Roasting ROAS.

ROAS, or Return On Ad Spend, may as well be the go-to metric for performance marketing. Other than ROI (Return On Investment), the metric only considers the money that’s spent buying ad space. The higher the return on media investment, the better the campaign performance. Seems logic, right? Yes, in theory. In practice? I have my doubts.

The whole idea of performance marketing is doubtful to me. Of course, I look at numbers and use them as a way to optimize campaigns. But I just can’t understand why this is different from marketing in general. Isn’t that the very nature of the trade? For some reason it seems to be a good idea to only focus on the numbers and forget about the rest, which makes ROAS the measure of choice.?

Here are 5 reasons why ROAS can as easily work against your campaigns.

1. It neglects the bigger picture. Why wouldn’t you consider your full investment (and return)? It’s like buying a car by only looking at the miles per gallon. Then you have some performance marketer that discovers the car could run on pure alcohol. While alcohol may be cheaper than gasoline, they don’t measure the damage to the engine.?

2. Following the analogy of cheap fuel, if ROAS is your metric then you’re incentivized to buy dead cheap media. Even though it's the shittiest media out there, big brands advertise on all these annoying clickbait websites. The result is a very bad experience for the end users.

3. If your only goal is the return on media spend, then it can go south easily. Obviously you should also consider your management resources. If you have 300% on 50k media, but you use 100k to manage it, you might not get the result you’re after. While ROAS can be a good measure to compare results so that “all other things kept equal,” usually it’s not used this way.

4. How do you define “return” anyway? Revenue? Margin? Clicks? As an advertiser you want margin. But your agency might not agree... They could argue that since they don’t control the website the best they can do is deliver qualified leads. It’s the website's job to do the sale. Understandable, but then the shit hits the fan. If you can’t define a qualified lead (and you can’t with all the changes in tracking due to GDPR and e-privacy) any click will do. You want money, you get clicks. These clicks might trick people and even piss them off. This is what happens all the time: you have your eye on a discounted product, you click, but the product is not available. From a marketing perspective this is totally devastating. From a ROAS perspective it’s a best practice.

(see post)

5. There are zero long term objectives. Any piece of communication should consider the brand and help it build. However in a ROAS context, the brand is completely neglected. ROAS would favour an unbranded ad that clicks better. Any ad that’s not clicked is a waste. Since most ads aren’t clicked on, the majority of the campaign is a loss. This is not how advertising works. There is also no consideration on returning customers or customer lifetime value. If the CLV were high, it would justify even a ridiculous low ROAS.?

"Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes." - Goodhart’s law

Sure, ROAS can be used in good ways. There's definitely room for it on a balanced scorecard, but just understand it’s not the holy grail and can be easily misused. Abusing ROAS is a very good example of what we like to call: Death by KPI. I recently learned there’s something called Goodhart’s law that states: “Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes." In other words: when a measure becomes a target, it ceases to be a good measure. That might be the biggest problem with ROAS - it quickly becomes a target and not a measure of interest.

This article was written as an editorial for our new newsletter. Do you want to receive our newsletter too??Feel free to sign-up here.