Kubernetes Cost Management: The Build vs. Buy Dilemma (And How To Solve It)

When starting their cloud journey, most teams turn to native cost management from cloud providers. But once your cloud footprint scales into multiple services, accounts, and providers, these tools won’t cut it anymore.

So, you’re left with a choice:

Is it better to buy a third-party solution or build one in-house, possibly based on open source? 🤔

Hasty decision-making here might lead you down the path of poorly scoped development, costing you way more than a teammate with a habit of overprovisioning by 300%. 💸

You need to take care of these three things before running off to build your custom tooling.

3 things to mull over when deciding whether to build or buy

1. How much will it all cost?

Investing in a homegrown cloud cost management tool needs to make sense from a financial perspective.

Do you know how many engineers will build the new tool or how many hours it will take? You can’t calculate the cost of your DIY solution without knowing that.

Tech giants might have some 5-10 engineers to spare for building and maintaining their cost tooling. Do you have that kind of flexibility?

This is how you discover the opportunity cost of building your own cloud cost management tool:

• Start by assessing the number of engineers you’d need,
• Multiply this number by their average annual salary, 
• Then add some more for overhead costs (someone will have to oversee the project and run QA tests).

2. Are you really the best team to build this tool? 

It’s time to ask some questions:

• Why is your team better suited to solving the cloud cost problem than a team of full-time engineers who have dedicated the past few years to it?
• And even if you have top talent on board, are you sure that this is what they should be doing in the first place? 

The mission of your business is to solve problems for your customers. 

Shifting your focus to cloud cost optimization might get you a win on the cloud bill front. But what wins in your main business are you going to sacrifice?

3. How strategically important is cutting cloud costs to your business?

Saving millions of dollars on the cloud – as the mobile marketing company Branch did – is a great way to boost your business. 

But will optimizing the cloud deliver value to your customers? 

Will it help you compete in the market?

When engineers focus on cloud costs, they’re no longer involved in building new features.  

So, ask yourself these questions:

• Is cloud cost optimization a top priority? 
• How does it compare to other items on your roadmap?

Research shows that high-performing organizations make sure their internal teams focus on core aspects of their business and buy all the software that’s not strategic as SaaS. There’s an important lesson to be learned here.

Is turning to open source the answer?

The tech industry wouldn’t be where it is today if it weren’t for open-source. But is it a good candidate for cloud cost optimization as well?

Even if you take the route of implementing a bunch of ready-made solutions, you’ll still have to put in the work to keep this patchwork in good shape. You’ll be fiddling with the software to no end that way.

As stories like the Log4j fiasco show, stability can also be a problem. The downside here is that you don’t have a support team jumping on the issue ASAP or any SLAs for your business to rely on.

Is there a solution?

Many teams decide to build a cost management tool in-house because they feel that legacy cost optimization solutions aren’t up to the task when dealing with cloud-native technologies like Kubernetes. 

Now, we're biased here because we’ve built a cost management platform based on automation.

But CAST AI was actually born the same way. Out of the need to manage cloud costs better when our founders were running an entirely different company.

We now have tens of engineers working to solve this problem, and if you ask any of them, they’ll tell you that automating cost optimization without sacrificing reliability is no easy feat.

Ultimately, the choice is yours. Building a cost tool in-house gives you all the control but also the responsibility of seeing the entire project through and maintaining it.

I hope the considerations above help you make the right decision.

Cheers,

Allen

Found this email useful? Forward it to your friends and colleagues who need more Kubernetes best practices in their lives!