mffap / zitadel-python-mau-per-org

Calculate monthly active users per organization in zitadel using Python.

Monthly active users per organization in ZITADEL

Why would you need this?

Assumption is that we define Monthly Active Users as users who refreshed their token once per month.

There is currently no single call that can extract all user.token.added events with the UserIds AND the organization-id. There is also no call that returns all UserIds across all organizations available.

Given these limitations, you have to do the following steps:

1. Pull all user.token.added (new token created) to get a timestamp and a UserId
2. Get all organizations for an instance
3. Loop through all organizations and get all the UserIds
4. Match the UserId in the token to the Users's organization ID
5. Uniquely count token created events per organization and slice by month

Prerequisites

• Python 3.x
• Service User PAT with Manager role IAM_OWNER_VIEWER

Install

Create a file called .env with the following contents.

PAT="D1...qQ"
BASE_URL="abc.zitadel.cloud"
START_YEAR=2023

Create the output folder.

mkdir ./output

Create and activate virtual environment.

python3 -m venv venv
source venv/bin/activate

Install dependencies.

pip install -r requirements.txt

Generate the ZITADEL client including dependencies (takes a couple of seconds). This will create several package folders.

buf generate https://github.com/zitadel/zitadel#format=git --include-imports

Run

Activate the environment.

python3 -m venv venv

Download the data from ZITADEL. Output will be saved as pickle (pkl) file in /output.

python3 main.py

Transform the data and save the output as csv file in /output. You can also use the Notebook (.ipynb) instead.

python3 mau_per_org.py

Debug

If something breaks with the connection to ZITADEL use the following flags to get more information about the grpc request.

GRPC_VERBOSITY=DEBUG GRPC_TRACE=http python3 main.py