TL;DR
- Showback exposes cloud and infrastructure costs to the teams that incurred them — typically as a dashboard or report — without moving money internally.
- Chargeback goes one step further: the cost is recharged to the consuming team's budget through internal accounting entries, making it a real P&L line.
- Most organisations start with showback to build awareness, then move to chargeback once allocation accuracy is high enough to defend internal recharge.
- Both models depend on robust cost-allocation tagging; without it, neither showback nor chargeback can be done credibly.
Definitions#
Showback and chargeback are two points on the same spectrum: cost allocation. Both answer the question 'who consumed this?' What differs is whether the answer is informational or financial.
- Showback — costs are attributed to consuming teams and reported back to them, but the budget remains central. The team sees what they spent; central IT or finance still pays.
- Chargeback — costs are journalled from the central account into the consuming team's cost centre, becoming part of that team's P&L.
When Each Is Right#
| Stage | Model | Why |
|---|---|---|
| Early FinOps practice | Showback | Tagging is incomplete; visibility itself drives most of the behaviour change. |
| Mature FinOps practice | Chargeback | Allocation accuracy is defensible; finance demands real budget accountability. |
| Shared platform (e.g. internal Kubernetes) | Showback or partial chargeback | Shared overhead is hard to allocate at line-item level; pool unattributable costs. |
| Multi-tenant SaaS product | Chargeback | Customer cost is part of unit economics; needs to flow into the product P&L. |
What You Need Before You Can Do Either#
Both models require the same foundational data plumbing. The difference is only in what you do with the output.
- Consistent cost-allocation tags applied across every resource — typically cost-centre, team, env, project, application.
- An allocation rule for untagged costs — usually proportional distribution or assignment to a 'shared' bucket.
- Treatment for shared services — internal platforms, central observability, shared egress. Either flat-rate cross-charged or proportionally distributed.
- Treatment for commitments — Reserved Instance and Savings Plan discounts should be amortised fairly across consumers, not concentrated in whichever account holds them.
- An agreed cadence — typically monthly close, with daily showback dashboards for in-flight visibility.
Common Failure Modes#
Cost-allocation programmes fail more often through process than through tooling. The recurring failure patterns are well known.
- Tag drift — tagging policy exists on paper but is not enforced at resource creation. Allocation accuracy collapses.
- Surprise chargeback — finance moves from showback to chargeback without giving teams notice, time to react, or means to optimise.
- Shared-services arguments — teams dispute the allocation of central platform costs because the model was never agreed.
- Commitment hoarding — central IT keeps all RI/SP discounts and recharges teams at list price, making the central function look efficient at the teams' expense.
Chargeback without trust is worse than showback. If teams do not believe the numbers, they will spend more time arguing the bill than optimising the workload.
Implementation Pattern#
A typical journey across an 18-month FinOps maturity arc looks like the following. The point of the staged approach is to build trust in the numbers before they become real money.
- Month 0-3 — establish tagging policy, instrument FOCUS-conformant billing ingestion, build a single source of truth.
- Month 3-6 — launch showback dashboards per team, with month-over-month trend and forecast.
- Month 6-12 — agree allocation rules for shared services and commitment discounts. Publish a chargeback methodology document.
- Month 12-18 — switch to chargeback for top-level cost centres, keeping showback in parallel for finer-grained workloads.
On Yobitel#
Yobitel exposes FOCUS-conformant cost data through Yobibyte's billing API, which makes both showback and chargeback straightforward for customers running multi-team workloads on Yobitel capacity. Tags applied at workload creation propagate through to the billing record, so cost-centre and project attribution requires no separate enrichment pipeline.
References
- FinOps Foundation — Allocation capability · FinOps Foundation
- AWS Cost Allocation Tags · AWS
- Azure cost allocation · Microsoft Learn