Go to homepage

Table of Contents

What is ROAS?

ROAS stands for Return On Ad Spend.

The ROAS Formula = Gross Sales / Ad Spend

ROAS is the most commonly used KPI that performance marketers are measured on, especially in e-commerce.

ROAS grew popular in an era when tracking via cookies was far more efficient than it is today. Therefore, it was a great KPI for e-commerce companies to measure the efficiency of their paid marketing efforts in channels like Google ads and Facebook ads.

The popularity of ROAS was also due to the fact that tracking sales on an e-commerce website was relatively easy for all ad platforms compared to any other value-creating metric like profit, Customer Lifetime Value, etc.

The difference between ROAS and blended ROAS

  1. The first and the original way is only to include Sales attributed to ads.
  2. The second is to look at "blended ROAS," which means the total Gross Sales divided by the total ad spend.

What is the difference between ROAS and Cost of Sales?

Very little. At least not blended ROAS and Cost of Sales. It's just the inverse.

The cost of sales formula: Marketing Spend / Gross Sales

So, it's a different way of looking at it and answering a slightly different question.

ROAS answers, "How much more revenue do I get if I spend one dollar on ads?"

Cost Of Sales, or COS, answers, "As a % of sales, how much marketing have we spent?"

If you look at ROAS, not the blended ROAS, the inverse is ACOS, which stands for advertising cost of sales. That is not the regular cost of sales but only the sales attributed to the ads.

The different misaligning risks caused by ROAS

Different scenarios mislead the actual effectiveness of ad spending. It's often about over-investing ad spending in low-margin products or markets while ignoring the non-marketing costs.

The most common examples of when a ROAS can be misleading

  1. Low-margin products might lead to better ROAS.
  2. Low profitability orders, meaning low AOV, might lead to higher ROAS
  3. Sales in countries with higher shipping costs might have higher ROAS.
  4. Best sellers get higher ROAS than slow movers, leading to early stockouts of the best sellers and slow movers with too much stock.

A real-world example of how bad ROAS can be

An e-commerce company with a high gross margin and worldwide shipping once asked me to analyze its profitability. Despite its success, the company sensed its marketing could be more effective, relying on a performance marketing agency driven by Return on Ad Spend (ROAS). This agency, with a standard analytics setup, was not mismanaging its duties; it was simply maximizing ROAS, unaware of the broader implications.

When I dug into the numbers, I found what’s usually the case for these types of companies. While 80% of their marketing budget was efficiently allocated, achieving targeted ROAS across prioritized markets, it was the remaining 20%—the "Rest of the World Experiment Budget"—that caught my attention. This segment had much higher ROAS, seemingly benefiting the company's overall metrics by tapping into core customers worldwide. However, this success masked a costly oversight.

Delving deeper, I analyzed the average shipping costs to these experimental markets, revealing that shipping to 18 out of 20 countries was so expensive that the company lost money on each order without considering the marketing spend. Since there was such a low volume in these markets, so they’ve been overlooked, with no negotiated shipping terms, leading to too high costs.

Further examination showed that the actual return on ad spend, or what we at Dema call "epROAS" (the definition of epROAS can be found here), was less than 20% of the marketing expenditure in these countries. The high shipping costs rendered the orders unprofitable despite the seemingly favorable ROAS.

After this analysis of the clear difference between perceived and actual profitability, the company's way of working with the agency and ROAS changed. Now, the agency gets more profit data and is steered on epROAS.

This experience underscored the importance of looking beyond surface-level metrics, also known as Vanity Metrics, to understand the true impact of marketing strategies. It was a harsh reminder that in the world of e-commerce, profitability killers are everywhere, and to find them, you need a comprehensive analysis that includes all operational costs.

Aligning Marketing and Company Interests by steering on epROAS

When understanding that ROAS might not be the great metric to measure your marketing spend's efficiency, you might wonder, "So what should I use instead of ROAS?"

The short and easy answer is epROAS.

The more honest answer is that it depends and is not that easy anymore. You can't optimize for only one metric, and you won't be able to track everything; therefore, you need more advanced tools to shape your analytical skills to ensure profitable and sustainable growth. You typically need to work with Marketing Mix Modeling, which we've written another blog article about here.

But first things first - let's go through the necessary shift from ROAS to epROAS

All the cases when ROAS is sub-optimizing marketing spend and make your marketing dollars be invested in profitable alternatives can easily be solved if you start to evaluate the marketing (and buying/product team) on epROAS and Net Gross Profit 3, also known as Contribution Margin 3. Read more about Contribution Margin 3 in this blog article here.

But is it really that simple that this whole problem vanished just by starting to evaluate on epROAS?

In part, yes. It's not more complicated than that. However, the tricky part is getting to a point where the marketing and the whole commercial team, including buying/product teams, have this data and can follow it in real time and at a granular enough level. But that's also when you can choose to just plug into Dema and be up and running in no time.

To prove the point here, let's take a simplified example (see table below).

Product A is sold for 100. We have a ROAS target of 500%. For whatever reason, the product then gets discounted by 30%, and it starts to convert very well and is picked up by the Google algorithm in a performance max campaign, resulting in a very high ROAS and high sales. The gross margin worsened when applying the discount, and the average order value also decreased, i.e., people didn't buy more per order as a result of the discount. This results in less Gross Profit 1 to cover the exact fulfillment cost per order (shipping cost, etc).

As you can see, ROAS goes up, and epROAS goes down. This is the easiest way of showing that steering the marketing team on epROAS is better for sustainable and profitable growth than trying to "keep it simple" and let the marketers work with ROAS and then try to change that ROAS-target depending on all possible variables that affect the gross profit 2.

Implementing epROAS for Effective Decision-Making

So, what do you need to get started with measuring epROAS and ensure aligned teams toward profitable growth for your e-commerce business?

You need

  1. All order data, not only the orders tracked by the script on your website
  2. All inventory data to map purchasing cost (COGS) to all orders
  3. All marketing costs per channel daily (to make it more actionable, it should be down on ad, product, and keyword level)
  4. Detailed averages of all fulfillment costs per market, per shipping provider, per payment provider, product, and order, as granular as possible.
  5. Detailed averages of all returns and the associated costs per country and product.

You need all these data points connected. If they're not, the commercial team will have difficulty acting on them. Let's say you only get the input that epROAS is down from 150% to 75% in a specific market—nothing more. What can you do? Well, you could move spending to another market, but you also want to know, "Why is it down?".

That's why you need a proper data warehouse setup that supports this way of working toward profitable growth. While you can invest heavily in setting up your own data warehouse, you should always look at the alternative cost to that. What could we do with the same time and money? It takes a lot of time and requires a lot of talented data engineers and data scientists and a bunch of tools to build and maintain a data warehouse that is optimized for analyzing the commercial side of e-commerce.

Consider Jeff Bezos's mantra: "Focus on the things that make your beer taste better." An analogy is taken from the history of German beer makers who understood that outsourcing power generation was probably better than building up capacity and competence to start generating that energy themselves. So, just because you desperately need something (because you do if you don't have this setup today), it's not always the best thing to build it yourself.

An analytics platform like Dema could align your team around epROAS and contribute with other necessary metrics and possibilities to understand and improve your short and long-term profitable growth. Book a demo today if you are interested in understanding how.


  1. Switch from steering your marketing team on ROAS to aligning the marketing team with the product/buying team and steering the whole commercial team on epROAS.
  2. As Jeff Bezos famously said, get your data in order, but spend your time and money on “...the things that make your beer taste better”. Don’t think your e-commerce company needs to hire a big team of data engineers and data scientists; there are solutions, like Dema.ai, that will have you up and running with epROAS in zero time.


Do you like what you read?

Then, you will love Dema's platform. Or at least, it will speak to you, and it is worth a demo.

Book a demo

Want to read about the top 1%?

The e-commerce company Ridestore has used Dema to improve its short—and long-term profitable growth. Read the Case Study to learn more.

Learn how Ridestore uses Dema