Every Monday: the 5 accounts most likely to churn this month

From your database. Count distinct active users per account per week, and add the seat count and MRR you store for each account. In PostgreSQL, date_trunc('week', …) returns the Monday of the week:

SELECT a.id                AS account_id,
       a.name              AS account_name,
       date_trunc('week', e.created_at)::date AS week_start,
       COUNT(DISTINCT e.user_id) AS active_users,
       a.seats             AS seats,
       a.mrr               AS mrr
FROM events e
JOIN accounts a ON a.id = e.account_id
WHERE e.created_at >= now() - interval '12 weeks'
GROUP BY 1, 2, 3, 5, 6
ORDER BY 3, 1;

Rename the tables and columns to match yours. If you only store today's seats and MRR (no weekly history), that is fine: use the current values. Upload every Monday and MondayKeep builds the history, because new uploads merge into earlier ones.

From Stripe. The Stripe Dashboard can export customers, subscriptions and invoices as CSV. In a spreadsheet, join them by customer ID: the subscription quantity is usually your seat count, the subscription amount is your MRR, the card's expiry month and year give card_exp (write it as YYYY-MM), and failed invoice payment attempts that week give failed_payments. Usage (active users) comes from your own product data.

From a spreadsheet. Keep one tab with one row per account per week, then use File → Download → CSV.

One row per account per week with account_id, week_start, active_users and seats. Add account_name, mrr, card_exp and failed_payments for more signals. Most teams export it from their database or billing tool in a few minutes; the template and the guide on this page show how.

A 3-person agency and a 300-seat company never look alike. Average-based alerts flag small accounts every week and miss big ones sliding slowly. MondayKeep uses each account's own median of the previous 8 weeks and its usual week-to-week swing.

Fixable risk: something you can act on this week, like a failed payment, a card about to expire, seats removed, a downgrade or usage clearly below that account's normal range. Noise: a dip that is within the account's normal swing, a tiny team where one person is on vacation, or a week when your whole customer base dipped together.

No. It is a fixed set of rules with points, listed on this page. The same data always gives the same score, and every reason shows the numbers behind it.

For a Monday email, yes: upload the newest week before Monday 13:00 UTC. New uploads merge into earlier ones, so you can upload only the latest week. If the data is more than 14 days old on Monday, you get a short reminder instead of a report.

No. You upload a CSV and nothing else. MondayKeep never asks for API keys or database passwords.

It is used only to build your report. We keep the last 12 weeks, and one button on the dashboard deletes it right away.

US$ 29 per month, billed by Stripe. Cancel anytime.