How to Build a Cost Allocation Model With FOCUS Data

Last updated: June 2026

A cost allocation model is the logic that assigns every line of cloud spend to an owner, a team, product, environment, or cost center, so each dollar has someone accountable for it. FOCUS, the FinOps Open Cost and Usage Specification, is an open standard that gives every provider's billing data the same columns and definitions. Building allocation on FOCUS means one model covers AWS, Azure, Google Cloud, and OCI instead of a fragile parser per cloud. This is the See step of FinOps made concrete: tagging plus normalized FOCUS data so every dollar has an owner, which is the foundation everything else in governance rests on.

This article is part of our governance, tagging and allocation cluster. For the full picture start with the complete guide to cloud cost governance, the pillar this piece links up to. Allocation only works if your tags are well designed, which is covered in what is a tagging taxonomy and how to design one.

What is FOCUS and why use it for allocation?

FOCUS is an open specification that defines a common schema for cloud billing data across providers, maintained by the FinOps Foundation. Using it for allocation means every bill, AWS, Azure, Google Cloud, OCI, and many SaaS vendors, shares the same column names and definitions for cost, usage, tags, and account identifiers, so a single allocation model runs across all of them. That removes the brittle work of normalizing each provider's native export by hand and keeps the logic portable as you add or change clouds. The standard and its column definitions are published at the FOCUS project site.

Step 1: How do I ingest billing data in FOCUS format?

Export each cloud's billing into the FOCUS schema so every provider's cost data shares the same columns before you write any allocation logic. The major providers offer FOCUS-formatted billing exports, or you map their native export to FOCUS once. Land all of it in one place, a data warehouse or lake, with a consistent FOCUS table per provider or a unified table. This single normalized dataset is what makes the rest of the model provider-agnostic, and it is the See-step data layer.

Step 2: How do I define allocation dimensions?

Decide what you allocate cost to and which FOCUS columns and tags carry those dimensions. Typical dimensions are team, product or service, environment, and cost center, each of which should map to a specific tag key or account structure. Agree these with finance and engineering up front, because the dimensions determine what reports you can produce and what owners will be held to. Keep the list short and stable; a sprawling set of dimensions is harder to keep clean than a focused one.

Step 3: How do I map tags and accounts to owners?

Build an explicit mapping from tags, account IDs, and project IDs to the owning team or product so every line item resolves to an owner. Resource tags carry most of this, but accounts, subscriptions, and projects often imply ownership too, and a lookup table fills gaps where tags are missing. Maintain the mapping as version-controlled reference data, not a one-off spreadsheet, so it stays current. The cleaner your tagging taxonomy, the smaller this mapping and the more line items resolve automatically.

Step 4: How do I allocate shared and untagged cost?

Apply a documented, agreed rule to split shared platform cost and to surface untagged spend rather than dropping it. Shared cost, such as a common Kubernetes platform, network, or observability stack, gets divided by a transparent rule, by each team's share of usage or an even split, that owners accept in advance. Untagged cost should appear as its own visible line that drives tagging cleanup, never absorbed silently into other teams. Transparency is what makes the numbers trustworthy; a hidden allocation rule destroys confidence in the whole model.

Step 5: Why reconcile the model to the invoice?

Reconcile because an allocation model is only trusted when its total equals the actual bill, including discounts and credits. Sum the allocated cost across all owners and confirm it matches the invoiced amount after commitments, negotiated rates, and credits are applied. FOCUS carries both list and effective cost, which makes this reconciliation tractable across providers. If the model and the invoice disagree, owners will dismiss the numbers, so reconciliation is the gate before any showback or chargeback goes out.

Step 6: How do I publish cost per owner?

Publish cost per team or product on a regular cadence as showback so owners see and act on their spend. Showback reports each owner's cost for visibility without moving money, which drives accountability at low friction and is the usual first step. Once the model reconciles cleanly and owners trust it, you can move to chargeback, billing cost back to budgets. Feed the published figures into the rest of governance, including budgets and guardrails described in how to set up cloud budgets that trigger automated actions.

Frequently asked questions

What is FOCUS in cloud cost?

FOCUS, the FinOps Open Cost and Usage Specification, is an open standard that defines a common schema for cloud billing data across providers. It gives AWS, Azure, Google Cloud, OCI, and SaaS bills the same column names and definitions, so a single allocation model can run across all of them without provider-specific parsing. It is maintained by the FinOps Foundation.

Why use FOCUS for cost allocation?

FOCUS makes multi-cloud allocation possible with one model instead of one per provider, because every bill shares the same schema for cost, usage, tags, and account identifiers. That removes the brittle mapping work of normalizing each cloud's native billing export by hand, and it makes the allocation logic portable as you add or change providers.

How do you allocate shared and untagged cloud cost?

Allocate shared cost with a documented, agreed rule, such as splitting by each team's share of usage or by an even split, and surface untagged cost explicitly rather than hiding it. The key is that the rule is transparent and consistent so owners trust the numbers. Untagged spend should be reported as a line that drives tagging cleanup, not absorbed silently.

What is the difference between showback and chargeback?

Showback reports each team's cloud cost for visibility without moving money, while chargeback bills that cost back to the team's budget. Showback is the usual first step because it drives accountability with low friction; chargeback follows once the allocation model is trusted and reconciles cleanly to the invoice.

The short version

Build a cost allocation model with FOCUS data by ingesting normalized billing, defining allocation dimensions, mapping tags and accounts to owners, allocating shared and untagged cost with a transparent rule, reconciling to the invoice, and publishing cost per owner as showback. FOCUS lets one model cover every cloud, and reconciliation is what makes owners trust it. When you want that allocation model built and reconciling from day one, that is exactly what our FinOps implementation service delivers.

Written by Morten Andersen and reviewed by Fredrik Filipsson, applying the See, Cut, Lock, Run method. Independent and vendor neutral.