FinOps X 2024 - Looking back at a great event

Last year (2023), I felt like I’d finally found my home, I was with my people. FinOps X is the only place on the planet I can go and be surrounded by people who understand what I do. This year felt very similar, but it did have two noticeable differences.

1. It felt a whole lot bigger.
2. It felt like there were a lot of people at the start of their journey.

Neither of these things are bad, as an open learner it was a great experience for me to share my own journey and knowledge of FinOps with people who hadn’t heard it before. It was also great to meet people who listen to the Podcast I do with Frank (What’s New in Cloud FinOps) and meet a few people already watching Synyega TV.

Starting with the keynotes, I thought they were excellent, both days. I feel JR opened superbly around the future of FinOps, exciting the crowd for not just the conference but what’s the come in general and then probably one of my highlights of the conference - Elly Rauch of Disney hit the stage. Living up to her colleague Ally Whitman who was person of the conference for me in 2023, Elly lit up the stage with a hugely engaging and thought provoking talk on forecasting.

Uber came next, telling us some quite staggering numbers regarding errors they’d found by the cloud vendors in their billing data, meaning they were owed millions of dollars which most customers who weren’t as diligent would never have even noticed. They said their number was over 6.5% which far larger than I ever experienced, but I do ponder whether some of that comes from a particular vendor who I worked with less in my time. I’ve never seen it much above 0.5%. Very interesting either way!

Next we had the launch of FOCUS 1.0 (and the new exam - which I passed this morning), which was great news for those of us very hopeful that standardisation of data may lead to better spent time in the world of FinOps.

I took a slightly different approach this year, on day 1 I did a lot of chalk talks. For those who have never been, Chalk Talks are interactive sessions where a room of people get split into groups to discuss a topic and then present back. I only did one last year and did far more normal sessions (with a speaker at the front and slides). While these chalk talks were very well run, and good. For me they didn’t actually add as much value as I’d hoped and because they were twice as long this year, each one meant I missed two talks. So I felt on day two, I had to change tact and went back to more conventional talks.

The two chalk talks I went to, were about non-obvious optimisations and Demonstrating the Value of cloud.

In the optimisations session, let by some real FinOps legends, we had a really friendly mixed group. We discussed some quite complex topics and some quite basic ones. I brought up complex commitment strategies and explained them to the group. Someone also brought up tactics around centralising Cloudtrail to reduce costs, which I thought was great. We also discussed containerisation versus FaaS. Some really great topics, but I always felt a little aware that we had such a mixed experience group, that we needed to try to keep our concepts accessible, otherwise people wouldn’t be getting value. This was a hard balancing act.

On the demonstrating business value chalk talk, we ended up speaking largely about unit economics, the two people running the room were excellent, but I feel we maybe as a set of groups, didn’t really answer the questions they set (in hindsight). Anyway, it was a rewarding group to be in as I met some brilliant people and shared some war-stories.

After this I went to a very good talk by Mastercard’s Vic Saluja and Mike Jaco, where they said (without saying it specifically) that Green Pixie is miles ahead of other GreenOps tools out there. Something I strongly agree with, but they have been working with it at scale and proven the point.

End of day one and we had a very good time on the Midway aircraft carrier again, before getting up early to hit day two!

Day 2 keynotes again very good, Mike Fuller on tooling I found very interesting as it’s something I’ve worked on a lot, and then the product announcements from the cloud vendors was great - special shoutout to Sarah from Google who’s presentation outshone the rest.

So into day 2 proper - what a great day of talks it was.

First up Optimizing for Billions from Coinbase - extremely detailed talking about some great concepts of where they continued to find greater and greater efficiencies in terms of performance and cost. Using containerisation as well as looking at how to boot, how to pre-load instances for scaling and then continually trying to improve their performance and cost. Never resting on their successes, always looking for more. Great talk.

Unfortunately for them… the next talk was even better and is really the one I can’t help but remember. Ziva and Dvir from Wix walked us through “Taking FinOps to the next levels: Beyond Optimization”. In this talk, Dvir and Ziza eloquently brought us through some of the unique approached they have taken in FinOps. Not so much in the fact they have shifted left - which while not common yet, isn’t unheard of - but in the way they’ve done it. They had a brilliant process for engaging in the design phase, which is worth watching Youtube for alone. More than that though, they had Financial Incident tickets automation. Which meets the engineers where they are in sending them tickets via the flows they are used to around specific financial anomalies that need investigating. Again, a must watch from the video when it comes out. However the thought that really stuck with me, was 100% utilised wastage. Resources that may look 100% used by normal metrics, but when you change the perspective you suddenly realised you’re wasting. Caching services being a great example, where by nature 100% of memory will always be in use, but the true metric is hit rate. Genius! Thank you both for sharing your brilliant minds with the community.

In the afternoon I jumped into a FOCUS session, talking about the standard and what’s next. I really want people to go watch this and to go do the Introduction to Focus course which is free. So I won’t spoil it here.

My talk of the afternoon was by Andrew Feig and Craig Shine from JPMorgan Chase) - “Achieving Cost Excellence: Shifting Left, Compute Optimization and tackling Anomaly Detection”. What was really excellent here was the self-built metrics they’d created for both sides of this talk. It ended up going more down one route for the audience, but I think both were excellent points and I was staggered by how much of the work had been done by one very capable man. Definitely worth a watch again!

My mind blown, I just joined a final chalk talk on focus which was about use cases - definitely go look into the use case library to learn more about this!

So all in all, an incredible conference. Thanks so much to everyone who spoke and everyone who said hello. I’m already looking forward to next year.