Google Cloud's serverless data warehouse, billed per TiB scanned
BigQuery is the data warehouse to pick if your product data already lands in Google Cloud, Firebase or GA4: the exports are one click and the first terabyte of queries each month is free. The bill is driven by bytes scanned, not by seats, so a disciplined team with partitioned tables pays very little. The risk is the opposite: one careless SELECT * over a year of events costs real money.
Are we wrong about your product? Tell us.
BigQuery is Google Cloud's serverless data warehouse. You load or stream data into tables, query it with standard SQL, and Google runs the query across as many machines as it needs. There are no clusters to size and nothing to patch. It is the default destination for Firebase and Google Analytics 4 event exports, which is why many Indian consumer apps end up on it without deciding to.
You pay two meters: storage (per GiB per month, cheaper for tables not edited in 90 days) and compute. Compute is either on-demand (per TiB of data a query reads) or capacity (slots bought by the hour in Standard, Enterprise and Enterprise Plus editions). Mumbai (asia-south1) and Delhi (asia-south2) are both regions, so data can stay in India.
Quick facts: Google Cloud · Serverless, standard SQL · Mumbai and Delhi regions · Free: 1 TiB of queries and 10 GiB storage a month · Native GA4 and Firebase export.
A query over a few terabytes finishes in seconds without any capacity planning; you pay for the bytes it read.
Raw event-level data from Google Analytics 4 and Firebase lands in BigQuery daily or streaming, with no ETL tool.
Partition tables by date and cluster by user or event name so a query reads only the slices it needs. This is the single biggest cost lever.
Train regression, classification and forecasting models in SQL, and use Gemini models on text columns from inside a query.
asia-south1 (Mumbai) and asia-south2 (Delhi) keep data in India for RBI and DPDP conversations.
Access per dataset, per table, per row policy and per column tag, with audit logs in Cloud Logging.
Pay for what queries read and what tables store. The free allowances renew every month and are enough for a small product's event data.
| Plan | Price | What you get |
|---|---|---|
| Queries, on-demand | First 1 TiB a month free, then $6.25 per TiB | Billed on bytes each query reads; partitioned tables read less |
| Active logical storage | First 10 GiB a month free, then $0.000031507 per GiB-hour | About $0.023 per GiB a month in US regions; tables edited in the last 90 days |
| Long-term logical storage | $0.000021918 per GiB-hour | About $0.016 per GiB a month; tables untouched for 90 days drop to this rate automatically |
| Capacity, Enterprise edition | $0.06 per slot-hour (pay as you go, US) | $0.054 with a 1-year commitment, $0.048 with 3 years; Standard and Enterprise Plus have their own rates per region |
Source: cloud.google.com/bigquery/pricing, checked 22 Oct 2026. Prices in US dollars, US regions; Mumbai and Delhi rates are listed on the same page and run slightly higher. GST applies on Google Cloud India invoices.
In the order a small team would do them.
Pick the Mumbai region when you create the dataset; it cannot be moved later without copying every table.
In IAM & Admin → Quotas, cap 'Query usage per day' for the project. Start at 1 TiB; raise it when you know your pattern.
Link the GA4 property to the project; daily export is free of GA4 charges and lands in a date-partitioned events table.
Partition by event date and cluster by user_id or event_name; require a partition filter on large tables.
Point your BI tool at a small, aggregated table rather than the raw events table so dashboards do not rescan history.
If BigQuery is not the right fit, these are the pages to read next.
Cloud-neutral warehouse billed per compute-second with separate storage; runs on AWS, Azure and GCP, including Mumbai.
Choose when: You are on AWS, or want the same warehouse across cloudsOpen-source columnar database for sub-second analytics on event streams; self-host for free or use ClickHouse Cloud.
Choose when: You need real-time dashboards and want to control the cost model