The biggest mistake people make with Return on Ad Spend is treating it like a final score in a video game. It feels good to see a high number, but chasing that single metric often leads you to make decisions that feel productive but actually choke your company's growth.
This leads to a classic trap: spending most of your money on retargeting campaigns. These campaigns will almost always have the highest ROAS because they're showing ads to people who already know you, have been to your site, or are even past customers. The problem is that this isn't growth. You're not acquiring new customers; you're just paying to remind existing ones you're still there. Taylor Holiday made this point perfectly on Ecommerce Conversations when he argued against advertising to existing customers. The fix for this is to segment your metrics. You should have one ROAS target for prospecting and new customer acquisition, and a much higher one for retargeting campaigns.
A second common error is aiming for an arbitrary ROAS target, like 4x or 5x, without knowing if that's actually profitable for your specific business. People do this because they hear benchmarks in podcasts or see them on Twitter, but every business's margins are different. This is dangerous. You could be turning off campaigns that are generating profit just because they don't hit your magic number, or worse, burning cash on campaigns that meet the target but still lose money after accounting for costs. The simple fix is to calculate your breakeven ROAS. It's just 1 divided by your product's profit margin. If your margin is 33%, your breakeven ROAS is 3x. As Brendan Hughes explained on The eCom Ops Podcast, effective ad spend optimization is about tracking the right key metrics for your business, not someone else's.
Relying solely on the ROAS numbers from inside your Facebook or Google Ads dashboard is a third major pitfall. It’s convenient, but since a few years ago, platform reporting has become increasingly unreliable and incomplete. You are making decisions with bad data. When you trust it completely, you risk cutting channels that are actually driving sales and scaling channels that are taking credit for sales they didn't really generate. The solution is to have a