The release of AWS Savings Plans is going to change how companies optimize their AWS costs. And make no mistake — those changes are going to impact your FinOps practice. The question is how much and how quickly.
(A full video of the presentation is available for FinOps Foundation Members)
The short answer? We don’t know. Yet. As with any release or major change, there’s going to be a period of six months to a year before best practices are fully developed. That being said, you can be sure that all of the experts in the FinOps Foundation jumped right into Savings Plans, even if only to figure out how they work.
Last week, over a hundred of them got together to chat about Savings Plans, strategies for adoption, and potential pitfalls. The meeting included discussion and a quick, five-question survey. Check out a summary of the responses below, then dig into detail with the meeting recording in the Member area.
#1: Do you see RIs being completely replaced by Savings Plans?
Many people have positioned Savings Plans as the replacement for RIs, and, at first glance, this seems like a solid conclusion. But there are a few key differences that keep it from being such a cut and dry answer.
As of now, there are certain key limitations cited by those in the “No” camp, including the fact that RIs apply to non-compute services (like RDS) while Savings Plans are limited to EC2 and Fargate.
#2: For those who are going to keep buying RIs, why?
Digging a little deeper into the reasons to keep using RIs yielded a variety of different reasons, with central themes around maturity adoption, systems, and Savings Plans limitations.
Additional concerns were raised, such as increased specificity and control with RIs or the ability to sell/capitalize them. This might change in time, since Savings Plans are in their infancy. Being more mature, RIs have incredibly robust processes, systems, and best practices surrounding them. If nothing else, these will lead to a delay in Savings Plans adoption.
#3: How soon do you plan to start buying SPs?
While some members weren’t confident about replacing RIs, the majority were still embracing adoption of Savings Plans. Many had already purchased Savings Plans, although mostly in an experimental capacity.
The key takeaway from this, however, was the experimental approach. F2 members are largely in the middle of evaluating Savings Plans and figuring out how they’ll fit into existing systems. (There are a lot of in-depth discussions going on in the member Slack channel.) As a result, a lot of the purchases in the blue and yellow bars are small so FinOps teams can get an idea of how Savings Plans apply and appear in the CUR.
#4: Are you going to buy any more RIs?
This question is really the telling one. Savings Plans are being positioned as RI replacements, but it’s going to take FinOps expert buy-in to make that happen. And on that front, people are mostly being strategic about their approach.
The majority are still going to buy some RIs for the reasons above (non-computer coverage, etc.), but the jury’s still out on whether SPs will be a full replacement. We’re sure these opinions are going to develop a lot over the coming months — especially after any announcements or releases at re:Invent.
#5: What do you expect to be the hardest challenge to solve with Savings Plans?
Any financial savings tool is only as effective as it allows you to track the savings and improve on them. With RIs, these techniques are fully mature, allowing companies to reach coverage levels of 80-90%. Being a new product, the same can’t be said of Savings Plans.
Looking through the responses, it’s obvious there’s a need for further defined methodologies and best practices to track, allocate, and manage Savings Plans. It’s also obvious that the FinOps experts are planning to put a lot of work into figuring those out.
Conclusion: A good strategy is essential for Savings Plans adoption
Throughout all of the responses and discussions around them, one thing was made clear: the only way to get the most out of Savings Plans is to build a solid strategy around adopting them. Before you purchase large quantities, you should be able to answer key questions like:
- How will Savings Plans apply to my usage?
- How will Savings Plans work with my existing RI portfolio?
- How much are my costs and usage that are eligible for Savings Plans?
- How much would Savings Plans to cover them cost? (Remember: Savings Plans are based on your spend after the discount is applied!)
- How can I allocate spend under Savings Plans vs. non-Savings Plans spend?
- How can I make sure that Savings Plans are used when and where we want them to be used? (e.g. ensuring they’re used for production and not sucked up by R&D efforts)
There are obviously more questions to be answered before you can invest with 100% confidence, but that doesn’t mean you should necessarily wait to buy Savings Plans. The field of FinOps and cloud is always evolving, so we should all approach it with an experimental mindset.
There are many ways to do this. One member recommended splitting your Savings Plans investment out into equal percentages and buying 1-year plans slowly over the course of the year. By the time you reach the other end, the plans will start expiring, so you’ll have the opportunity to reevaluate and adjust accordingly every month.
Obviously that approach won’t work for all companies, but it’s the kind of thinking that everyone should doing towards Savings Plans. It’s an exciting time that could completely change how AWS finances are managed. Like everything with the cloud, let’s embrace that change strategically and use to reach even higher heights.
Join the FinOps & Savings Plans discussion
Make sure you stay up to date on the latest best practices, insights, and strategies by joining the FinOps Foundation. With member-only resources, an exclusive Slack channel, and local FinOps meetups, it’s the best way to make sure you’re at the top of your FinOps game.