The economics of a streamed desktop
What a cloud workspace actually costs to run, why per-minute metering beats per-seat pricing, and how we keep a session hour honest.
By The SisiDesk Team
Most remote desktop products bill per seat per month. It is a comfortable model for the vendor and a bad one for the customer, because a seat that runs four hours a week costs the same as a seat that runs forty.
Where the money actually goes
A streamed workspace has three real cost lines: the compute instance while it is awake, the persistent volume while it exists, and the egress of encoded frames to the device. Only the first scales with how hard you work. Storage is small and constant; egress is meaningful but predictable once a codec and bitrate ceiling are fixed.
- Compute is metered per minute of an awake instance, not per calendar day.
- Idle suspension parks a workspace automatically so a forgotten tab is not a bill.
- Storage persists across suspension — that is the whole point of the product.
Why we publish the margin logic
Every session in your account shows minutes and metered cost. We hold ourselves to a gross margin target rather than a hidden markup, because a workspace you cannot forecast is a workspace finance will eventually kill.
A bill should never be a surprise. If you cannot predict it, you do not really control it.