Retail Media ROAS & Incrementality Calculator
Retail media dashboards report flattering ROAS because they credit the ads with sales you would have won regardless. This calculator does the honest version: it removes the non-incremental sales and shows the return, the profit, and the break-even your spend really needs to clear.
Reported ROAS is attributed sales divided by spend. Incremental ROAS multiplies that by the share of sales the ads actually caused. Enter your spend, the attributed sales from the retailer, your gross margin, and an incrementality estimate; the tool returns your incremental ROAS, the real profit after product cost, and the break-even incrementality — the minimum share of sales that must be truly caused by the ads for the spend to pay for itself.
Retail Media ROAS & Incrementality Calculator inputs and result
| Incrementality | Incremental ROAS | Incremental profit | Verdict |
|---|
How to use this calculator
- Pick your retailerChoosing a network loads a sensible starting incrementality and a benchmark note. It is a starting point, not a verdict — your own measured lift always wins.
- Enter spend and attributed salesUse one period and one scope. Take both numbers straight from the retailer’s dashboard so the reported ROAS matches what your team sees.
- Set your gross marginEnter margin after cost of goods. Retail media often runs on thin retail margins, which is exactly why incremental profit, not ROAS, is the number that matters.
- Estimate incrementalityEnter the share of those sales the ads actually caused. If you have run a holdout, use that figure; if not, treat the preset as a hypothesis to test.
- Read the verdict and actCheck incremental ROAS, real profit, and your break-even incrementality, then use the table to see how the answer changes if your true lift is higher or lower. Export it for the next budget conversation.
RGM Expert Says
Retail media is the fastest-growing line in most consumer budgets, and also the easiest to fool yourself with. The networks attribute generously, so a sponsored-products campaign that mostly intercepts shoppers already reaching for your brand can post a four or five times ROAS while adding almost nothing to the top line. We built this calculator because the first question we ask a retail media report is never “what was the ROAS” — it is “how much of this would have happened anyway.”
The number that changes decisions is the break-even incrementality. Once a client sees that their spend needs, say, 48% of attributed sales to be genuinely incremental just to cover product cost and media, the conversation shifts from chasing reported ROAS to defending real margin. Branded sponsored products and loyalty-heavy audiences are usually where the leakage hides; new-to-brand placements and off-site inventory are usually where the incremental growth is.
Use this to size the prize before you invest in measurement, then close the loop with a real test. A geo holdout or a PSA control gives you a defensible incrementality figure, and once you have it, this tool turns it into the profit and break-even your finance team will actually trust. The estimate gets you a smart hypothesis; the holdout gets you the truth.
How it works
The arithmetic is deliberately transparent. Reported ROAS is what the dashboard shows; everything else strips it back to what the ads truly earned.
- Attributed sales — the revenue the retailer credits to your ads for the period.
- Incrementality — the share of those sales that would not have happened without the ad; the one input worth measuring properly.
- Break-even incrementality — the minimum incrementality at which profit reaches zero. Above it you make money; below it you are subsidizing existing demand.
Reported ROAS and the incremental adjustment are standard; the break-even incrementality is RGM’s framing of the same identity. Measure incrementality with a holdout rather than trusting platform attribution — see the academic case in Gordon et al., Marketing Science (2019).
Why reported ROAS quietly overstates retail media
Retail media networks attribute a sale whenever a shopper who saw or clicked an ad later buys, which sweeps in a great deal of demand the brand already owned. For a category leader bidding on its own branded terms, much of that “return” is simply demand the shopper arrived with. The result is a dashboard ROAS that looks like growth but is partly an accounting artifact.
The fix is incrementality measurement, and the discipline is the same one RGM applies everywhere: hold out a comparable audience or set of geographies, run the campaign everywhere else, and measure the difference. Industry and academic work on advertising measurement has shown repeatedly that platform-attributed return can overstate true causal lift by a wide margin, which is why a measured holdout beats any attribution model.
This matters most as budgets scale. A program that is comfortably profitable at 80% incrementality can be deeply unprofitable at 40%, and you cannot tell which world you are in from the ROAS alone. Knowing your break-even incrementality tells you how much measurement error you can absorb before the spend stops paying for itself.
Reported ROAS is not incremental ROAS
Reported ROAS varies widely by network and placement; these are rough public ranges, not targets. Your incrementality — and therefore your real return — is almost always lower than the dashboard figure, especially on branded sponsored placements.
| Network / placement | Default incrementality (RGM) | Incrementality watch-out |
|---|---|---|
| Amazon Ads | ~55% | High branded-term & Subscribe & Save overlap |
| Walmart Connect | ~60% | Mid-funnel on-site search; mixed |
| Instacart | ~72% | In-basket; ~45–60% new-to-brand |
| DoorDash / Uber Eats | ~66–68% | Marketplace discovery; confirm with Sales Lift |
| Best Buy / Chewy | ~64% | Considered-purchase; conquest can be incremental |
| Kroger / Albertsons / Target | ~58–62% | Loyalty audiences include planned buyers |
| Warehouse (Costco / Sam’s) | ~56–58% | Planned bulk trips lower true lift |
| Branded sponsored products | low | Frequently the least incremental placement |
What measurement leaders say
Attribution tells you which ad got the last touch; incrementality tells you which ad caused the sale. Only one of them belongs in a budget decision.
Platform-reported return can overstate true causal lift substantially; randomized holdouts remain the dependable way to measure advertising’s real effect.