True ROAS vs platform ROAS
Platform ROAS is the return an ad platform says it drove. True ROAS is the return your own tracking can actually tie to real orders. The gap between them decides where your budget should go.
Always quoted with an attribution model and window, for example 30-day first touch.
What is true ROAS?
Every ad platform reports its own ROAS. Meta, Google and the rest each count the purchases they think they caused, using their own rules, and then mark their own homework. Add up what every platform claims and you'll often get more revenue than your store actually made.
True ROAS starts from the other end. It takes the orders that really landed in your store, works out which ads each customer actually interacted with, and credits revenue using one attribution model across every channel. One order, one owner, no double counting.
How it's calculated
True ROAS needs three things joined together: the order, the customer journey that led to it, and the spend behind each touchpoint. BooleanMaths merges orders at the order ID from three streams, so an order isn't lost when one source drops it.
| Input | Where it comes from | What it gives us |
|---|---|---|
| Touchpoints | Server-side pixel on the site | Which ads and channels each visitor came from, and when |
| Orders | Shopify order and checkout webhooks | Every order and its value, including third-party checkouts |
| Delivery status | Shipping partner data | Which orders were delivered, cancelled or returned to origin |
| Spend | Ad platform APIs | Cost per channel, campaign and ad |
Why platform ROAS runs higher
The gap isn't random. It comes from a handful of rules that all push the platform's number up.
Retargeting tends to be the most inflated, because it reaches people who were already close to buying and claims their purchase on a view.
Worked examples
Example 1: One Meta account, one week
An illustrative brand spends ₹5,00,000 on Meta in a week across three campaigns. Ads Manager reports a ROAS of 3.80. Tracked against real orders, the account returned 2.60.
| Campaign | Spend | Platform revenue | Platform ROAS | Tracked revenue | True ROAS | Inflation |
|---|---|---|---|---|---|---|
| Prospecting, broad | ₹2,00,000 | ₹6,40,000 | 3.20 | ₹4,60,000 | 2.30 | 39% |
| Advantage+ shopping | ₹2,00,000 | ₹7,80,000 | 3.90 | ₹5,80,000 | 2.90 | 34% |
| Retargeting | ₹1,00,000 | ₹4,80,000 | 4.80 | ₹2,60,000 | 2.60 | 85% |
| Total | ₹5,00,000 | ₹19,00,000 | 3.80 | ₹13,00,000 | 2.60 | 46% |
- The ranking flipsOn platform ROAS, retargeting is the star at 4.80. On true ROAS, Advantage+ leads and retargeting drops to second.
- RetargetingClaims nearly twice the revenue it can be tied to. Much of it is view-through credit on people who were already on their way to buy.
- The account₹19,00,000 claimed against ₹13,00,000 tracked. Budget planned on 3.80 would overspend on every campaign.
Example 2: Two ads, and the one you'd wrongly scale
Two ads in the same campaign, same spend. Scale on platform ROAS and the money goes to the wrong one.
Using true ROAS to decide
Never use platform ROAS for decisions, not even for small tactical tweaks. Judge every ad on true ROAS, under the model that fits its funnel stage, against your break-even ROAS. These are starting rules to tune to your margins.
Common mistakes
- "Just for small tweaks"Daily budget nudges on platform ROAS add up to a budget plan built on inflated numbers.
- Mixing models across channelsComparing Meta on 7-day click to Google on last click isn't a comparison. Use one model for all.
- Applying a flat discount"Meta over-reports by 30%" isn't true across the board. Inflation varies by campaign, as the example shows.
- Ignoring the windowA true ROAS number without its model and window can't be compared with anything.
Frequently asked questions
Why is Meta's ROAS higher than what I see in my store?
Meta credits itself for view-throughs, modelled conversions and sales other channels also touched. Your store only has each order once, so the sum of platform claims often exceeds real revenue.
Is true ROAS always lower than platform ROAS?
Usually, but not always. When a platform loses signal, for example on broken checkout journeys, it can under-report, and true ROAS comes out higher.
Which attribution model should true ROAS use?
Match the model to the ad's job: 30-day first touch for prospecting, 14-day linear for middle of funnel, 3-day last touch for retargeting. The attribution models page explains each.
Should I still feed purchase data back to Meta?
Yes. The platform's algorithm needs conversion signal to optimise delivery. Send it accurate data through a Conversions API, but make your own budget decisions on true ROAS.
Related metrics
See true ROAS next to what platforms claim
BooleanMaths tracks every order server-side and shows true ROAS for every channel, campaign and ad, under the attribution model you choose.
True ROAS vs platform ROAS
Platform ROAS is the return an ad platform says it drove. True ROAS is the return your own tracking can actually tie to real orders. The gap between them decides where your budget should go.
Always quoted with an attribution model and window, for example 30-day first touch.
What is true ROAS?
Every ad platform reports its own ROAS. Meta, Google and the rest each count the purchases they think they caused, using their own rules, and then mark their own homework. Add up what every platform claims and you'll often get more revenue than your store actually made.
True ROAS starts from the other end. It takes the orders that really landed in your store, works out which ads each customer actually interacted with, and credits revenue using one attribution model across every channel. One order, one owner, no double counting.
How it's calculated
True ROAS needs three things joined together: the order, the customer journey that led to it, and the spend behind each touchpoint. BooleanMaths merges orders at the order ID from three streams, so an order isn't lost when one source drops it.
| Input | Where it comes from | What it gives us |
|---|---|---|
| Touchpoints | Server-side pixel on the site | Which ads and channels each visitor came from, and when |
| Orders | Shopify order and checkout webhooks | Every order and its value, including third-party checkouts |
| Delivery status | Shipping partner data | Which orders were delivered, cancelled or returned to origin |
| Spend | Ad platform APIs | Cost per channel, campaign and ad |
Why platform ROAS runs higher
The gap isn't random. It comes from a handful of rules that all push the platform's number up.
Retargeting tends to be the most inflated, because it reaches people who were already close to buying and claims their purchase on a view.
Worked examples
Example 1: One Meta account, one week
An illustrative brand spends ₹5,00,000 on Meta in a week across three campaigns. Ads Manager reports a ROAS of 3.80. Tracked against real orders, the account returned 2.60.
| Campaign | Spend | Platform revenue | Platform ROAS | Tracked revenue | True ROAS | Inflation |
|---|---|---|---|---|---|---|
| Prospecting, broad | ₹2,00,000 | ₹6,40,000 | 3.20 | ₹4,60,000 | 2.30 | 39% |
| Advantage+ shopping | ₹2,00,000 | ₹7,80,000 | 3.90 | ₹5,80,000 | 2.90 | 34% |
| Retargeting | ₹1,00,000 | ₹4,80,000 | 4.80 | ₹2,60,000 | 2.60 | 85% |
| Total | ₹5,00,000 | ₹19,00,000 | 3.80 | ₹13,00,000 | 2.60 | 46% |
- The ranking flipsOn platform ROAS, retargeting is the star at 4.80. On true ROAS, Advantage+ leads and retargeting drops to second.
- RetargetingClaims nearly twice the revenue it can be tied to. Much of it is view-through credit on people who were already on their way to buy.
- The account₹19,00,000 claimed against ₹13,00,000 tracked. Budget planned on 3.80 would overspend on every campaign.
Example 2: Two ads, and the one you'd wrongly scale
Two ads in the same campaign, same spend. Scale on platform ROAS and the money goes to the wrong one.
Using true ROAS to decide
Never use platform ROAS for decisions, not even for small tactical tweaks. Judge every ad on true ROAS, under the model that fits its funnel stage, against your break-even ROAS. These are starting rules to tune to your margins.
Common mistakes
- "Just for small tweaks"Daily budget nudges on platform ROAS add up to a budget plan built on inflated numbers.
- Mixing models across channelsComparing Meta on 7-day click to Google on last click isn't a comparison. Use one model for all.
- Applying a flat discount"Meta over-reports by 30%" isn't true across the board. Inflation varies by campaign, as the example shows.
- Ignoring the windowA true ROAS number without its model and window can't be compared with anything.
Frequently asked questions
Why is Meta's ROAS higher than what I see in my store?
Meta credits itself for view-throughs, modelled conversions and sales other channels also touched. Your store only has each order once, so the sum of platform claims often exceeds real revenue.
Is true ROAS always lower than platform ROAS?
Usually, but not always. When a platform loses signal, for example on broken checkout journeys, it can under-report, and true ROAS comes out higher.
Which attribution model should true ROAS use?
Match the model to the ad's job: 30-day first touch for prospecting, 14-day linear for middle of funnel, 3-day last touch for retargeting. The attribution models page explains each.
Should I still feed purchase data back to Meta?
Yes. The platform's algorithm needs conversion signal to optimise delivery. Send it accurate data through a Conversions API, but make your own budget decisions on true ROAS.
Related metrics
See true ROAS next to what platforms claim
BooleanMaths tracks every order server-side and shows true ROAS for every channel, campaign and ad, under the attribution model you choose.