# Weekly executive review from BigQuery, published in Notion

> Every Monday morning, turn last week's revenue, churn, and top-account moves into a written exec briefing in Notion and email the team a teaser.

- Workflow type: agent
- Services: Google BigQuery, Notion, Gmail
- Categories: Operations, Finance
- Published: 2026-07-17

## What it does

- Pulls last week's rollups from your BigQuery warehouse — revenue by product line, new customers, churn, top ten expansion accounts, and top ten usage drops — against the prior week
- Writes a plain-English weekly review with Wins, Concerns, and Watch List sections, calling out the specific accounts driving each movement
- Publishes the review as a new page inside your Weekly Business Review database in Notion, titled with the Monday it covers
- Emails a two-paragraph teaser with a link to the full write-up to your exec distribution list

## What you'll need

- A Google BigQuery project with the tables that hold your revenue, customers, and account usage
- A Notion workspace with a Weekly Business Review database shared with your Notion connection
- A Gmail account that can send email to your exec distribution list

## Prompt

Every Monday at 7am, run a synchronous SQL query in Google BigQuery that returns last week's business rollups against the prior week: revenue by product line, new customer count, churn count, the top ten accounts by expansion, and the top ten accounts by usage drop. Point the query at the datasets in the user's BigQuery project that hold customer, revenue, and account usage data, and compute both the current week and the week before so you can talk about deltas rather than absolute numbers alone.

Turn the raw rows into a written weekly executive report with three sections: Wins, Concerns, and Watch List. Write in commit-quality prose — clear, specific, and confident, not dashboard-speak. In each section, name the specific accounts driving each movement: the top expansion accounts under Wins, the biggest churn contributors and revenue softness under Concerns, and the largest usage drops under Watch List. Explain what changed and why it matters instead of restating the numbers.

Publish the report as a new page inside the Notion database called "Weekly Business Review". Title the page "Weekly Business Review — week of {Monday date}" where the date is the Monday that opens the week being reviewed. Put a short numbers block at the top (revenue, new customers, churn, expansion versus the prior week) and format each of the three narrative sections as its own heading with prose underneath.

Then send an email through Gmail to the exec distribution list. Keep the email itself short: a two-paragraph teaser that names the single most important win and the single biggest concern, followed by a link to the Notion page. Use the subject line "Weekly Business Review — week of {Monday date}".

If last week's data is missing or clearly incomplete for any of the metrics, say so explicitly in the report rather than reporting zeros. Before the first run, ask the user for the specific BigQuery table names to query, the ID of the Weekly Business Review Notion database, and the email address of the exec distribution list.

## How to customize

- Change the day and time the review runs (Sunday night, Monday morning, mid-week check-in)
- Swap in your own metrics — churn dollars, gross margin, active seats, product usage
- Adjust who receives the teaser email and how brief or detailed the summary reads

## FAQ

### Do the exec recipients need Notion access?

They need view access to the Weekly Business Review database if you want them to open the full report from the email link. The teaser email itself works for anyone with a mailbox.

### What if last week's numbers are missing or partial?

The agent is instructed to flag data gaps in the report explicitly rather than reporting zeros, so you can chase down the missing pipeline before the exec team draws the wrong conclusion.

### Can we add more sections, like product usage or customer sentiment?

Yes. Extend the SQL to pull the extra numbers and tell the agent which section (Wins, Concerns, or Watch List) they belong in. The narrative structure is a starting point, not a hard rule.

### Does it overwrite last week's page?

No. Each run creates a new page inside the Weekly Business Review database, titled with the Monday it covers, so you keep a full history.

### Can we point it at Snowflake or Postgres instead of BigQuery?

The prompt is written for BigQuery, but the pattern is the same for any warehouse we support. Swap the query step to your data source and keep the rest of the workflow the same.

Use this prompt in General Input: https://www.generalinput.com/prompts/weekly-executive-review-from-bigquery-published-in-notion