As engineers sometimes we must make architecture decisions like whether to use SQL or NoSQL DB. These technical decisions are probably something that end users don’t care much about, at least not initially. However, eventually people become interested in analyzing business data and they are especially interested in self service data analysis. If a startup isn’t looking at its business data, that startup will probably fail.

It’s common that stakeholders want to show stats to investors, unfortunately stakeholders often have to interrupt developers to compile a one off data report but one off static reports have a significant disadvantage. Static reports fail to answer additional questions that arise.

In this article I’m going to show how I’ve equipped stakeholders with interesting and attractive real time data. I’ll also describe how to enable teams to do their own data analysis. Wouldn’t it be much easier if stakeholders had the tools to answer all their business questions themselves?

Metabase is a great tool for that purpose. Metabase offers connections to multiple data sources and allows the user to create appealing dashboards using illustrative visualization objects.

Metabase is easy to set up and maintain. You can manage Metabase on Heroku, AWS, Kubernetes, or you can use a hosted service like Metabase Cloud which will maintain the service for you so that you don’t need to care about the hosting details.

Metabase officially supports these databases:

–   BigQuery (Google Cloud Platform)

–   Druid

–   Google Analytics (version 3; if you’re interested in support for version 4, let us know by adding your +1 to this issue (#12905))

–   H2

–   MongoDB (version 3.6 or higher)

–   MySQL (version 5.7 or higher, as well as MariaDB version 10.2 or higher)

–   Oracle

–   PostgreSQL

–   Presto

–   Redshift (Amazon Web Services)

–   Snowflake

–   SparkSQL

–   SQL Server

–   SQLite

–   Vertica

If you don’t see your particular driver for your datastore, there’s still hope. There are also other drivers that support more databases but aren’t officially released. You can find the community supported drivers here.

Another interesting feature of Metabase is that it supports multiple engines (SQL, NoSQL or services like Google Analytics) that you can combine to create your own integrated dashboards.

A few other features that this tool offers:

–   Alerts over email (you could for example trigger an email when you got less than X new users on a day)

–   Alert over Slack (You could built a daily report that would show on slack)

–   Share dashboards (you could have it working as an iframe inside your application or you could have just a link that you can share)

–   Share/Download results as XLS, JSON or CSV

A powerful feature is the ability to easily share the results of a query as JSON with other services for further processing. Metabase is fast and open source and has flexible costs; the tool itself is free and you can decide how much you want to spend on hosting.

There are some things to consider before deciding if Metabase is the right tool for you. First, it has only 22 official data source drivers available while other tools like Tableau have up to 300.

I’ll also emphasize that to use Metabase effectively, you must already know what metrics you want to track. Although Metabase makes it easy to get started and explore, depending on your KPIs, you may find other tools provide the visualizations your particular business needs.

And finally, Metabase is a community driven project and it improves consistently with every release, however some aspects of Metabase may be behind more expensive competitors. Therefore you’ll want to do some collaborative requirements gathering with stakeholders to ensure Metabase can provide the visualizations you’re looking for out of the box.