Showback vs. chargeback: allocating cloud costs without starting a war
Splitting the cloud bill across teams sounds simple until shared services enter the picture. How to choose an allocation model that creates accountability instead of resentment.
Every scaleup reaches the moment when one shared cloud bill stops working. Teams want autonomy, finance wants accountability, and the platform team is stuck in the middle dividing costs for shared Kubernetes clusters, central logging, and a data platform everyone uses. The model you pick — showback or chargeback — determines whether that conversation builds cost awareness or political conflict.
Showback: visibility without the invoice
Showback reports each team's share of spend without actually moving money between budgets. It is the right starting point: teams learn what they cost before anyone is held to the number. Start here, always. Teams that see their spend for the first time typically find 10–20% waste on their own, without any pressure.
Chargeback: real money, real accountability
Chargeback moves the cost into each team's actual budget. It drives harder decisions — teams genuinely weigh new features against infrastructure cost — but it only works when the allocation is trusted. Introduce it before the data is credible and teams will spend their energy disputing the bill instead of reducing it.
Shared costs are the hard part
Direct costs are easy: tagged resources belong to a team. Shared costs need an allocation key that feels fair: proportion of usage where usage is measurable, proportional to each team's direct spend where it is not. Publish the method, keep it stable, and never let one team's optimisation subsidise another team's waste.
The practical path
1. Tag everything and accept that 90% coverage beats 100% perfection.
2. Run showback for at least two quarters.
3. Agree the shared-cost allocation keys with team leads before anyone sees a chargeback number.
4. Move to chargeback gradually — shared services last.
bluebill automates the tagging audits, computes the shared-cost splits from real usage data, and gives every team a live view of its own number — so the model you choose runs on data people trust.