenue (because their window is conservative) and over-investing in channels that are over-claiming credit (because their window is generous).

Attribution window: the period of time after a user interacts with an ad during which a conversion (purchase, lead, install) will be attributed to that ad. A 30-day click window means a conversion occurring 29 days after an ad click will still be credited to that campaign.

Click-through attribution: attributing a conversion to an ad that the user clicked. The most conservative and most defensible attribution type - the user expressed direct interest by clicking.

View-through attribution (VTA): attributing a conversion to an ad that the user saw but did not click. If someone views a Snapchat ad on Monday and purchases from your website on Thursday (without clicking the ad), a 7-day view-through window credits Snapchat with that conversion.

Last-click attribution: within a multi-touchpoint path, the last ad clicked before conversion receives 100% of the credit. This is the default for most platform reporting and systematically disadvantages upper-funnel channels that appear earlier in the path.

Key takeaways

Google Ads, Meta, Snapchat, and Pinterest each use different default attribution windows

Google Ads defaults to 30-day click, Meta to 7-day click plus 1-day view, Snapchat includes 28-day view-through, and Pinterest uses 30-day click plus 1-day engagement. A side-by-side ROAS comparison across these four is measuring four different methodologies with one label.

Moving Meta from 7-day click plus 1-day view to 7-day click typically reduces attributed conversions by 15 to 25 percent

This reduction reveals the view-through inflation embedded in the default setting. The delta is not a drop in actual performance - it is the removal of credit for conversions that occurred without a click.

A channel with a 30-day view-through window reports 2 to 4 times more conversions than 1-day click attribution

The difference is purely methodological, not performance-based. The same channel, the same campaign, and the same customer behaviour produces dramatically different numbers depending solely on which window is applied.

The only defensible cross-channel ROAS comparison normalises all channels to the same attribution window

Typically this means 7-day click, 0 view, applied consistently across all platforms before drawing budget allocation conclusions. Any comparison that preserves each platform's default window is comparing methodologies, not channels.

Defunding a channel based on lower platform-reported ROAS may mean cutting the most conservative measurer

A channel using a shorter attribution window will report fewer conversions than the same channel measured on a longer window. Teams that defund based on reported numbers without normalising windows routinely defund their best-performing channels.

Attribution windows are only one side of the model question. The other side is which touchpoint gets credit within the window. Last-click attribution combined with a 30-day window systematically erases the first 2-3 months of a typical B2B SaaS journey, even with a longer-window attribution model. See why last-click attribution is broken for the broader model failure pattern.

Default attribution windows by platform - 2026

Interpreting the table: Snapchat's 28-day view-through window is the most generous of any major platform. For a brand running Snapchat at scale, a high percentage of conversions will have a Snapchat impression somewhere in the prior 28 days - meaning Snapchat claims credit for a large fraction of all purchases. Google's default has no view-through component at all - it only counts conversions from users who clicked an ad within 30 days.

Why different windows produce different ROAS numbers for the same campaign

A concrete example. A DTC brand runs simultaneous campaigns on Meta (7-day click), Google Shopping (30-day click), and Snapchat (1-day swipe + 28-day view-through) with equal £1,000/day budgets.

In any given week:

• Google Shopping attributes 40 conversions worth £8,000. ROAS: 8x.
• Meta attributes 35 conversions worth £7,000. ROAS: 7x.
• Snapchat attributes 32 conversions worth £6,400. ROAS: 6.4x.

Reading this, you might conclude: Google > Meta > Snapchat in efficiency. But consider:

1. Google's 30-day window includes conversions from clicks up to 30 days ago - many of those customers were initially acquired by Meta or TikTok prospecting campaigns earlier in the month. Google gets last-click credit on the branded search that closed the sale.

2. Snapchat's 28-day view-through means those 32 attributed conversions include many customers who never clicked a Snapchat ad - they just had a Snapchat impression somewhere in their path. Remove view-through attribution and Snapchat's conversion count might drop to 8-12.

3. Meta's 7-day window misses conversions from prospects who engaged with a Meta ad in week 1 of the month but converted in week 2 - those show as unattributed or as Google/organic conversions.

The real story may be: Snapchat prospecting drives awareness, Meta retargeting drives consideration, Google branded search closes the sale - but platform-reported ROAS credits Google with most of the revenue while Snapchat looks like the weakest channel.

Prooflytics

Turn attribution into decisions, not debates

One brief across every channel, with the memory of what each one drove.

Start free trialSee pricing

14 days free · no credit card

How to normalise attribution windows for cross-channel comparison

Three methods, in order of rigour:

Method 1 - Align windows manually. Set all platforms to the same attribution window before comparing. The most conservative comparable window is 1-day click only (no view-through) - this creates an apples-to-apples comparison but dramatically lowers reported ROAS on every channel, particularly Snapchat. A more practical option: 7-day click, no view-through across all platforms.

How to adjust:

• Meta: Ads Manager to Columns to Attribution Setting to 7-day click, no view
• TikTok: Ads Manager to Attribution Settings to 7-day click
• Snapchat: Ads Manager to Attribution Manager to 7-day click, turn off view-through
• Google Ads: Account Settings to Attribution to set to 7 days (affects future attribution, not historical)

Method 2 - Use GA4 as a platform-neutral source. GA4 with data-driven or linear attribution applies a consistent model across all traffic sources. GA4 will show lower ROAS numbers than platform native reporting for all channels - because GA4 does not include view-through attribution and handles cross-device less generously. The advantage: the same methodology applies to all channels simultaneously.

Prooflytics can connect GA4 attribution data alongside platform-native data when both are linked. This lets you see "Meta platform-reported ROAS: 7x, GA4-attributed Meta ROAS: 4x" - the gap shows how much of Meta's claimed revenue is from view-through or cross-device paths that GA4 does not credit to Meta.

Method 3 - Incrementality testing. The only method that measures true causal contribution rather than correlation-based attribution. Run a geo holdout test or audience holdout test: turn off a channel for one audience segment or region for 4-6 weeks, compare conversion rates against the control group. The conversion lift in the treated group vs control is the true incremental ROAS. See incrementality testing in marketing for the full methodology.

The view-through attribution problem: when it inflates ROAS

View-through attribution (VTA) is the most contested measurement choice in paid media. It credits an ad for a conversion that occurred after someone merely saw the ad - without clicking. The case for VTA: some ad formats (display, video) influence purchase intent even without a click; excluding VTA undercounts their contribution. The case against: VTA credits any ad served to anyone who later happened to purchase, regardless of causal relationship.

When VTA inflates ROAS significantly:

• High impression volume campaigns on Snapchat or Pinterest reaching large audiences (most purchasers will have seen an ad somewhere in 28 days)
• Retargeting campaigns with high frequency (every person who saw the ad was already considering purchase - the ad may not have changed their decision)
• Brand campaigns running alongside direct-response campaigns (the brand ad gets credit for conversions driven by the direct-response ad)

A practical test for VTA inflation: compare your platform-reported ROAS (including view-through) to your click-only ROAS on the same window. If click-only ROAS is less than 50% of view-through ROAS, your reporting is heavily dependent on view-through attribution - which is a signal to test incrementality before making budget decisions based on that ROAS number.

How Prooflytics handles cross-channel attribution

Prooflytics syncs platform-native ROAS data from each connected channel - including the attribution window that platform uses by default. In the daily briefing, each channel's ROAS is displayed alongside its attribution context, so you know whether a number reflects a 30-day click window or a 7-day click + 28-day view window.

For teams connected to GA4 alongside their paid channels, Prooflytics can surface both the platform-reported ROAS and the GA4-attributed ROAS for each channel in the same view - making the attribution gap visible. This gives you both the "how the platform sees it" number and the "what platform-neutral measurement says" number, side by side, before you make a budget decision.

For deeper treatment of multi-touch attribution models and how they distribute credit across the customer journey, see the multi-touch attribution guide. For DTC brands using post-purchase surveys as a measurement layer on top of platform attribution, see the post-purchase survey attribution guide.

Bottom line

• Default attribution windows differ substantially across platforms: Google 30-day click, Meta 7-day click + 1-day view, Snapchat 1-day swipe + 28-day view-through - these are not comparable without normalisation
• Summing platform-reported attributed revenue across channels produces a number higher than actual revenue due to double-counting
• To compare ROAS cross-channel fairly: set all platforms to the same window (7-day click, no view-through is the practical standard)
• Snapchat's 28-day view-through is the most generous window - ROAS figures that include this should be verified against click-only data before making budget decisions
• Prooflytics displays each channel's ROAS alongside its attribution context, making the window difference visible in the daily brief

You can read independent reviews of Prooflytics on G2 and compare it to alternatives in the marketing analytics category.