Why Platform Conversions Add Up to More Than Your Orders

Ad platforms double counting conversions is the reason your weekly report never adds up. Meta says it drove 140 purchases, Google says 95, TikTok says 60, and the shop recorded 210 orders. Nobody made a mistake in the maths: each platform counts by its own rules, sees only its own ads, and credits any sale that lands inside its window. This guide explains why the sum of platform conversions is higher than your Shopify orders, how to measure that gap, how to deduplicate conversions across channels, and what to put in front of a client or a founder instead.

The numbers above are an illustration, not a benchmark. Your own ratio is what matters, and the method below shows you how to find it.

Quick answer

Each ad platform attributes a sale to its own ads whenever a click or view falls inside its attribution window, and none of them can see the other platforms. A shopper who saw a TikTok ad, clicked a Meta ad and then searched on Google can be claimed three times. Add view-through credit, modelled conversions and different counting rules, and the platform sum overshoots your orders. Measure the gap by dividing the platform sum by store orders, use your orders as the single source of truth, and assign each order to one channel with UTMs.

Wevion connects Meta, Google, TikTok, Taboola, Snapchat and Outbrain through their official APIs, syncs them into one cross-channel view about every 15 minutes, and links Shopify or WooCommerce orders, so the platform sum and your real orders sit side by side with true profit and POAS (in beta).

Why each platform claims the same sale

The overlap is not a bug in any one platform. It comes from four separate counting choices that are each reasonable on their own and inflate the total when you add the platforms together.

Every platform only sees its own ads

Meta knows about Meta impressions and clicks. Google knows about Google. TikTok knows about TikTok. When a purchase event fires on your site, each platform checks whether that person interacted with one of its ads recently, and if so, it takes the credit. TikTok's own help centre puts it plainly: "Every conversion is attributed only to one touch point such as click, view, or engaged view." That rule holds inside TikTok. It says nothing about Meta or Google, which run the same check on the same order.

The result is that one purchase by one person can appear as one conversion in three or four dashboards. If you want the longer story of how attribution models assign credit inside a single platform, our guide to first-touch vs last-touch attribution covers the models themselves.

Attribution windows are long and they differ

The window is how long after a click or view a platform will still claim a sale. The defaults are not the same across platforms, so the same order can fall inside one window and outside another:

Google's help page states that for click-through conversions "the default window is 30 days", and that you can set it "anywhere from 1 to 90 days, depending on the source of conversions." Snap's Ads API lists SWIPE_28DAY_VIEW_1DAY as the default conversion window for an ad squad. TikTok's ad group settings offer click windows of 1, 7, 14 or 28 days, and its TikTok Shop Ads attribution page uses a 7-day click and 1-day view window. Settings can also be changed per campaign, so check what each account actually uses rather than assuming the default.

Longer windows mean more chances for overlap. A shopper who clicked a Snapchat ad three weeks ago and a Google ad yesterday is inside both windows today.

View-through credit counts people who never clicked

A view-through conversion is a sale credited to an ad the person saw but did not click. Google's default view-through window is 1 day, and its engaged-view window is 3 days. TikTok's engaged view-through attribution "assigns credit to an ad when a user watches at least 6 seconds of the ad." Meta's developer docs for ROAS rules note that "if the attribution window includes both click and view windows, the computed ROAS is the sum of the values."

View credit is not worthless: upper-funnel video does influence people. But a feed impression is cheap to collect and easy to overlap. The same buyer can scroll past ads on three platforms in a day, and each one can claim the view.

Modelled conversions fill in what the platform cannot see

Since browser and mobile privacy changes, platforms estimate conversions they cannot observe directly. Google is open about this on its conversion modelling page: its Conversions column reports both modelled and observed conversions. It adds that modelled conversions can take up to five days to settle in reporting. Other platforms model too, with less public detail.

Modelling is a sensible answer to missing signal, but it means part of each platform's total is an estimate made without knowing what the other platforms are also estimating. Two estimates of the same order do not cancel out.

Counting rules and credit models differ

Platforms also disagree on what one conversion is. Some count every purchase event, some count one per click or per person. Google Ads offers two attribution models, last click and data-driven, and says "the 'Data-driven' model is the default attribution model for most conversion actions" (Google Ads Help). Data-driven credit can split one order into fractions across Google touchpoints, while another platform gives itself the whole order. Reporting time differs too: some platforms log the conversion on the day of the click, your shop logs it on the day of the order.

How to measure the gap

The fastest way to see how much your platforms over-claim is a simple ratio you can rebuild every week. It needs nothing more than the reported purchases and your order export.

1. Pick one period and one time zone. Use a full week or month, and convert every source to the same time zone as your shop.
2. Add up platform conversions. Sum the purchase conversions reported by Meta, Google, TikTok, Snapchat and any other channel for that period. Note each platform's attribution setting next to its number.
3. Count real orders. Take paid orders from Shopify or WooCommerce for the same period, excluding test orders, cancellations and fully refunded orders.
4. Divide. Platform sum divided by store orders is your over-claim ratio. A ratio of 1.0 means the platforms agree with the shop. A ratio of 1.6 means the platforms together claim 60% more orders than happened.
5. Repeat for revenue. Do the same with reported conversion value against net sales. The revenue ratio is often different from the order ratio, which tells you something about which platforms claim the larger baskets.

Two caveats keep the ratio honest. Some orders come from organic search, email or direct visits and were never touched by an ad, so the platform sum should arguably sit below total orders, not at it. And the ratio moves when you change attribution settings, so log any change on the day you make it.

The inflation shows up on revenue too, and our reported ROAS vs true ROAS framework walks through putting both numbers on one page. If you are still deciding whether your current number is trustworthy at all, seven signs you can't trust your ROAS is a quicker diagnostic.

How to deduplicate conversions across channels

Deduplication cannot happen inside the ad platforms, because no platform can see another platform's claims. It happens outside them, in a layer that starts from your orders and works backwards.

Make the store the single source of truth

Your shop knows how many orders happened, what they were worth and what they cost. That makes it the only neutral referee. Every report that adds channels together should start from store orders and allocate them, not start from platform conversions and add them up. Our guide to multi-platform ad reporting with a single source of truth covers the setup in detail.

Tag every ad with consistent UTMs

UTM parameters on every ad link let your store or analytics tool record which channel and campaign sent each visit. Use one naming convention across all platforms: the same spelling for utm_source, a consistent utm_medium for paid social and paid search, and campaign names that match what is in the ad account. Inconsistent tags create a large "unknown" bucket that cannot be allocated to anyone.

Assign each order to one channel with one rule

Once orders carry UTMs, choose a single rule for who gets the credit, such as last paid click inside 7 days, and apply it to every order. Each order then belongs to exactly one channel, so the channel totals add up to your order count. The rule will be imperfect, and that is fine. What matters is that it is consistent and that everyone reading the report knows which rule it is.

Use first-party order data, not pixels alone

Browser pixels miss conversions and platforms then model them back. Order data from Shopify or WooCommerce does not depend on cookies. Joining ad spend by channel with first-party orders gives you a deduplicated cost per order and, once product costs are in, profit per channel. Our article on Shopify profit tracking from ROAS to POAS shows how to add margin to that join.

Test incrementality when the stakes are high

Deduplication tells you which channel gets the credit for an order. It does not tell you whether the ad caused the order. For that, there are two higher-level tools:

• Holdout or geo tests. Pause or withhold a channel in some regions or for a test group, and compare sales with a control. The difference is the incremental lift.
• Marketing mix modelling (MMM). A statistical model that relates weekly spend by channel to total sales over a long period. It needs months of history and enough spend variation to be useful.

Both are worth the effort for larger budgets. For most teams, a clean order-based report solves the day-to-day problem first.

What to report instead

A good cross-channel report shows both views and labels them clearly, so nobody mistakes a platform's claim for a fact.

• Store orders and net revenue as the headline. This is what happened.
• Deduplicated orders per channel, using your single attribution rule, so channels add up to the headline.
• Blended cost per order and blended ROAS: total ad spend across platforms against total store orders and revenue.
• Profit and POAS, once product costs, shipping and fees are included. A channel can win on orders and still lose money.
• Platform-reported conversions, kept in a separate column with each attribution setting named, used for optimising inside that platform.
• The over-claim ratio, tracked weekly, so a sudden jump prompts a check of tracking or attribution settings.

Platform numbers still have a job. Each algorithm optimises on its own conversions, and comparing ad sets inside Meta with Meta's numbers is fair. The mistake is adding those numbers across platforms and treating the total as sales. For running the wider operation across channels, our pillar on how to manage ads across multiple platforms puts reporting in context with launching, rules and budgets.

Where Wevion fits

Most of the work above is plumbing: pulling spend and conversions from several ad accounts, matching them to orders, and keeping the two in the same view. Wevion's cross-channel analytics bring Meta, Google, TikTok, Taboola, Snapchat and Outbrain into one view, synced about every 15 minutes, so the platform sum is visible without exporting six reports.

Connect Shopify or WooCommerce and profitability analytics (in beta) put real orders, true profit and POAS next to what each platform claims. When the gap between the two widens, you can ask Wavo, the AI copilot, about it on your own account data, and any budget change it proposes only happens after you approve it.

You can try this on your own accounts with a 14-day free trial on the Pro plan, with no card needed. Start your free trial.

Key takeaways

• Ad platforms double count conversions because each sees only its own ads and claims any sale inside its window.
• Windows differ: Google defaults to a 30-day click window, Snapchat to 28-day swipe and 1-day view.
• View-through credit and modelled conversions widen the gap further.
• Measure the over-claim ratio: platform sum divided by store orders, same period and time zone.
• Deduplicate outside the platforms with store orders, consistent UTMs and one attribution rule.
• Report store orders, profit and POAS as the headline, and keep platform numbers for in-platform optimisation.

Sources (checked 22 October 2026): Google Ads conversion windows, Google Ads attribution models, Google Ads conversion modelling, Meta for Developers on attribution defaults, Meta ROAS ad rules, TikTok attribution overview, TikTok ad group attribution settings, TikTok engaged view-through attribution, TikTok Shop Ads attribution, Snap Ads API ad squads.