We are all used to a flat rate CapEx model for on-premises VDI deployments. Even though that infrastructure is over-provisioned for peak usage and 80+% under-utilized, there is no possibility of runaway costs. And this is a model, we know and understand for multiple decades.
Not surprisingly, many customers want that same predictable model in the cloud with Desktop as a Service (DaaS). And they end up opting for 3 year flat rate pricing which can come with significant 50-60% discounts from the on-demand pricing.
After working with hundreds of DaaS customers, our recommendation is simple: “Never Use Flat Rate Pricing for DaaS. Well, Almost Never!”
Let us explain why.
(1) DaaS Usage Patterns (60-120 hrs/user/month out of 720 hours)
We see very different usage patterns across many different use cases with our customers. In the screenshots below you see data from 4 different customers. The first column is the average usage per user per day. The second column is the number of hours an average user works in a single day. The last column is the average number of days the user works in a month. The total number of hours of usage in a month range from about 60 hours for contractor use cases to 120 hours for full-time employees. Remember that in a 30-day month, there are 720 hours, so the percentage of hours in a month a desktop is in use ranges between 8-15%. These numbers are extremely important as we calculate total cost of IaaS.
(2) DaaS is an Embarrassingly Consumption-Friendly Workload
Most workloads are not consumption friendly. They need to be re-written to be cloud-native so they can take advantage of cloud consumption billing. However, desktops are an embarrassingly consumption-friendly workload. Each desktop (hence user) can be optimized for consumption independently of the other desktops. Each desktop can be started/resumed or stopped/paused completely independently based on usage requirements.
(3) Just-in-time & Just-enough optimizations
(3) 45+% Savings Over Flat Rate with Optimized Consumption Models
In the table below, we have captured the savings optimized consumption models can generate compared to 3 year flat rate pricing – 47%-75%.
Are there use cases where usage justifies flat rate pricing? Yes. But for the vast majority of use cases, optimized consumption models will generate significant savings. This changes the way IT budgets, and that is going to be the biggest tradeoff for customers – lower cost vs. greater predictability.
