Palette's TCO calculator - Methodology

Understand the true cost of Kubernetes and understand your spend and ROI if you used a DIY approach or Palette as an enterprise Kubernetes management platform

Back to Palette TCO calculator

How does this calculator work?

Every business, and every Kubernetes environment, is different. To truly calculate Total Cost of Ownership, or even better Return on Investment, you need to factor in many possible variables: from downtime to cloud spend, dev productivity to security exposure, hardware to field engineering costs.

But one of the biggest cost factors within your control is how you spend your team’s time. Are you burning thousands of hours a year on manual Kubernetes operations? How much time could you save? That’s what this calculator is for.

To show you how Palette can self-fund, we’ve looked at the three-year lifecycle of a Kubernetes environment including design, build and operate stages.

And to help you get to a representative figure faster, we’ve simplified the focus of this calculator and made some assumptions.

First, we compare fully costed staff time per year with and without Palette.

We start with an estimate of hours spent managing each cluster each year, drawn from research with ESG and projects conducted with real Palette customers. To account for differing degrees of advanced automation, we apply a maturity multiplier against this time estimate.

To turn that into a dollar cost, we multiply those hours against average fully loaded Kubernetes engineer salaries, based on StackOverflow, ZipRecruiter and PayScale. The figure we used is $200,000 annual, or $96 per working hour.


Against the management time cost we factored in license costs for Palette. This is not a precise quotation — you’ll need to talk to us to get one of those. But it’s a handy approximation based on your answers to questions about the size and nature of your infrastructure. We need a few numbers to fit into our pricing model, because Palette’s pricing is:

  • Usage-based in cloud and virtualized environmentsints, based on the number of cores running on your worker nodes in each cluster. We’ve assumed an average of four cores per node across dev, test and production clusters, and that half of your nodes are running in a cloud-managed Kubernetes environment like EKS or GKE.
  • A flat annual fee for edge and bare metal servers, with the specific price based on the size of the server. We’ve assumed an average size.