A good sous chef tracks cost per plate. They know that if a dish costs $8 to make and sells for $24, the kitchen is running well. But that single ratio doesn’t tell them whether the front-of-house upsell drove the order, whether a competitor’s closure sent extra foot traffic their way, or whether certain menu items only sell because they pull people in for something more profitable. The number is real, but the story behind it is more complicated than it looks.
Return on ad spend works the same way. As a marketing metric, ROAS tells you how much revenue was generated for every dollar spent on advertising. It’s clean, it’s actionable, and it’s one of the most widely used measures of advertising efficiency in the industry. But the number you’re looking at is only as trustworthy as the system doing the measuring, and most marketers are working with platform-reported figures that have some significant blind spots built in.
For brands running multi-channel ad campaigns across Google Ads, Meta, CTV, and beyond, understanding not just what ROAS is but also where it comes from and what it can’t tell you is the difference between making budget decisions with confidence and optimizing toward a number that flatters your ad platforms more than it reflects your actual business performance.
Key takeaways
- ROAS (return on ad spend) measures how much revenue is generated for every dollar spent on advertising, calculated as total revenue generated divided by total ad spend. A 4:1 ratio is a commonly cited benchmark, but good ROAS varies significantly by industry, profit margins, and business model.
- Platform-reported ROAS and modeled ROAS are two different numbers that answer different questions. The figure your advertising platform shows you comes from its own attribution model, which has a built-in incentive to claim credit for as many conversions as possible.
- Ad platforms can and do double-count conversions. When multiple platforms each take credit for the same sale, your reported ROAS numbers can add up to more revenue than your business actually generated, making multi-platform comparisons unreliable without an independent measurement layer.
- A high ROAS does not guarantee a profitable campaign. ROAS focuses solely on revenue relative to ad spend and doesn’t account for profit margins, business costs, fulfillment expenses, or operating overhead.
- Lower-funnel channels like branded search and retargeting will almost always show higher ROAS than upper-funnel awareness campaigns, but that doesn’t mean awareness is underperforming. It means standard ROAS can’t see how awareness campaigns drive the demand that lower-funnel campaigns then convert.
- Halo effects, including organic traffic lifts, branded search increases, and Amazon revenue driven by non-Amazon campaigns, are systematically undercounted in platform-reported ROAS, which leads brands to undervalue the channels most responsible for building demand.
- Modeled ROAS, produced by Prescient’s marketing mix model (MMM), gives you an unbiased, cross-channel view of advertising efficiency by using statistical modeling rather than platform attribution, making it a more reliable foundation for budget allocation decisions and long-term marketing strategy.
What is ROAS and how do you calculate it?
Before getting into the nuances, it helps to start with the ROAS formula itself, because it’s one of those marketing metrics that’s genuinely simple on its face. Return on ad spend is calculated as:
ROAS = Revenue generated by ad campaign รท Cost of ad campaign
So if you spent $10,000 on an advertising campaign and it generated $50,000 in revenue, your ROAS is 5:1, meaning $5 in revenue earned for every $1 in ad spend. That’s the basic ROAS calculation, and it’s the version you’ll see surfaced in most advertising platforms and analytics dashboards.
The 4:1 benchmark gets cited frequently as the threshold for good ROAS, and while it’s a reasonable starting point for many ecommerce businesses, it’s not a universal rule. A brand with healthy profit margins might run a profitable campaign at 2:1. A brand with thin margins, high fulfillment costs, and significant advertising spend might need 8:1 or higher just to break even. Your break even ROAS depends entirely on your unit economics, not on what a benchmark report says.
It’s also worth noting that ROAS as a metric focuses specifically on the relationship between ad spend and the revenue earned from advertising. It doesn’t factor in non-advertising business costs like cost of goods sold, warehousing, or customer service. That’s where a lot of marketing teams run into trouble, particularly when presenting ROAS numbers to finance teams who are thinking about overall profitability rather than advertising efficiency in isolation.
ROAS vs. ROI: what’s the difference?
These two metrics get used interchangeably a lot, but they measure very different things, and confusing them can lead to meaningful misalignments between your marketing team and your leadership.
ROAS focuses solely on the advertising efficiency of your ad spend. It tells you how much revenue your ad dollars are generating, but it doesn’t tell you whether that revenue is actually profitable once you account for all associated costs. ROI, by contrast, calculates the overall profitability of an investment after subtracting the total costs involved, including production, fulfillment, overhead, and anything else that goes into delivering the product or service to a paying customer.
A specific ad campaign can show strong ROAS and still lose money. If your ROAS is 4:1 but your cost of goods is 60% of revenue, your actual return after associated costs is much lower than the top-line number suggests. This is why some marketing teams also track metrics like marketing efficiency ratio (MER), which looks at total revenue divided by total marketing spend across all channels, giving a blended view that’s harder for any individual advertising platform to inflate.
For most day-to-day campaign performance tracking, ROAS is the right tool. For strategic budget decisions and financial reporting, it needs context from broader profitability metrics. The key is understanding which question you’re trying to answer before you pull the number.
What counts as a good ROAS?
There’s no single answer here, which is genuinely frustrating when you’re trying to set a target ROAS for your campaigns. Comparing ROAS across businesses without accounting for their differences in margin structure, channel mix, and business model is a bit like comparing average ROAS across restaurants that serve completely different cuisines at different price points. The number means something different in every context.
Profit margins are the most important variable. A DTC wellness brand with a 70% gross margin has a lot more room to run campaigns at a lower ROAS than an apparel brand operating at 35% margins. Your minimum ROAS threshold, the point below which a specific campaign is losing money, should be calculated from your actual unit economics before you set any bidding strategies or performance targets.
Average order value and customer lifetime value also matter significantly. If your customers tend to make repeat purchases, a campaign that looks borderline on first-order ROAS might be very profitable over a 12-month window. Brands that sell subscription products or have high repurchase rates often make the deliberate choice to accept a lower acquisition ROAS because the long-term revenue justifies the upfront advertising costs.
The role of each channel in your funnel is another factor that shapes what ROAS you should expect. Lower-funnel campaigns that capture in-market demand, like branded search or retargeting, will naturally generate higher ROAS than top-of-funnel awareness campaigns on YouTube or CTV. Comparing these directly when allocating budgets is one of the most common mistakes in marketing strategy, and we’ll go deeper on this in the next section.
When tracking ROAS as a guide for decision-making, the most useful benchmarks are your own historical figures segmented by campaign type, channel, and audience, not an industry average from a third-party report. Good ROAS is the ROAS at which your business grows profitably, and only your own data can tell you what that number is.
Platform-reported ROAS vs. Modeled ROAS: Why the numbers don’t match
This is where the real conversation about return on ad spend gets interesting, and where a lot of brands are quietly leaving money on the table without realizing it.
The ROAS figure you see when you log into your advertising platform comes from that platform’s own attribution model. It’s looking at conversions that occurred within a set window and claiming credit for them based on its own rules. The problem is that every platform runs this process simultaneously, using different attribution logic. Meta claims credit for the customer who saw your Facebook ad. Google claims credit because the customer later searched your brand name before converting. TikTok claims credit for the view from seven days earlier. The result is that your reported revenue across multiple platforms can add up to significantly more than your actual total revenue.
This isn’t a theoretical concern. Some brands discover when they total up their in-platform reporting that attributed revenue across their advertising channels exceeds their actual business revenue. That’s mathematically impossible, which means one or more platforms are taking credit for conversions that other channels also influenced. The incentive structure here is real: ad platforms have every reason to show favorable results because favorable results justify continued ad spend.
Modeled ROAS, a metric we use within our marketing mix model, takes a fundamentally different approach. Instead of relying on conversion tracking or attribution windows, we use statistical modeling to analyze the relationship between your actual ad spend and your actual revenue over time, across all of your advertising channels at once. It doesn’t have a stake in any individual campaign or channel performing well. It’s looking at your whole marketing ecosystem and asking: when we spent more here, what actually happened to revenue?
The two numbers, platform-reported ROAS and Modeled ROAS, will often differ. That gap is useful information in itself. When a platform’s reported ROAS is dramatically higher than the modeled figure, it’s a signal that the platform is overclaiming credit. When they align, you can be more confident that the platform’s attribution is reasonably accurate. Looking at both side by side gives you a more honest view of advertising efficiency than either number provides on its own.
How ROAS varies by channel and why that matters for allocating budgets
Not all advertising channels are built to close sales, and measuring all of them against the same ROAS benchmark is one of the most reliable ways to make bad budget decisions.
Channels that sit closer to the point of purchase, like branded search, retargeting, and email, will almost always generate higher ROAS because they’re capturing demand that already exists. Someone who searches your brand name on Google and clicks through to buy was probably already close to converting. The search campaign gets the credit, but the awareness campaign that introduced them to your brand weeks earlier did the harder work of creating that demand in the first place. This is the dynamic that makes ROAS a misleading signal when you’re looking at it by channel in isolation.
Upper-funnel channels like CTV, YouTube, and paid social prospecting tend to show lower direct ROAS because they’re reaching people who aren’t yet in-market. Their job is to plant the brand in someone’s mind so that when they are ready to buy, you’re the brand they think of. Standard conversion tracking can’t see that mechanism, so those advertising campaigns consistently appear to underperform relative to what they’re actually contributing to your revenue.
This matters a lot for budget allocation. If you use ROAS alone to decide which ad campaigns to scale and which to cut, you’ll systematically defund the campaigns that build demand while over-investing in the campaigns that harvest it. Eventually you run out of demand to harvest, ROAS on your lower-funnel campaigns starts declining, and the reason isn’t obvious from the data.
The right approach is to understand what each campaign type is supposed to do and evaluate it against goals that match its role. ROAS is the right signal for lower-funnel performance. Reach, brand search lift, and halo revenue are more appropriate signals for upper-funnel evaluation. Blending these rather than applying a single ROAS target across your entire marketing strategy gives you a far more accurate picture of where your advertising dollars are actually going.
The halo effect problem: what standard ROAS can’t see
One of the most significant gaps in platform-reported ROAS is its inability to account for halo effects, the downstream revenue that an advertising campaign generates through channels other than the one running the specific ad.
When someone sees your Meta prospecting ad, doesn’t click it, and then searches your brand name on Google three days later and converts, your Meta campaign gets zero credit in standard reporting. Your Google branded search campaign gets all of it. But the Meta campaign created the awareness that made the Google search happen. Remove that awareness spend and your branded search performance would look very different. This is a version of the entire customer journey that platform attribution simply cannot reconstruct.
The same dynamic plays out for Amazon. A brand running CTV or paid social campaigns can drive significant revenue to their Amazon store through customers who discovered the brand via those ads and then went to Amazon to purchase, because that’s where they prefer to buy. None of that revenue shows up in the CTV or social ROAS figures because those ad platforms can’t see what happened on Amazon. The campaign looks like it underperformed when it was actually doing exactly what it was supposed to do.
These halo effects are real, measurable, and often substantial. For brands running significant upper-funnel spend, the revenue that surfaces through organic traffic, branded search, and direct conversions as a result of paid campaigns can rival or exceed the directly attributed revenue. Ignoring them doesn’t make them disappear. It just means your ROAS calculations are leaving a large portion of the value your campaigns create off the books.
The limitations of ROAS as a standalone metric
ROAS is a useful marketing metric, but it has real limitations that matter when you’re using it to guide major advertising decisions. It’s worth knowing what it can’t tell you before you let it drive your strategy.
It doesn’t account for profit margins or business costs. A campaign generating 6:1 ROAS might still be unprofitable once you factor in cost of goods, shipping, customer acquisition cost, and overhead. The revenue earned from advertising is not the same as profit, and treating it that way is a surprisingly common and costly mistake.
It can’t tell you whether you’re hitting saturation on a specific ad campaign. Every campaign has a point at which additional ad dollars start generating diminishing returns. Without saturation curve data, a brand can keep scaling a specific campaign past its efficient range and see ROAS decline gradually, without understanding that the campaign has simply reached its ceiling at that spend level.
It doesn’t distinguish internal performance from external factors. When ROAS changes, it could be because your ad creatives improved, because a competitor pulled back spend, because the season shifted, or because of any number of external variables that had nothing to do with your advertising efforts. Standard ROAS reporting can’t tell you which of those things happened, which makes it hard to draw reliable conclusions from performance changes when tracking ROAS over time.
And as discussed, it systematically undercounts revenue driven by upper-funnel campaigns and cross-channel halo effects, which creates a structural bias toward valuing lower-funnel, last-touch activity over everything that happened upstream. For brands making budget decisions based purely on reported ROAS, this bias is invisible but consequential.
How to improve ROAS: Where most strategies start and where they fall short
The standard playbook for improving ROAS involves a few well-established levers: tightening audience targeting, testing different ad creatives, refining bidding strategies, cutting underperforming campaigns, and shifting budget toward higher-performing advertising channels. These are all legitimate tactics and they’re worth doing. But they operate within the limitations of your measurement system, which means they can only improve what you’re able to accurately see.
Tracking ROAS at the individual campaign level rather than the channel level is one of the most impactful improvements most marketing teams can make without changing their measurement stack at all. Channel-level ROAS averages out performance across every campaign running within that channel, which can obscure significant variation. A Google Ads account running five campaigns might have two generating strong returns and three dragging the average down. Channel-level ROAS doesn’t surface that. Specific ad campaign ROAS does.
Beyond tactical optimization, improving ROAS is also a function of improving how you measure it. If you’re only looking at platform-reported figures, you’re optimizing toward numbers that each platform produces about itself. Adding an independent measurement layer, like a marketing mix model, gives you ROAS figures that aren’t shaped by any single platform’s attribution logic. That’s often when brands discover that the campaigns they thought were underperforming were actually driving more revenue than the platforms were crediting, and the campaigns with the highest reported ROAS were overclaiming.
Better measurement isn’t just a diagnostic tool. Brands that understand their true advertising efficiency across all channels can allocate advertising budgets based on actual performance rather than platform-reported performance, which compounds over time into meaningfully better outcomes. It’s one of the more durable competitive advantages available to performance marketing teams.
Where Prescient comes in
The ROAS number in your ad platform is a starting point, not a verdict. Prescient AI’s marketing mix model gives you a second number: Modeled ROAS built from the statistical relationship between your actual ad spend and your actual revenue, with no pixel dependency and no advertising platform bias. You can compare both figures side by side directly in the Prescient dashboard, which makes it easy to see where your platforms are overclaiming and where the numbers align.
Prescient attributes revenue at the campaign level, not just the channel level, so you can see which specific ad campaigns are driving base revenue through direct engagement with your ads and which are generating halo revenue through organic traffic, branded search, direct traffic, and Amazon store activity driven by your off-Amazon campaigns. The campaigns that show low ROAS in your platform reports are often the ones quietly generating the most value in channels the platforms can’t see. Prescient makes that visible.
Beyond attribution, Prescient surfaces saturation curves and confidence scores at the campaign level. That means you know not just what your current ROAS is, but whether you’re spending at an efficient point on the curve. A campaign operating below its saturation point may have room to scale profitably. A campaign past its saturation point is generating diminishing returns, and Prescient shows you exactly where that threshold is. That turns ROAS from a backward-looking report card into a forward-looking tool for making smarter, faster budget decisions.
Daily model updates mean you’re not waiting weeks for the picture to clear. As your ad spend changes and your campaigns evolve, Prescient’s model reflects those changes in near real time, giving you a measurement foundation that’s built for the speed at which performance marketing teams actually need to operate. Book a demo to see the platform for yourself.
Return on ad spend FAQs
What is a good ROAS for my industry?
There’s no universal benchmark that holds across industries, because good return on ad spend (ROAS) is fundamentally a function of your profit margins and business model rather than your category. A brand with 70% gross margins can be profitable at 2:1 ROAS. A brand with 30% margins might need 7:1 or higher to justify the advertising spend. The most reliable way to set a meaningful ROAS target is to work backward from your unit economics: calculate your minimum return on ad spend first, then use your historical campaign data to understand what high-performing and low-performing campaigns actually look like for your specific business. Industry benchmarks are a starting point for curiosity, not a basis for strategy.
Why is my platform ROAS higher than my actual revenue growth would suggest?
This is one of the most common signs of multi-platform attribution overlap. When multiple advertising channels each claim credit for the same conversion, the sum of reported revenue across platforms can significantly exceed your actual revenue. It doesn’t necessarily mean your advertising isn’t working. It means each platform’s attribution model is crediting itself for sales that were influenced by multiple touchpoints across the entire customer journey. An independent measurement approach like an MMM can help you understand what’s actually driving conversions by analyzing the statistical relationship between your ad spend and your total revenue, rather than relying on each platform’s self-reported figures.
What’s the difference between ROAS and MER (marketing efficiency ratio)?
Return on ad spend (ROAS) measures the revenue efficiency of a specific ad campaign or channel relative to that campaign’s ad costs. MER, or marketing efficiency ratio, measures total revenue divided by total marketing spend across all of your channels, including spend that doesn’t have direct conversion tracking. MER is a blended metric that’s harder for individual platforms to inflate because it looks at your whole business rather than isolated campaigns. Many brands use MER as a top-line health check alongside campaign-level ROAS, because MER reflects the cumulative impact of all your advertising spend rather than the sum of each platform’s individually claimed results.
How do I measure ROAS for upper-funnel or awareness campaigns?
Standard conversion tracking is a poor fit for awareness campaigns because it can only see revenue that converts within a direct click path, and awareness campaigns rarely drive purchases that way. The more accurate approach is to measure the downstream effects of awareness spend: branded search lift, organic traffic changes, and direct revenue increases that correlate with awareness campaign activity. A marketing mix model is particularly useful here because it can isolate the contribution of upper-funnel spend to total revenue even without direct conversion data, which gives you a much more complete view of what your awareness campaigns are actually doing for your business.
Can ROAS go down even when my marketing is actually improving?
Yes, and this happens more often than most marketers expect. As you scale spend on an ad campaign, you typically reach a saturation point where additional dollars generate incrementally less revenue. The campaign hasn’t stopped working; it’s just operating in a less efficient range of the spend curve. Return on ad spend (ROAS) will decline even if the campaign is still generating significant, profitable revenue. External factors can also affect ROAS independently of campaign quality: a competitor pulling back spend might improve your ROAS without any change in your advertising, while a competitor ramping up might suppress it. Tracking ROAS in isolation, without understanding saturation dynamics and external context, can lead you to cut campaigns that are performing well and scale campaigns that are past their peak efficiency.
How does an MMM calculate ROAS differently than a platform?
Ad platforms calculate ROAS by attributing conversions to touchpoints within a defined window, using their own rules about which interactions get credit and how much. This means the figure they report is shaped by their attribution methodology and, to some extent, their incentive to show favorable results. A marketing mix model calculates ROAS by modeling the statistical relationship between your actual ad spend and your actual total revenue over time, across all channels simultaneously. It doesn’t rely on pixel-based conversion tracking or attribution windows, and it doesn’t have a stake in any particular channel performing well. The result is a modeled ROAS that reflects how your spend actually moved your revenue, which is often a different and more accurate number than what your platforms report.
Should I use ROAS or ROI to evaluate my campaigns?
Both, but for different purposes. ROAS is the right metric for day-to-day campaign performance tracking because it’s specific to your advertising spend and easy to calculate at the individual campaign or channel level. ROI is more appropriate for strategic budget decisions and financial reporting because it accounts for the full cost of generating that revenue, including cost of goods, operating costs, and overhead. A campaign with strong ROAS can still have negative ROI if the margins aren’t there to support it. Using ROAS for campaign optimization and ROI for business-level profitability analysis gives you the clearest picture of both how your advertising is performing and whether it’s actually making your business money.
