From 609a8d65d4be8ae0ff7465e87fa3cebd350e7041 Mon Sep 17 00:00:00 2001
From: FAQ Bot <faq-bot@datatalks.club>
Date: Fri, 6 Feb 2026 19:46:09 +0000
Subject: [PATCH] NEW: How can I estimate the cost of executed queries in
 BigQuery within a spe

---
 ...326f6fa9_bigquery-cost-estimate-queries.md | 27 +++++++++++++++++++
 1 file changed, 27 insertions(+)
 create mode 100644 _questions/machine-learning-zoomcamp/general/114_08326f6fa9_bigquery-cost-estimate-queries.md

diff --git a/_questions/machine-learning-zoomcamp/general/114_08326f6fa9_bigquery-cost-estimate-queries.md b/_questions/machine-learning-zoomcamp/general/114_08326f6fa9_bigquery-cost-estimate-queries.md
new file mode 100644
index 0000000..5ff9a85
--- /dev/null
+++ b/_questions/machine-learning-zoomcamp/general/114_08326f6fa9_bigquery-cost-estimate-queries.md
@@ -0,0 +1,27 @@
+---
+id: 08326f6fa9
+question: How can I estimate the cost of executed queries in BigQuery within a specific
+  time window, and identify which users ran them and when?
+sort_order: 114
+---
+
+Overview:
+This snippet shows how to estimate the cost of executed queries in BigQuery over a time window using the INFORMATION_SCHEMA.JOBS_BY_PROJECT history metadata. The data is grouped by date, user, job_type, and statement_type to identify who ran queries and their cost.
+
+```sql
+SELECT
+  DATE(creation_time) as date,
+  job_type,
+  statement_type,
+  user_email,
+  SUM(total_bytes_billed) / 1099511627776 * 6.25 as estimated_cost_usd,  -- on-demand model costs are defined US$ 6.25 per TB
+  COUNT(*) as query_count
+FROM `project-0c3c5223-416f-4242-b0f.region-us.INFORMATION_SCHEMA.JOBS_BY_PROJECT`
+WHERE creation_time >= DATE_ADD(CURRENT_TIMESTAMP(), INTERVAL -60 DAY)  --Interval can be modified
+GROUP BY date, user_email, job_type, statement_type
+ORDER BY estimated_cost_usd DESC -- Order by the highest cost execution based on total bytes billed
+;
+```
+
+Output:
+With this query we will be able to find insights about in which dates and what users did query executions with highest costs and optimize ones that are not efficient.