Will a scheduled BigQuery report run up my bill?

That is the real risk with recurring BigQuery work, since you pay per byte scanned and an hourly schedule multiplies one careless query by 720 runs a month. SaturnSQL enforces a per-query scanned-bytes cap on the connection, so a query that would scan more than your limit is rejected before it runs.

Can I post a daily GA4 summary to Slack?

Yes, and it is one of the most common uses. Query the GA4 BigQuery export for yesterday's sessions, users, and conversions, then schedule the result into a channel each morning. Filter on the partition column so the query only scans the day you need.

How large a result can I post to Slack?

Slack caps a message section at roughly 3,000 characters. SaturnSQL fails the delivery rather than truncating it, so Slack suits aggregates and short lists. For full extracts, schedule the same query to Google Sheets instead.

BigQuery → Slack

BigQuery to Slack, without the surprise bill

Q: How do I send a BigQuery query to Slack automatically?

Connect BigQuery to SaturnSQL with a service account, write and save your query, then add a schedule with Slack as the destination and pick a channel. The query runs on your cadence and the result posts itself. No Cloud Function, no Scheduled Query plus Pub/Sub plumbing.

Post a scheduled BigQuery result straight into a channel, with a per-query scanned-bytes cap so a recurring report cannot quietly become expensive. From €19/month.

Start free for 14 days See pricing

Service account auth · Per-query cost cap · No credit card required

ga4_daily.sqlbigquery

SELECT
  count(DISTINCT session_id) AS sessions,
  count(DISTINCT user_pseudo_id) AS users
FROM events_*
WHERE _TABLE_SUFFIX = '20260726';

Every morning

#growth08:00

SaturnSQLAPP

GA4 yesterday

Sessions: 18,402

Users: 11,238

Signups: 96

Open in SaturnSQL

Why the cap matters

BigQuery bills per byte scanned, and a schedule multiplies one careless query by up to 720 runs a month. The connection-level cap rejects a query that would scan more than your limit before it runs, which is the difference between a report and an incident.

Set it up in five minutes

No Cloud Function. No Scheduled Query wired to Pub/Sub. No webhook to host.

Connect BigQuery

Add a service account key and set a scanned-bytes cap, so a scheduled query has a known worst-case cost.

Filter the partition

Write the query against one day rather than the whole table, and check the bytes scanned before you save.

Schedule to a channel

Pick the Slack channel and a cadence, and optionally template the message into a sentence.

Patterns

Three flows worth wiring up

#growth

The GA4 morning digest

Yesterday’s sessions, users, and conversions from the GA4 export, posted at 08:00. Answers the question people would otherwise open four Analytics tabs for.

#marketing

The spend check

Ad spend and cost per acquisition by channel, daily. Cheap to run when the query filters on the date partition, and it catches a runaway campaign the same day.

#data-alerts

The pipeline heartbeat

Row counts from the tables your loads write to, with the query returning nothing when everything landed. Silence means the pipeline is healthy.

Working with the GA4 export specifically? Our guide to querying GA4 data in BigQuery covers the table layout and the partition filters that keep these queries cheap.

Message templates

A sentence beats a table

Alias a column in SQL, reference it in braces, and the post reads like something a person wrote rather than a grid nobody parses on a phone.

Template

GA4 {date}: *{sessions}* sessions, {users} users, {signups} signups.

Posts as

GA4 2026-07-27: 18,402 sessions, 11,238 users, 96 signups.

Placeholders resolve against the first result row. Built-ins {results}, {row_count}, {query_name} and {date} are always available. An unknown placeholder fails the delivery rather than posting a blank, because a silently wrong metric is worse than a missing one.

Where Slack is the wrong destination

Slack caps a message section at roughly 3,000 characters, and SaturnSQL fails the delivery rather than truncating it. Half a table posted as if it were the whole answer is worse than a delivery you get told about.

With BigQuery this rarely bites, because the queries that belong in Slack are aggregates, and aggregates are also the cheap ones to scan. If the honest answer is a large extract, schedule it to Google Sheets and post a short summary alongside. A schedule has one destination, so that is two schedules on the same saved query.

FAQ

Frequently asked questions

How do I send a BigQuery query to Slack automatically?

Connect BigQuery with a service account, write and save your query, then add a schedule with Slack as the destination and pick a channel. No Cloud Function and no Scheduled Query wired through Pub/Sub.

Will a scheduled report run up my BigQuery bill?

That is the real risk with recurring BigQuery work: you pay per byte scanned, and hourly scheduling turns one careless query into 720 runs a month. The connection carries a per-query scanned-bytes cap, and a query that would exceed it is rejected before it runs.

Can I post a daily GA4 summary?

Yes, and it is the most common use of this. Query the GA4 export for yesterday, filter on the partition so you scan one day rather than the whole history, and schedule it into a channel each morning.

How large a result can I post?

Around 3,000 characters, which is a Slack limit rather than ours. SaturnSQL fails rather than truncating, so you are told instead of quietly misinformed. Aggregate for Slack, and send full extracts to a sheet.

Which other databases can post to Slack?

All of them. See Redshift to Slack, or the overview for Postgres, MySQL, SQL Server, and ClickHouse.

Your BigQuery numbers, in the channel

From €19/mo, with a cost cap on every scheduled run.

Start free for 14 days

No credit card required.