FinOps X 2024 — There are no runners
Overview
For those new to the FinOps community, FinOps X is the FinOps Foundation’s (F2) global conference. It brings together practitioners, tools and services vendors and, for the first time this year, the three leading public cloud providers. This was the third FinOps X, and the Datacom team have been attending since last year, you can check out Eva Yu’s breakdown of FinOps X 2023 here. This year I had the great privilege of attending with Nick Cowan from the Cloud Financial Management team at Datacom.
As with prior years, this year was another round of exponential growth for the foundation with more than double the attendance of prior years (~1200 last year to over 2500 this year), even more sponsors, and representation from 9 out of the Fortune 10 in the room at the event. This lines up with the growth of F2 community members, also doubling from 11,000 at last year’s event to over 22,000 this year. I have been a member of F2 from the inception and it’s wonderful to see such a generous and well-run community grow and maintain its core mission even at this scale.
The other noticeable difference was the resolution of one of the major talking points last year: AWS has finally joined the F2 and was a Platinum sponsor of FinOps X along with Microsoft and Google Cloud. This resulted in a rarely seen spectacle, VPs from all 3 platforms on stage, discussing the needs of their customers and the future of FinOps. It’s one of the few areas where public cloud vendors agree, that standardising FinOps practices and data across platforms will drive huge value for both customers and platforms alike. You can view this along with the other details about the conference and more detail on the keynotes here.
Key takeaways:
It wouldn’t be a post-conference blog without a few key takeaways. As I run a consulting practice these days, the things I am looking for and that catch my attention are a little different than prior conferences, but overall, I think the following themes resonate broadly with everyone I spoke to at the event.
A change in tone for the F2
The first observation was a change in tone from the F2 and the associated community. It’s moved on from its beginnings, where passionate community members were doing all the heavy lifting to get the community working while doing very complex and demanding day jobs. There is now a very capable F2 community infrastructure that’s driving the momentum and helping gather and distribute the learnings in a programmatic way. This wasn’t an overnight change and has been the hard work of the F2 team for years, but the impact of that work came through loud and clear in the organisation and tone of the event. The team are quickly learning and making obvious improvements every year. While not without hiccups (the web agenda/conference app experience will no doubt get a bunch of attention before the next one) the event is approaching the scale and professionalism of any major industry event.
Given the attendees and the quality of content, it’s a no-brainer for anyone seriously investing in FinOps, you should have representation at this event.
That said, as with all large-scale events, you do have to work harder to network than in prior events. I did find myself caught in a lot more vendor conversations (something that is valuable but can be done in other contexts) and I had to make the conscious decision to avoid those outside of more structured conversations with vendors at booths. We also split up at lunch and at networking events to cover more ground. I got incredibly lucky many times and even got to spend the flight back from San Diego to San Francisco chatting with Josh Collier from Grammarly. Always wear your FinOps conference t-shirt on the way home, you never know who you might meet. As with all prior FinOps community events, when you do get into conversation people are friendly and generous with their time. It’s hard to overstate the value of these connections and the conversations had over those few days, especially now so many more people are focusing on FinOps as a mature practice.
FOCUS
Data standards and quality from Cloud platforms have been a headache for FinOps professionals since the early days of Public Cloud. As the market moved to a Multi and Hybrid Cloud model, the issue grew in complexity and scale. Now almost every company leveraging these platforms spends time and resources cleaning and normalizing Public Cloud platform data. FOCUS (FinOps Cost and Usage Specification) codifies the effort to normalise Cost and Usage data across vendors and contexts. It has been a major theme in the community since before the last FinOps X with the 0.5 version of the standard being announced at last year’s conference. This year saw the release of the 1.0 FOCUS standard, allowing more teams to start to leverage it in anger. We are also seeing widescale adoption by the platforms with each of the three major platforms making FOCUS announcements at or prior to the event.
My impression of FOCUS 1.0 is it’s a very strong step in the right direction. AWS have made clear that they believe there should be extra columns (noted by their FOCUS report tools having the option to add AWS-specific columns in) and there have been other discussions about teams making a lot of additional customizations. It’s early days with the standard and as JR states very clearly in the Keynote, it needs people to use and provide feedback to get better. So, if you haven’t already, take a minute to click the link above, review how the standard might help your business and most importantly take time to provide feedback. You can route this via your Cloud Provider, Tools vendor or directly to the FOCUS team. The Datacom CFM practice is also contributing to the standard so we would also welcome any discussion on the topic.
SaaS
As the benefits of FinOps are starting to materialise in many companies, the conversation is shifting towards including more and more of the IT spend base. SaaS/Licensing has been added as a key persona in the FinOps framework and SaaS was a discussion point in most FOCUS conversations I had. The effort to get normalised data from SaaS vendors is now a more organised campaign, and it’s one of the bigger bets from many FinOps/Cloud Platform management tool vendors. I think we will see more and more in this space in the coming year.
SaaS also represents one of the most interesting conversations between FinOps, ITAM (Information Technology Asset Management) and SAM (Software Asset Management) capabilities. The lessons of consumption economics and agility learned by FinOps over the last 10 years provide a range of new and powerful perspectives that ITAM and SAM teams can benefit from. Allowing teams to move away from the pure asset-driven mindset of physical hardware and license ownership to one aligned to the ever more ubiquitous demand-driven delivery models. In parallel the infrastructure of accountability, governance and audit traditionally owned by ITAM and SAM teams can accelerate FinOps maturity. I have covered the power of strong governance in another blog which you can find here.
and AI
Most FinOps practitioners are no strangers to AI in some form or another. FinOps has been leveraging machine learning for Anomaly Detection and predictive modelling for Forecasting for almost as long as the practice has existed. As you would expect, the main topic of conversation this year was the impact of LLMs (Large Language Models) on the industry. This was not so much the potential for LLMs to assist in FinOps (which was refreshing), but the cost they were incurring. E.g. FinOps for AI more than AI for FinOps.
At this point, almost every platform seems to be getting a LLM or SLM (Small Language Model) addon. Some are adding these in elegant and useful ways (usually associated with coding) but most as awkward bolt-ons with “Beta” tags and warnings about accuracy. This was represented in each of the announcements from the 3 major cloud vendors, as they are including natural language interfaces into their optimisation capabilities in each console. All three fit in the awkward bolt-on category (with all the caveats and beta tags) and for now I wouldn’t recommend spending a lot of time trying to integrate LLMs into your practice. I have no doubt a lot of this will become more useful later, but I think it will be paired with new paradigms in AI development rather than just better LLMs. That said I am no expert here so as with other major tech trends, the team will be keeping a close eye on future developments.
In the FinOps for AI discussion, it came in two parts; On stage, as per the diagram below, it was mostly about understanding the new trade-offs. AI, particularly LLMs are predicted to be major drivers of not only cost but energy use and carbon emissions in the coming years. Carbon awareness in teams is becoming an ever more important consideration for FinOps practices and there are signs of GreenOps and FinOps becoming more and more integrated. Sustainability is also another aligned role called out in the FinOps framework.
Off stage, the discussion was very focused on the secretive and unstable nature of a lot of LLM services and their pricing models. Unlike most consumption-based constructs of the last 10 years, AI services are prone to rapid price rises, low transparency in cost and performance metrics, and strict NDAs with clear restrictions on what information consumers can share publicly. I think this reflects the sense of urgency around AI and LLMs, and how incredibly fast the industry is trying to capitalise on new capabilities. Even with this in mind it’s unsettling and I think there will be more than one casualty of the race to LLM supremacy in the years to come.
Summary — There are no runners
There is no way I could cover everything of value at the conference here. Not only as it would test the patience of any dedicated reader but because the conference has become so packed with content that one person can’t get across everything that interests them. The good news is the content will make its way to YouTube over the coming months and you can see it all, including last year’s sessions and much more on the F2 YouTube channel here. With that in mind, I would like to cover one last and very important point made by JR in the day 1 Keynote:
There are no runners. FinOps, much like Security has become everyone’s job. While there are central teams that push the capabilities forward, it doesn’t work without the cooperation of everyone in the broader team. Security doesn’t work if you don’t lock the door on the way out, and cost management won’t work if you don’t switch off the lights as you exit. As such it encompasses the constant evolution of markets, the IT industry, company culture and a hundred other variables that change what value means for a business. In this sea of constant change, you can get parts of your practice to the Run state, but never everything and certainly not for very long. The FinOps journey is never complete and there are endless opportunities to drive more value for your organization. The F2 community is a great place to learn how and for that I am very grateful.
