What We Know About Google Cloud's FOCUS 1.4 Billing Report

A report from shattered.io, surfaced through Google News, says Google Cloud's cost-and-usage data now aligns with version 1.4 of the FinOps Foundation's FOCUS specification, and that GPU compute costs climbed 79% in the period the report covers. The underlying dataset, methodology, and exact timeframe behind that 79% figure are not described in the material available from the report's own headline and summary, so the number should be treated as a claim to verify against a billing account, not a confirmed industry-wide figure.

What is FOCUS, and why does a version number matter?

FOCUS — the FinOps Open Cost and Usage Specification — is an open, vendor-neutral schema for formatting cloud billing exports so that cost line items look the same whether they come from Google Cloud, AWS, or Azure. A cloud provider "hitting" a specific FOCUS version means its billing export columns, cost categories, and usage units map to that version's schema, which in theory lets a finance team run the same cost report across multiple clouds without rebuilding it for each vendor. The shattered.io report frames Google Cloud's billing data as having reached FOCUS 1.4, according to the outlet's headline.

Where does the 79% GPU cost figure come from?

That number is attributed to shattered.io's own report, as reflected in the headline aggregated by Google News: "Google Cloud Hits FOCUS 1.4 as GPU Costs Surge 79%." No breakdown of which GPU instance types, regions, or billing periods produced that percentage is available in the source material reviewed for this explainer. Anyone relying on the figure for budgeting should pull the comparable numbers directly from a Google Cloud billing export rather than treating a single secondhand percentage as a planning input.

"Google Cloud Hits FOCUS 1.4 as GPU Costs Surge 79%" — headline, shattered.io, via Google News

What two numbers does the report actually put on record?

• FOCUS 1.4 — the specification version shattered.io says Google Cloud's billing data now matches.
• 79% — the reported increase in GPU compute costs, per shattered.io's headline figures, with no stated baseline period disclosed in the available summary.

Those are the only two figures tied to this story. No dollar totals, instance types, or comparison windows accompany them in the material reviewed here.

How reliable is a single-outlet report like this?

Shattered.io is the only outlet cited for both claims, and the report has not been cross-checked here against a second independent source, Google Cloud's own documentation, or a FinOps Foundation release note. That is a meaningful limitation: a 79% swing in GPU pricing or usage is the kind of figure that typically needs a stated time window and a defined cost category — on-demand, committed-use, or spot — before it means much to a procurement team. Until that detail surfaces, the figure is best read as a headline claim awaiting confirmation rather than a settled fact.

What does a FOCUS-aligned report change for a GPU buyer?

In practice, a FOCUS-conformant billing export lets a team filter cost data by charge category, pricing unit, and resource type using the same column names FinOps Foundation member companies agreed on. For a team running GPU workloads across clouds, that standardization is supposed to make it easier to compare, say, an H100 hour on one provider against a comparable instance elsewhere without manually reconciling different invoice formats. Whether Google Cloud's specific implementation delivers that in practice is a separate question from whether the 79% cost figure is accurate — the two claims in the original headline are not causally linked in the available source material.

What should a team watching GPU cloud spend do next?

• Pull a current billing export and check whether it already reports costs in FOCUS-style columns.
• Treat the 79% figure as a prompt to check a specific account's GPU line items over the same stretch of months, not as a confirmed market-wide trend.
• Watch for a follow-up from shattered.io or Google Cloud itself that defines the time window, instance types, and whether the increase reflects price changes, usage growth, or both.

Glossary

FOCUS — FinOps Open Cost and Usage Specification, an open schema for cloud billing data. FinOps — the practice and vendor coalition focused on making cloud financial operations accountable and standardized. GPU compute cost — the billed charge for renting graphics-processor capacity, typically priced per hour or per usage unit. Billing export — the raw, downloadable record of charges a cloud provider generates for an account.

This explainer is based on the available Google News summary of the shattered.io report. The full shattered.io article, with any supporting data tables, was not independently reviewed for this piece; the originating item can be found via Google News.