Find what made your BigQuery bill jump
When the BigQuery bill jumps, Claude finds the SKU, project and dataset that moved, then names the jobs behind them with INFORMATION_SCHEMA.
Runs in
Claude · Claude Code
First optimization recommendation
The next day
ROI
Pays for itself with the first recommendation you act on
Our BigQuery bill jumped last month. Which SKUs, projects and datasets caused it, and which jobs are behind them?
BigQuery +€3,100 vs the previous 3 months. On-demand Analysis +€2,300 in analytics-eu: EU has no reservation, so these jobs bill per TiB. Top job: the dbt service account, one query hash, 41 runs on events_raw.
The data platform team decides which jobs, reservation or datasets to change.
Playbook
What the agent does
From what Costory notices to what you decide. Costory suggests the next step in conversation, and you decide whether to go ahead.
Costory notices
Last month's BigQuery cost comes in above the previous three.
Costory investigates
01
Load the skill
Claude calls get_skill for bigquery, then get_context for your currency.
02
Open the dashboard
The BigQuery dashboard template shows which SKU, project and region moved.
03
Name the jobs
Claude writes INFORMATION_SCHEMA SQL that groups jobs by user, query hash and table.
04
Act
Create an alert, or map job and dataset labels to teams.
Where you see it
Claude
You decide
The data platform team decides which jobs, reservation or datasets to change.
Why it matters
The BigQuery invoice lists SKUs and projects, not queries. When the bill jumps, nobody can say which jobs, projects or datasets caused it.
Who it’s for
Data platform teams, analytics engineers
Where it runs
Claude · Claude Code
What you get
What lands in front of your team
The cause, not just a number
SKU, project and dataset that moved.
Jobs with a name
User, query hash and table behind the cost.
A fix to decide on
Reservation, query or dataset change for the owner.
Sources and assistants
What it connects
Data sources
Example prompts
Or drive it yourself
Investigate
Our BigQuery bill jumped last month. Which SKUs, projects and datasets caused it, and which jobs are behind them?
Name the jobs
Write the INFORMATION_SCHEMA query that lists last week's expensive BigQuery jobs by user, query hash and table.
Close the loop
Create an alert if daily BigQuery on-demand Analysis cost goes above €150 and post it to our #data-platform Slack channel.
FAQ
Questions about this playbook
Do we need job labels?
No. The investigation starts from invoice data: SKU, project, region and dataset. dbt job labels add cost by model and package when you want that level.
How long does this take to set up?
Connecting your billing data and the Costory MCP takes under 30 minutes. Add the MCP to Claude with {"costory": {"url": "https://app-api.costory.io/mcp"}}, and the BigQuery skill loads when you ask about BigQuery cost. The first billing data lands the next day.
Can it split slot cost by query?
Not by default. Slot cost lands on the reservation. Reallocation to jobs is set up by the Costory team on request; until then, the slot usage of each job shows who used the reservation.
Does Costory change anything in our cloud accounts?
No. Costory reads billing exports and the metrics you connect, and has no write access to your cloud. Actions it takes create Costory objects only: alerts, reports, dashboards and events. Your team stays in control of every change.
Step-by-step guide in the Costory docs
Related playbooks
Run these alongside it
Costory notices
A dbt model's cost jumps or a new expensive model appears.
BigQuery cost by dbt model, package and team
Label BigQuery jobs with dbt metadata and allocate processing and storage costs to models, packages and owners.
See the playbook
Costory notices
Spend on a team, service or account deviates from its baseline.
Investigate a cost spike and find the deploy behind it
When spend moves, the agent finds the deploy, team and service behind it, checks incidents and provider events from the same window, and posts the likely cause.
See the playbook
Costory notices
Uncovered steady Google Cloud spend persists for 90 days.
Committed Use Discount planning for Google Cloud
Import CUD metadata, see coverage per project and team, and size spend-based commitments on your steady baseline.
See the playbook
Put this playbook on call
Included in every plan, from €250/month. 30-day free trial: connect in under 30 minutes, get your first optimization recommendation the next day.
First optimization recommendation
The next day
ROI
Pays for itself with the first recommendation you act on
Setup
About 30 minutes
Data residency
EU
Access
Read-only on your cloud