What Is a Good ROAS? The Real Answer by Channel and Business Model

A good ROAS is the ROAS at which your business is profitable after accounting for the cost of goods, overhead, and the specific way your business model generates revenue. That number is different for every business, which is why benchmarks like "3x ROAS is good" or "you need at least 4x" are nearly useless without context.

The question worth asking isn't "what's a good ROAS" — it's "what's the minimum ROAS I need to run a profitable paid channel, and what ROAS do I need to reach my margin target?"

The Minimum Viable ROAS Formula

Your minimum viable ROAS (the floor below which paid media loses money) is a function of your gross margin:

Minimum ROAS = 1 / Gross Margin

If your gross margin is 70% (revenue minus the direct cost of delivering the product), your minimum viable ROAS is 1 / 0.70 = 1.43x. Below 1.43x, every dollar you spend in advertising costs you more in gross profit than you're recovering.

If your gross margin is 40%, your minimum viable ROAS is 1 / 0.40 = 2.5x. Below 2.5x, you're losing money on gross profit before paying a single employee salary or software subscription.

This math clarifies why software companies (often 75-85% gross margins) can tolerate much lower ROAS targets than direct-to-consumer companies (often 40-60% gross margins). A 2x ROAS that works for a SaaS business would destroy a DTC brand.

Why Many Teams Are Running at the Wrong ROAS Target

The most common mistake is using revenue ROAS (what the ad platform reports) when business decisions need to be made on gross profit ROAS or contribution margin ROAS.

A campaign that shows 5x ROAS in Meta Ads Manager looks healthy. But if 30% of purchases are refunded within 30 days, the real ROAS after returns is closer to 3.5x. If gross margin is 45%, the contribution margin from that 3.5x ROAS is $1.575 per $1 spent — before any overhead. If overhead (salaries, platform fees, creative costs) eats another 30% of revenue, the business is actually losing money on a campaign that looks like 5x ROAS.

Final Round AI encountered this problem before switching to AIMA for media management. Their Google UAC campaigns were showing 3.8x ROAS in the platform. After accounting for LTV cohort data (users who purchased via those campaigns had 8% lower 90-day retention than organic), the real ROAS on profitable users was closer to 2.9x — below their breakeven threshold. They'd been scaling a losing channel because platform-reported ROAS didn't account for cohort quality.

ROAS Benchmarks by Channel (With Context)

These are real-world ranges, not targets. Use them only for orientation, not as goals.

Meta Ads (social): 1.5x to 4x for most e-commerce and subscription businesses. Higher for lifestyle categories where social proof works well. Lower for technical B2B products. The reason for the wide range: Meta's attribution inflates ROAS by crediting view-through conversions for users who would have converted organically.

Google Search (intent-based): 4x to 10x for high-intent commercial terms. Google Search typically shows higher ROAS than social because users are actively searching for what you're selling. The floor is higher because CPCs are higher.

Google UAC / Apple Search Ads (app install): 2x to 6x measured at 30-day LTV window. The wide range reflects differences in app category, monetization model, and how well LTV is measured.

TikTok: 1x to 3x for DTC brands with strong creative. Lower ROAS floor than Meta because CPMs are lower, but conversion rates are also lower. TikTok ROAS inflates significantly with view-through attribution.

CTV / Connected TV: Often shows negative or near-zero direct ROAS because it's an awareness channel with difficult attribution. Brands using CTV typically measure it via brand lift studies or incrementality testing rather than direct ROAS.

Why Blended ROAS Hides the Real Problem

Blended ROAS (total revenue / total ad spend across all channels) makes channels look better than they are when organic and brand are included.

A company doing $1M in monthly revenue with $200K in total ad spend has a 5x blended ROAS. But $600K of that revenue is direct traffic and organic search, with $400K coming from paid. The actual paid ROAS is $400K / $200K = 2x, which may be below their minimum viable threshold.

Blended ROAS is a useful sanity check but a terrible optimization signal. Channel-level ROAS — and ideally, incremental ROAS — is what decisions should be made on.

Setting the Right ROAS Target

Work backward from your margin targets instead of forward from benchmarks:

1. Define your gross margin (product revenue minus direct COGS)
2. Calculate your minimum viable ROAS (1 / gross margin)
3. Add your target contribution margin — if you want 20% contribution margin after ad spend, your target ROAS = 1 / (gross margin - target margin %)
4. Add a buffer for attribution inflation (platform-reported ROAS is typically 10-30% higher than true ROAS due to attribution model bias)
5. The resulting number is your real target ROAS for each channel

Managed Growth tracks gross-profit-adjusted ROAS by channel and cohort rather than relying on platform-reported revenue figures, which gives a more accurate picture of whether paid channels are actually profitable.

Subscription App ROAS and the LTV Window Problem

For subscription apps, ROAS measured at one point in time can look very different from ROAS measured at another. A user who installs and subscribes in month one generates immediate revenue. Whether they renew at month three, month six, or cancel after the first billing cycle changes the actual ROAS of the campaign that acquired them.

The window you choose to measure ROAS determines which campaigns appear effective. Short-window ROAS (7 to 30 days) rewards campaigns that drive fast first purchases. Long-window ROAS (90 to 180 days) rewards campaigns that find users who retain.

These two do not always correlate. A campaign driving high-intent users with fast first purchases may have mediocre retention, producing strong 7-day ROAS and poor 90-day ROAS. A campaign driving users who take a free trial before subscribing looks weak at 7 days but often shows strong 90-day ROAS because trial users who convert tend to retain better than direct purchasers.

Most teams make budget decisions on the wrong window because it is the only one where data is immediately available. The practical solution: use short-window ROAS for day-to-day campaign optimization, but validate channel-level budget allocation decisions against 90-day cohort data at least monthly. If a channel's 7-day ROAS is consistently strong but its 90-day user retention chronically underperforms baseline, scaling that channel based on short-window data means scaling a problem.

ROAS and Incrementality

Platform-reported ROAS and incremental ROAS are not the same number, and the gap between them varies significantly by channel and campaign type.

Retargeting campaigns typically show large gaps. The retargeting audience is already predisposed to convert because they have prior intent. When you run an incrementality test (a holdout where 15 to 20% of the retargeting audience sees no ads), conversion rates in the holdout group often run 60 to 80% of the rate in the exposed group. The actual incremental lift from retargeting spend is 20 to 40%, not the 100% that platform ROAS implies.

Brand search campaigns often show near-zero incremental ROAS for established brands because users typing the brand name were already planning to visit the site. The brand search campaign intercepts organic traffic and claims the conversion credit, inflating platform ROAS without generating any additional revenue.

Prospecting campaigns typically show the opposite: lower platform ROAS but higher incremental ROAS than the platform number suggests. Top-of-funnel touchpoints rarely receive last-click credit, so the platform underreports their contribution even though they initiated journeys that converted months later.

Understanding where your channel ROAS diverges from incremental ROAS is what allows you to set realistic expectations for what you will actually gain by scaling each channel.