top of page
  • Writer's pictureMatthew Lerner

How to Improve your ROAS

On Thursday in Barcelona, I was boarding a flight when a CFO friend texted me with an urgent question about how I’d calculate CPAs and estimate Return on Ad Spend (ROAS). (Funny what CFOs consider “urgent”) It reminded me of many conversations I’ve had where people fixate on this metric and miss a critical point. Obviously, advertising should make us money, right? Or… Q: Is it ever okay to have a negative ROAS?

A: Technically, it’s always okay to have a negative ROAS for some of your ads. If you’re spending money on advertising, your payback can come in two forms:

  1. New customers

  2. Learning

If your ads aren’t profitable in terms of customer acquisition, then at least make sure you’re learning! Q: Okay, but how do you measure learning?

A: Set a target for number of lessons learned (and documented) each week, including:

  1. What was your hypothesis?

  2. What actually happened?

  3. How did that surprise you?

  4. What will you do differently next week as a result?

This can be an email, a Slack post, a Google Doc, but don’t forget to save it somewhere for future employees so they don’t have to repeat experiments. Not sure how to “debug” a failed experiment? Check out this short video. Q: How much of your ad budget should be focused on learning vs. profitable acquisition?

A: It depends:

  • Early-stage: Pre-traction is all about learning - identifying your target audience, proposition, messages, defining the funnel.

  • Mid-Stage: Before you have product/market fit, keep your eyes on the prize! Focus on learning to lock in fit (clarify your ideal target audience, proposition, message, pricing). If your pre-PMF, your ROAS won’t be great, so don’t dump a ton of money into buying customers.

  • Post-PMF: Depending on your appetite, it’s probably 70/30 in favour of profitable acquisition. But ring-fence that learning budget to help you find new channels, messages, segments, designs, etc.


Q: Are some things better to test than others?

A: Absolutely. Don’t waste time testing random tactical changes (carousel vs. video, button colors, etc.). Focus on strategic lessons that help you find product-market fit. For example, how can you improve your targeting, refine your proposition, test pricing and messaging - insights that will make your whole business smarter, not just your marketer. If your ads can’t profitably lead to customer acquisition, they should at least bring you learning, so capture it and track it. (You get what you measure).


187 views0 comments

Recent Posts

See All

6 Scalable Alternatives to Paid Traffic

As I write this, we're heading towards Cyber-Monday, and Facebook CPAs are rising 10% - 20% per week. What can you do to rein those CPAs...


bottom of page