What Is a Good ROAS for Ecommerce?

If you’ve spent any time Googling this, you’ve probably noticed something annoying: every source gives you a different number. One article says 2.87x is the average. Another says 4x is the bare minimum. A third throws out 10x like it’s nothing. None of them are lying, exactly. They’re just measuring different things and calling it the same metric.

Here’s the short version. Across most 2025-2026 industry benchmark reports, the average blended ROAS for ecommerce sits somewhere between 2.5x and 4x, with 2.87x showing up as the most commonly cited figure and the median sitting closer to 2x (meaning half of all ecommerce brands are operating below a 2:1 return). A 4x ROAS is widely treated as a solid, sustainable target, and anything north of 5x is genuinely strong performance for most categories.

But that range is only useful as a starting orientation. Your actual “good” ROAS depends on your profit margin, your average order value, your customer lifetime value, and what stage your business is in. A skincare brand with 65% margins can be wildly profitable at 2x. A dropshipper with 22% margins can lose money at the exact same 2x. Same metric, opposite outcomes.

This guide walks through where those benchmark numbers come from, why they disagree with each other so much, and – more usefully – how to calculate the one ROAS number that actually applies to your store. If you want to skip ahead and run your own numbers, our ROAS calculator and break-even ROAS calculator can do the math while you read.

Why “average ROAS” isn’t really the question you’re asking

Most people who search for a good ROAS benchmark aren’t actually trying to find out what other stores average. They’re trying to answer one specific question: am I losing money on this ad spend?

That’s a margin question, not a benchmark question. ROAS only tells you revenue generated per dollar spent. It says nothing about cost of goods, shipping, payment processing, returns, or overhead. A campaign can post a 5x ROAS and still be unprofitable if your margins are thin enough, and another can sit at 1.8x and be perfectly healthy if your margins are wide and your customers come back.

Benchmarks are still worth knowing, though, because they tell you what’s realistic to aim for in your category and on your platform of choice. They just need to be read as a starting point for context, not a finish line.

ROAS benchmarks for ecommerce (and why they don’t agree with each other)

Pull up five different “average ROAS” reports and you’ll get five different numbers, sometimes for the same year. That’s not sloppy research on anyone’s part. It’s usually one of these five things:

1. Average vs. median. Average ROAS gets pulled upward by a small number of high performers. Median ROAS, which sits around 2x according to recent benchmark data, gives a more honest picture of where a typical store actually lands.
2. Blended vs. new-customer ROAS. A report measuring all revenue (including repeat buyers) will show a much higher number than one isolating first-time customer acquisition. More on this below, because it’s the single most overlooked distinction in ROAS reporting.
3. Attribution window differences. Meta’s default reporting window, Google’s data-driven attribution, and third-party analytics tools each count conversions differently, sometimes by 15-30%.
4. Sample composition. A report built from 35,000 Shopify brands skews differently than one built from 80,000 Meta video ads or a handful of agency case studies.
5. Timeframe. Full-year data, holiday-quarter data, and a single recent month tell three different stories, especially given how much ROAS swings seasonally (covered later in this article).

With that caveat in place, here’s how the numbers tend to shake out when you aggregate across the major 2025-2026 benchmark sources.

Average ROAS by ad platform

Platform	Typical blended ROAS range	Why it lands here
Google Search & Shopping	3.5x-4.5x	Captures shoppers actively searching with purchase intent already established
Meta (Facebook & Instagram)	1.8x-4.0x	Demand-creation platform; interrupts a scroll rather than capturing existing intent, typically running 15-25% below Google for the same brand
TikTok	1.2x-3.5x	Entertainment-first audience; performs well for visual, impulse-friendly categories like beauty and fashion, weaker for considered, research-heavy purchases
Reddit	2.3x-4.7x	An emerging channel that improved sharply following platform changes in late 2025, with lower CPMs than Meta in many verticals

A useful rule that comes up across multiple reports: Google tends to outperform Meta because it’s catching people who already decided to buy something. Meta and TikTok are better at creating that decision in the first place, which is a harder job to measure with a single ratio.

Average ROAS by product category

Category	Typical blended ROAS range	Context
General ecommerce (all categories)	2.5x-4x (avg. ~2.87x)	The broad band most stores fall into before segmenting by niche
Fashion & apparel	2.8x-4.3x	Visually driven and impulse-friendly, though high return rates quietly erode the real margin behind the headline number
Beauty & personal care	3x-4.2x, with top performers above 6x	Strong repeat-purchase behavior and a natural fit for short-form video formats
Health & supplements	2.3x-5.7x	Wide range because subscription-heavy brands can accept a lower first-order ROAS thanks to recurring revenue
Home & garden / furniture	4x-6.7x blended	High average order value inflates the blended figure, but margins are typically thinner and repeat cycles longer
Electronics & tech accessories	1.2x-2.8x	Thin margins, heavy price comparison shopping, and longer research cycles before purchase
Food, beverage & CPG	2.5x-4x	First-order ROAS matters less than repeat-purchase frequency over time

Treat these ranges as orientation, not a scoreboard. If you sell electronics and you’re sitting at 2.6x, you might be outperforming your category even though a fashion brand at the same ROAS would be underperforming theirs.

How to calculate your break-even ROAS (the number that actually matters)

This is the calculation most benchmark articles skip, and it’s the one that turns ROAS from a vanity number into a decision-making tool. Break-even ROAS tells you the minimum return you need just to cover your costs, before a single dollar of profit shows up.

The formula:

Break-Even ROAS = 1 ÷ Profit Margin

Here’s how to get there in five steps.

1. Find your average order value. Total revenue divided by total orders over a clean period (strip out fraudulent or test orders). Our AOV calculator handles this in a few seconds if you’d rather not do it by hand.
2. Add up your per-order costs. Cost of goods sold, shipping, packaging, payment processing fees, and a realistic allowance for returns. People consistently underestimate this last one.
3. Subtract step 2 from step 1 to get gross profit per order.
4. Divide gross profit by AOV to get your true profit margin. A profit margin calculator is useful here if you’re juggling several cost lines at once.
5. Divide 1 by that margin. That’s your break-even ROAS – the floor, not the goal.

Worked example, using a mid-margin DTC product:

Line item	Amount
Average order value	$58.00
Cost of goods sold	$19.00
Shipping	$6.00
Payment processing (≈3%)	$1.74
Returns allowance (≈5% of AOV)	$2.90
Total per-order cost	$29.64
Gross profit per order	$28.36
Profit margin	48.9%
Break-even ROAS	1 ÷ 0.489 ≈ 2.04x

In this example, anything above 2.04x is genuinely profitable. Anything below it is losing money, even if it looks fine on a dashboard. Most operators then add a buffer of 25-40% above break-even to account for overhead, salaries, software, and an actual profit margin worth running a business for, which in this case would put a realistic target somewhere around 2.6x-2.9x. You can sanity-check the whole chain, including what’s left over after ad spend, with our net profit calculator.

What a “good” ROAS actually looks like for three different stores

Numbers mean very little without a business attached to them. Here’s how the same concept plays out across three margin profiles.

Low-margin accessory store (22% margin). A store selling phone cases and similar accessories at a $24 AOV, with a 22% margin after product cost, shipping, and platform fees, has a break-even ROAS of about 4.5x. A lot of stores like this celebrate a 3x ROAS without realizing they’re losing money on every single order. For this margin profile, a real target needs to sit at 5.5x or higher, which is exactly why thin-margin categories show up at the bottom of most industry benchmark tables.

Mid-margin apparel brand (46% margin). A clothing brand running a $74 AOV with a 46% margin sits at a break-even ROAS near 2.17x. A realistic, comfortable target in the 3x-3.3x range leaves enough room for overhead and actual profit, which lines up neatly with the 2.8x-4.3x fashion benchmark range above.

High-margin subscription skincare brand (58% first-order margin). This is where the math gets more interesting. A skincare brand with a $42 AOV and 58% first-order margin has a break-even ROAS around 1.72x on the very first sale. If 35% of those customers subscribe and average 4-5 repeat orders over a year, the brand can intentionally run first-order ROAS as low as 1.4x-1.6x, technically below “break-even” on day one, because the customer lifetime value makes up the difference within a few months. Pairing this with a CAC calculator to compare acquisition cost against that lifetime value is what separates a sustainable subscription model from one quietly bleeding cash.

None of these three businesses should be using the same ROAS target, even though all three are “ecommerce.”

Blended ROAS vs. new customer ROAS: the gap most dashboards hide

Here’s a scenario that trips up a lot of growing brands. A store spends $50,000 on Meta ads in a month and generates $200,000 in total revenue – a 4x blended ROAS that looks excellent on the surface. But if only $80,000 of that revenue came from people who had never purchased before, the new customer ROAS is actually 1.6x.

That gap matters because returning customers convert more easily, spend more per order, and cost less to reach. Blended ROAS quietly gives ad spend credit for sales that loyal customers may have made anyway through email, branded search, or just typing your URL into the browser. It’s not dishonest, exactly, but it answers a different question than “is my acquisition spend working.”

A practical rule worth adopting: if your blended ROAS looks strong but most of that revenue traces back to existing customers, you’re not growing the business as much as the dashboard suggests. You’re retargeting people who already trust you. That’s not a bad thing to do, but it shouldn’t be confused with successful prospecting, and the two should be tracked and benchmarked separately wherever your platform or analytics setup allows it.

Why your ROAS swings throughout the year (and that’s normal)

If your ROAS drops every January, you’re not doing anything wrong. Seasonal swings of 50-60% between Q4 and Q1 show up consistently across benchmark data: ecommerce ROAS commonly peaks around 4-5x during the Black Friday through holiday window, then settles back to 2-2.5x in January and February as CPMs fall but so does buyer urgency.

A few patterns worth knowing if you’re setting targets across the calendar:

• November and December CPMs typically rise sharply (reports cite increases in the 35-41% range during peak holiday weeks), but conversion rates climb enough to often offset the cost. Fashion brands, for example, have reported Meta ROAS climbing as much as 17% during Black Friday and Cyber Monday compared to their yearly average.
• January is usually the cheapest month to acquire customers, with CPMs running roughly 20% below the annual average, even though ROAS itself often looks lower because buyer intent has cooled.
• Discounting during peak season changes your break-even math, not just your revenue. A 20% storewide discount lowers your margin and raises your break-even ROAS at exactly the moment you’re trying to scale spend, so it’s worth re-running the numbers through a discount calculator before locking in a holiday promotion.
• CPM volatility compounds the seasonal effect. Tracking your CPM trend month over month makes it much easier to tell whether a ROAS dip is a real problem or just the calendar doing what it always does.

The healthiest way to read seasonal ROAS data is to compare against the same period last year, not against last month. A 25% drop in early January, right after Cyber Monday demand gets temporarily exhausted, is expected. The same drop lasting into March is a different conversation.

Common mistakes that make a good ROAS look bad (or a bad one look fine)

A handful of mistakes show up again and again when people try to judge their own ROAS against the numbers above.

• Comparing blended ROAS to a break-even calculation done on new-customer math. These are two different questions wearing the same label.
• Treating platform-reported ROAS as ground truth. Google, Meta, and TikTok each use their own attribution windows and will sometimes claim credit for the same sale, inflating the combined total well above what your bank account shows.
• Leaving returns out of the margin calculation entirely. Categories with high return rates, apparel especially, can see effective margin swing by 10-15 percentage points once refunds are properly accounted for. Running your numbers through a return rate calculator before finalizing a target avoids a nasty surprise at month-end.
• Chasing one ROAS number across every campaign type. Prospecting campaigns and retargeting campaigns are not the same job and shouldn’t share a target. Retargeting commonly returns 5x-10x simply because it’s selling to people who already know the brand; judging a cold prospecting campaign by that same bar will make a perfectly healthy campaign look like a failure.
• Assuming higher ROAS always means more profit. Sometimes scaling a campaign brings the ROAS down slightly while total net profit goes up, because volume increased faster than efficiency declined. Checking ROI alongside ROAS (more on that distinction in the FAQ below) prevents over-optimizing for a ratio at the expense of actual dollars earned.

If your ROAS is below target, start here

Once you know your real break-even number, a ROAS that’s falling short stops being a vague problem and becomes a specific one to diagnose.

• Look at average order value first. Bundling, free-shipping thresholds, and post-purchase upsells raise revenue per order without touching your ad budget at all, which directly lowers the ROAS you need to hit profitability.
• Check click-through rate before blaming the audience. A weak CTR drives up cost-per-click, which drags ROAS down even with a perfectly good landing page. Our breakdowns on why CTR might be running low and how to improve CTR without raising spend cover this from the creative and targeting side.
• Watch for ad fatigue. Frequency creeping too high on the same audience quietly erodes performance over time. An ad frequency calculator makes it easy to spot when it’s time to refresh creative before ROAS starts sliding.
• Re-segment new vs. returning customer performance before deciding a channel “isn’t working.” A channel with a weak blended ROAS might be doing the hard, valuable work of acquisition while a different channel is just harvesting existing demand.
• Recalculate your break-even ROAS whenever costs shift. Supplier price increases, new shipping rates, and processing fee changes all move your margin, and your target ROAS needs to move with it. This is worth revisiting quarterly at minimum, not setting once and forgetting.
• If you’re running Google Smart Bidding, know that Target ROAS campaigns need real data to work properly. Google recommends at least 50 conversions in the trailing 30 days before relying on it, and advises setting your initial target at or below your historical performance rather than aspirationally above it.

ROAS will always be a noisy, imperfect metric, and that’s fine as long as it’s treated as one input rather than the whole verdict. The benchmark tables above are a reasonable starting point for context. Your break-even ROAS, calculated from your own margins, is the number that actually decides whether your ad spend is building a business or quietly draining one. If you haven’t run that calculation yet, it’s worth doing before the next campaign goes live, not after.

For more calculators covering the rest of your marketing math, from CPA and CPM to churn and customer lifetime value, the full marketing and advertising calculator library is free to use any time you need to sanity-check a number.

Frequently asked questions

Is a 2x ROAS good for ecommerce?

It depends entirely on margin. For a brand with 50%+ margins, 2x can sit close to or above break-even and be genuinely profitable. For a brand with 20-25% margins, 2x is usually a loss once product costs, shipping, and overhead are factored in. Run your own break-even calculation before deciding whether 2x is a win or a warning sign.

Is a 10x ROAS realistic?

For a sustained, account-wide blended average, 10x is rare outside of branded search terms or warm retargeting to an existing customer list. It does show up regularly at the campaign level, particularly for retargeting segments or low-CPC niches, but treating it as a baseline expectation across an entire account usually sets an unrealistic bar.

What’s the difference between ROAS and ROI?

ROAS measures revenue against ad spend alone (Revenue ÷ Ad Spend). ROI measures profit against total investment, including product cost, fulfillment, and overhead, not just media spend. A campaign can show a strong ROAS and a weak ROI at the same time if the underlying margin is thin. The ROI calculator is worth running alongside ROAS for that reason.

What ROAS should I set for Google Performance Max or Shopping campaigns?

Rather than picking an aspirational number, set your initial Target ROAS at or slightly below your account’s recent historical average, since Google’s own guidance recommends having that real performance data (ideally 50+ conversions in the trailing 30 days) before the bidding system has enough signal to optimize accurately.