by Om RathodIf you've ever asked what is return on ad spend ecommerce sellers actually track day to day, the answer is simpler than most dashboards make it look. ROAS is the revenue you generate for every dollar you put into ads. That's it. It's expressed as a ratio, like 4:1, or a percentage, like 400%, and either way it's telling you the same thing: how hard your ad dollars are working.
The formula:
ROAS = Ad Revenue / Ad Spend
Say you spend $10,000 on a campaign and it drives $40,000 in attributed revenue. That's a 4x ROAS. Simple math, but the number gets messy fast once you're pulling it from multiple platforms with different attribution rules (more on that later).
Ecommerce brands lean on ROAS harder than a typical retail business ever could, because the path from ad click to purchase is trackable end to end. A billboard can't tell you who bought a car because of it. A Meta ad can tell you, with reasonable confidence, that a specific $47 spend led to a specific $190 order. That direct line is why ROAS became the default health check for DTC marketing teams, sometimes to a fault, since a single ratio can hide a lot of what's actually happening underneath.
Calculating ROAS starts with two numbers from two different places, which is exactly where things start to go wrong for most teams.
Ad spend comes from Meta Ads Manager or Google Ads, pulled at the campaign or account level. Revenue comes from Shopify or GA4, and it needs to be revenue you can actually attribute back to that spend, not just total store revenue for the day.
Platform-level ROAS looks at one channel in isolation. Meta reports its own ROAS based on its own attribution model. Google does the same. These numbers rarely agree with each other because each platform likes to take credit for the same sale.
Blended ROAS is total revenue across all channels divided by total ad spend across all channels. It's a rougher number but harder to game, since it doesn't rely on any single platform's attribution logic.
One mistake shows up constantly: mixing gross revenue with net revenue. If you're calculating ROAS off gross sales but a chunk of that revenue gets refunded or discounted a week later, your ROAS was never real to begin with. Decide up front whether you're using gross or net, and stay consistent.
This is the exact reconciliation problem that eats hours every week for teams stitching together spreadsheets from four different exports. Trivas pulls ad spend and revenue automatically across Amazon, Shopify, Meta, and Google into one BI reporting view, so the blended number and the platform-level numbers sit side by side without a manual export in sight.
There's no universal "good" ROAS. Anyone who gives you a flat number without asking about your margins is guessing.
Rough ranges: a 2x-3x ROAS can be break-even or worse for a low-margin brand. Brands running typical 50-60% gross margins usually need something closer to 4x+ to be genuinely profitable after ad spend, not just revenue-positive.
Margin and AOV change the math completely. A $200 AOV brand with 70% gross margin has a lot of room, and can run profitably at 2x ROAS because the dollar amount of margin per order is high even at a thin ratio. A $30 AOV brand with 30% margin has almost no room. That brand might need 6x or higher just to cover CAC and still make money.
Channel matters just as much as margin. Branded search ROAS routinely runs 10x+ because you're capturing demand that already exists, people searching your brand name are close to buying anyway. Cold prospecting on Meta is a different game entirely. A 1.5x-2x ROAS on top-of-funnel Meta prospecting can still be the right call, because that spend is building the audience that eventually converts through retargeting and branded search at a much higher ratio. Judging a prospecting campaign by the same benchmark as branded search is one of the more common ways teams talk themselves out of growth spend that's actually working.
These three get used interchangeably in Slack threads and investor updates, and that's a problem, because they answer different questions.
ROAS
ROI
MER (Marketing Efficiency Ratio)
Here's the disconnect: a campaign can post a 5x ROAS and still lose money. If your product costs 40% of the sale price, shipping eats another 10%, and you've got payment processing and returns on top, that "efficient" campaign might barely break even once real costs are factored in. ROAS only ever looks at the top line.
Use ROAS to optimize at the channel or campaign level, day to day. Use MER as your gut check on whether marketing overall is trending healthy or bloated. Use ROI when you need the real answer on whether you're making money, because it's the only one of the three that actually asks that question.
Attribution windows are the quiet culprit behind most inflated ROAS reporting. Meta defaults to a 7-day click, 1-day view window, which means it's crediting itself for purchases that happened a full week after someone clicked, plus purchases from people who never clicked at all and just saw the ad. Shorten that window and the same campaign can look dramatically worse overnight, without spend or sales actually changing.
Relying only on Meta or Google's self-reported numbers is a mistake for the same reason. Both platforms are incentivized to take credit for as much revenue as possible. A source of truth like GA4, or a unified dashboard that reconciles platform claims against actual order data, gives you the number you can actually trust instead of the number each platform wants you to see.
Blending new and returning customer revenue into one ROAS figure is another one that quietly wrecks decision-making. If half your "ROAS" is coming from email flows and retargeting existing customers, your acquisition campaigns look far more efficient than they are. Separate new customer ROAS from returning customer ROAS before you decide whether prospecting spend is working.
Returns and refunds are the last piece, and they're the one most dashboards handle worst. If a $150 order gets refunded three days later but your ROAS calculation never backs it out, you're optimizing off revenue that doesn't exist anymore.
ROAS only matters if it changes what you do with your budget. Start by setting a minimum viable ROAS, the floor number based on your contribution margin below which a campaign is actively losing money. Anything above that floor is a candidate for more spend, anything below it needs to be fixed or cut.
Look at trends, not single days. A daily ROAS number bounces around for reasons that have nothing to do with campaign quality, weekends, payday cycles, a competitor's promo. A 7 or 14-day trend line tells you far more than yesterday's snapshot.
Watch for diminishing returns as you scale. If ROAS drops from 5x to 3x as you double daily spend, that's audience saturation talking. It doesn't mean stop, it means you've likely hit the ceiling of your current audience and need fresh creative or a wider targeting pool before pushing further.
Pairing ROAS with forecasting turns it from a rearview mirror into something you can actually plan with. If you're considering a 20% budget increase, forecasting and simulation tools can project how that increase moves revenue and margin over the next 30-60 days, rather than finding out the hard way that the extra spend just bought a saturated audience. This is where marketing leads asking what is return on ad spend ecommerce metrics can support usually get the most value, since the number itself is only useful in the context of what happens next.
ROAS is a starting point. It tells you how efficiently ad dollars are converting to revenue, nothing more. It won't tell you if you're actually profitable, and it won't tell you which channel is quietly propping up your blended number while another one bleeds cash.
Trivas unifies Amazon, Shopify, Meta, and Google ad data on one Redshift-backed dashboard, so your blended ROAS and your channel-level ROAS reflect what actually happened, not what each ad platform wants to claim credit for. If you'd rather see the math on your own numbers before committing to anything, run them through the ROAS calculator or start a trial to see blended ROAS across every channel in one place.
explore more insights
3 min read
3 min read
3 min read
.d53b12e5.png&w=3840&q=75 "Trivas.ai")