🌐 How does FinOps work in a heterogeneous cloud environment? 🚀

Hey everyone,

I'm diving into the world of FinOps and exploring its role in managing cloud costs, particularly when dealing with a heterogeneous cloud business - where you have resources spread across different cloud providers (AWS, Azure, Google Cloud, etc.) or services. I know that FinOps is all about visibility, accountability, and optimization, but I'm curious about how it adapts when you have multiple cloud environments in the mix.

I've heard that FinOps often works well with the Pareto Principle (80/20 rule), where 80% of costs come from 20% of resources. In this context, that seems like it would be pretty powerful in identifying the high-cost resources that need attention.

But how do you balance this approach when managing a heterogeneous cloud landscape? Does FinOps provide a solid framework for tracking, allocating, and optimizing costs across different providers? 🤔

Would love to hear your thoughts and experiences. Specifically:

• What tools or strategies do you use to get cost visibility across multiple cloud providers?

• How have you implemented the Pareto Principle (80/20) for cost optimization in a multi-cloud or hybrid setup?

• What challenges have you faced when aligning FinOps across different cloud environments?

Looking forward to your insights! 🚀

Hello, @Ajay R.

The first step would be collecting the data from all (three) providers into a repository or tool. This has to be done periodically. An Apptio tool like Cloudability (CLDY) can do that, given the appropriate permissions. CLDY handles the different products offered by the various providers pretty well.

We don't have CLDY, but we still collect the data into Apptio Cost Transparency.

Next, you need to be able to map the cloud costs to something like accounts, applications, etc. In general, you can do it with information coming from the billing data (e.g. Account Name/ID or something similar). That will give you an insight as to who is spending what money in the cloud. In some cases, you might need tagging information provided by the developers or the provisioning software, depending on your setup. 

In our case, we map accounts in GCP and AWS to Applications, and then follow the regular cost model allocation paths to Business Services and Business Units. For Azure, we leverage tagging data (actually, our FinOps team does that). The result is a breakdown of cloud costs by business owner. CLDY can do this, with some help depending on the setup.

Regarding Pareto, this is up to you, but I wouldn't ignore 20% of the costs. This might well be in the double-digit million range for large cloud spends, and might be difficult to justify to your Finance organization.

I hope this helps.

BTW, maybe this article (link) is helpful.

Thanks for sharing your insights on this topic @Guillermo Cuadrado

Hi, @Guillermo Cuadrado,

Thank you for the detailed explanation - this is really helpful! I completely agree on the importance of having clear mapping for cloud costs, especially when you're dealing with multiple providers. It's also good to know that the tool you're using, Apptio Cost Transparency, is capable of collecting the data even though you don't have CLDY.

Mapping the costs back to accounts, applications, and business units is something we're also looking to set up more effectively. In particular, I think the idea of leveraging tagging in Azure makes a lot of sense, as it seems to be a more flexible approach. Our team will likely need to refine how we gather tagging data, so I'll keep the FinOps perspective in mind.

Also, on the Pareto point - I see your point about not ignoring that 20%. Even if it's a small portion of cloud spend, it's still substantial and could definitely raise red flags in finance if it's not properly accounted for. I'll make sure we're not overlooking any potential areas.

Thanks again for sharing this insight! I'll check out the article you linked too.

Theres a few considerations here Ajay. 

When it comes to visibility across multiple CSPs ; having a single pane of glass thru tooling is key (or as I call it when I teach, a single glass of pain). This capability is normally provided by tooling such as IBM Cloudability. Now atop this, the other key consideration is tagging and having not only a policy for tagging but governance in place to ensure it occurs and its not ignored. Tagging is key to good visibility of the consumption of your cloud resources. Now remember, tooling is only part of the equation. With tooling you also need people and process to ensure that FinOps leaves a lasting affect in an organizations DNA (so its not forgotten when people move on).

We apply the Pareto Principle as a decision lens rather than a process. In FinOps, the first 80% of effort usually captures most of the value - the remaining 20% can demand significant time and coordination. We evaluate whether that last 20% justifies the effort or belongs to informed ignoring, keeping our focus on initiatives with the greatest return.

The core challenges you will face (and especially without tooling) is that each of the CSPs do things in their own nuanced ways. Learning the nuances is one of the largest challenges ; as well as the constant barrage of new things come to the CSPs to provide value to organizations - keeping up with that is a task in its own right. When it comes to aligning FinOps across our CSPs though remember FinOps is a framework like mentioned previously ; you simply adapt it to fit the (CSP) and your business to ensure that the right outcome is reached. 

I hope this helps in your FinOps journey - remember - FinOps is a journey. Its not a goal you must "reach"