Why ROAS is Not a Useful Metric

More often than not, we engage with seasoned ecommerce professionals tasked with running print programs while simultaneously holding some measure of responsibility for Profit & Loss statements. And as such, they tend to look at performance metrics with a ROAS (Return on Ad Spend) lens instead of a more critical metric – Contribution Per Order (CPO). Why is that important, and why does ROAS lack adequate insight as a key metric? Let me explain.

While ROAS is an efficient measuring metric, it cannot assign health to a channel, making it a relatively useless tool. The reality is it could be hurting your business as a primary form of indicator.  ROAS can’t tell you how much money you are making for every order received. As an example, a ROAS of 6x is more efficient than a ROAS of 5x, but that doesn’t mean you made more money (or any money). A ROAS of 6x on $10 did not make you more money than a ROAS of 5x on $10,000,000. The volume grew top-line demand, but it did not increase profits.

In contrast to ROAS, Contribution Per Order (CPO) brings in the magnitude of the campaign segment targeted (marketing cost and cost of goods). This metric will help with your payroll and to keep the lights on – in other words — adding top-line demand and bottom-line profits.

Regardless of which channel you operate in, each marketer must test to determine their organic demand levels and then know how each program contributes to revenue. ROAS can’t tell you that, and in fact, ROAS can only tell you have a slightly positive ROAS (but you may still be losing money as the remarketing bar below indicates!)


If you are a CMO and are solely relying on ROAS, how do you determine the channels that provide the best contribution or profits? In my chart above, Branded search really can’t be expanded in most cases because it relies on other efforts to grow beyond the organic level of demand that already exists.  (NOTE: be sure to separate brand vs. non-brand to gauge the performance because branded demand will skew performance if not separated, an article saved for another day!) Non-Branded search provides a decent 2.84 ROAS, but on a contribution basis, print prospecting performs nearly double the contribution. Meaning that for every order received, I’m making $20 more per order, which annually equates to adding approximately $100,000 in net revenue from this one segment alone! Again, in this example, I would be mining my database to engage with lapsed customers ($69.73 vs. $25.35) and mail deeper into all of these database segments. When combining those efforts, the CPO orders generate an average of $58.59 per order vs. $25.35 per order for non-brand search terms. In this example, I would be adding over $1,100,000 in net revenue in one year.

Admittedly, I’m only giving you one recent example. The exact opposite may be happening in your business, where digital marketing is outperforming print marketing. And that is fine! My point is to be sure you are analyzing contribution and then making decisions to add more revenue to your company’s bottom-line and long-term viability. I  should also mention that by adding lifetime value into the decision matrix will also pivot conversations around the health of the 0-12M file and downstream revenue changes.

If you have any questions on this data, please feel free to contact me at mhouston@cohereone.com.  We’d love to chat with you and help you with any of your marketing needs.

2 thoughts on “Why ROAS is Not a Useful Metric

  1. Helpful post

  2. Thank you, Chris.

Leave a Reply

Your email address will not be published. Required fields are marked *