Real ROAS is the revenue your store actually collected, net of refunds, divided by the ad spend you actually paid in the same period. Dashboard ROAS is what Meta, TikTok or Google calculate with their own attribution rules. They measure different things, which is why the 3.5x in Ads Manager rarely survives contact with your bank account.
You open Ads Manager, see a 3.5x ROAS and breathe easy. You open your bank and the numbers don't add up. You're not imagining it. The platform is answering the question "how many purchases can I connect to my ads?", and you need the answer to "how much money did my ads bring in?". Those questions sound alike and have very different answers.
Where the inflation comes from
Generous attribution windows
Meta's default attribution setting is 7-day click and 1-day view. Any purchase within seven days of a click, or within a day of someone simply seeing an ad, can be credited to that ad. Some of those people would have bought anyway: they were already on your email list, they searched your brand name, or a friend recommended you. The platform has no way to know, so it counts them.
Modeled conversions
Since iOS privacy changes and the spread of browser tracking protection, platforms see fewer conversions directly. To fill the gap they estimate, using statistical modeling, conversions they believe happened but couldn't observe. The modeling may be reasonable on average, but it is still an estimate, and it is not tied to an order in your Shopify admin.
Double counting across channels
A customer clicks a TikTok ad on Monday, a Meta ad on Wednesday and a Google Shopping ad on Friday, then buys. All three platforms can claim the full sale in their own dashboards, because each only sees its own touchpoints. Your store received one order; your dashboards report three.
Value and timing differences
Depending on how your pixel or Conversions API sends the purchase value, it may include tax and shipping, or be in a different currency. Refunds that happen later are not subtracted. And Ads Manager generally reports a conversion on the date of the ad interaction that led to it, not the day the order was placed, so the same day can look different in each tool.
The ads dashboard grades its own homework. You need a grade that matches your cash.
The formulas
Dashboard ROAS = Platform-attributed purchase value ÷ Platform spend Real ROAS (blended) = (Store revenue − Refunds) ÷ Total ad spend Real ROAS per channel = Net revenue of orders matched to that channel ÷ Channel spend Break-even ROAS = 1 ÷ Contribution margin before ads (%)
Blended real ROAS (often called MER) is the most honest number, because it can't be inflated by overlap: it uses every order and every cost exactly once. Per-channel real ROAS needs a matching rule, such as UTM parameters, click IDs or a first-party pixel that ties the order back to a session.
A worked example
Imagine a store that, in one month, spends $12,000 on Meta and $3,000 on TikTok. Revenue figures exclude sales tax.
| Source | Spend | Attributed / real revenue | ROAS |
|---|---|---|---|
| Meta (Ads Manager) | $12,000 | $42,000 reported | 3.50x |
| TikTok (Ads Manager) | $3,000 | $9,000 reported | 3.00x |
| Platforms combined | $15,000 | $51,000 claimed | 3.40x |
| Shopify, all orders | $15,000 | $40,000 collected | 2.67x |
| Shopify, net of $2,000 refunds | $15,000 | $38,000 | 2.53x |
The two platforms together claim $51,000 of sales. The store only sold $40,000 in total, including email, organic and direct traffic. At least $11,000 of claimed revenue is overlap or modeling that never turned into an order.
Matching orders to channels
Now match each Shopify order to the channel that brought the session, using UTMs and click IDs. Suppose $26,000 of orders match Meta, $5,000 match TikTok and the remaining $9,000 come from email, organic search and direct visits.
Meta real ROAS = 26,000 ÷ 12,000 = 2.17x (vs 3.50x reported) TikTok real ROAS = 5,000 ÷ 3,000 = 1.67x (vs 3.00x reported) Unmatched revenue = 40,000 − 26,000 − 5,000 = $9,000
Comparing with break-even
Suppose this store keeps 40% of net sales after COGS, shipping and payment fees, before ads. Its break-even ROAS is 1 ÷ 0.40 = 2.5x. Any ROAS above 2.5x makes money; anything below loses it.
Contribution before ads = 38,000 × 0.40 = $15,200 Profit after ads = 15,200 − 15,000 = $200 Meta: 26,000 × 0.40 − 12,000 = 10,400 − 12,000 = −$1,600 TikTok: 5,000 × 0.40 − 3,000 = 2,000 − 3,000 = −$1,000 Unmatched: 9,000 × 0.40 = $3,600 Refunds: −2,000 × 0.40 = −$800 Check: −1,600 − 1,000 + 3,600 − 800 = $200
Ads Manager says both channels clear 3x comfortably. In reality the store as a whole made $200 after ads, barely above break-even, and neither channel pays for itself on matched orders. That is a completely different decision than "scale both".
How to reconcile it in practice
The fix isn't another bid strategy: it's anchoring spend to your real orders.
- Use the same period and time zone everywhere. Set your ad accounts and your store to the same time zone, or convert before comparing.
- Use spend as the denominator, always from what the platform billed, converted to your store currency.
- Use store revenue excluding tax, and subtract refunds against the original orders.
- Tag every ad with consistent UTM parameters so orders can be matched to campaigns and ads.
- Track blended real ROAS daily and per-channel real ROAS weekly. Daily per-channel numbers are noisy.
- Compare everything to your break-even ROAS, recalculated when your costs or prices change.
Should you ignore Ads Manager?
No. The platform's numbers are still useful for comparing ads within the same platform: which creative, audience or placement converts better relative to the others. The attribution bias is roughly similar across ads in the same account, so the ranking is informative even if the absolute ROAS is inflated. What you shouldn't do is use the dashboard ROAS to decide how much total budget the business can afford.
If you want to know how much a channel really adds, the strongest test is incrementality: pause or reduce it in a region or period and measure what happens to total store revenue. Platforms offer lift studies for this, and you can run simple versions yourself by comparing blended revenue before and after a controlled change.
What you gain
Once real ROAS is your reference, you stop optimizing on sand. You kill campaigns that only burn budget, scale those that truly pay, and stop being surprised at month-end. ProfitFly does this reconciliation automatically: it crosses each channel's spend with your real Shopify orders and shows real ROAS and profit per campaign, with costs already deducted.
Frequently asked questions
Why is my Meta ROAS higher than my Shopify numbers?
Because Meta credits purchases within its attribution window, including view-throughs, models conversions it can't see, and doesn't know that other channels claim the same sale. Shopify only counts real orders once.
What is a good ROAS?
One above your break-even ROAS. Calculate it as 1 divided by your contribution margin before ads: with a 40% margin, break-even is 2.5x; with 25%, it is 4x.
Is real ROAS the same as MER?
Blended real ROAS is essentially MER (marketing efficiency ratio): total revenue divided by total ad spend. Subtracting refunds makes it stricter.
Should I change Meta's attribution window?
You can compare windows to understand how much of the reported value is view-through or late, but changing the setting doesn't change what really happened. Reconcile with orders instead.