Last month I went over to San Diego to speak at the FinOps X conference. The conference is ~2000 people from all kinds of backgrounds — engineers, accountants, analysts and other FinOps professionals. I’ve been aware of the foundation and the work they do for the last couple of years, so I was thrilled when our talk proposal was accepted. We shared our cloud cost journey at Monzo to-date, and dug into how we’re solving some of the unique problems a microservices architecture can present when it comes to accurately forecasting and attributing cloud spend.
I gave the talk with my colleague Selda, who leads our Cloud FinOps squad at Monzo. One of the key insights we shared was how we’ve got into a much better place with our cloud costs as a result of collaborating across different disciplines — namely by combining backend engineers with data scientists, analytics engineers and accountants. As a result, on forecasting cloud spend alone we’ve gone from a ~15% error rate in our forecasts to consistently predicting spend to over 95% accuracy. That’s improving every month, and we’re at around 97.5% at the time I’m writing this post (with our goal set to consistently be above 99%).