Anghami is an online music streaming service based in Lebanon. It is a sizable player in the Arabic music market with over 30M app downloads and 6M people using the app monthly. Besides running a free streaming site, it also has premium subscriptions, and has to deal with reporting to music content rights holders.
Like many cutting edge startups worldwide, Anghami prefers to use Open Source software whenever possible.
Prior to installing Metabase, they were using a custom coded analytics tool that sucked up engineering time to develop and maintain over the years.
One day Helmi (Growth & Analytics Lead @ Anghami) was searching on Google for “visualization tools for databases” and found Metabase. A Metabase instance was up and running via a Docker container in 5 minutes and the next day others in his company were using it.
Now, aside from helping write the occasional SQL view or complicated query, Anghami’s analytics system requires no maintenance time.
When Helmi first set up Metabase, Anghami was already using a separate data warehouse. This contained information about music streaming plays, user profiles and information about subscriptions. They paired this with Amplitude for interaction data and behavioral analytics.
They had recently moved to an AWS Redshift data warehouse. While this data is in a reasonable schema for analytical use, Helmi does occasionally create new aggregations that bring together data in a format that is easily and quickly consumed by non-technical coworkers via the Metabase GUI.
As might be expected in a business with a fairly large number of moving parts (content, subscriptions, rights management, etc.) there are a wide variety of questions that are asked within Metabase at Anghami. There are a set of nightly stats emails as well as a large number of dashboards.
About a third of the company uses Metabase weekly or more often — with users across Marketing, Advertising, Content, Financial, BizDev, Engineering as well as the Executive team. In a sign of a well structured and tuned data model, most of the company is able to use the GUI query builder to ask questions themselves.
By spending a bit of upfront time creating data models and marking up metadata, Helmi and the other engineers were able to almost entirely eliminate ongoing analytics engineering work.
Combining a bit of upfront work, a culture of data transparency and enabling any employee of the company to access Metabase via Google SSO provides a simple and easily maintainable self-serve, company-wide analytics system that requires minimal ongoing effort.