Collecting and reporting data from different sources is always a challenge and requires a little bit of creativity. And since we love challenges and are very creative, we were very excited to take part in the transformation of the tracking architecture for the Czech Radio, the public radio broadcaster of the Czech Republic.
Czech Radio is operating since 1923 and it is the oldest radio broadcaster in continental Europe and the second oldest in Europe after the BBC from the United Kingdom. It operates nation-wide with 11 national and 14 local stations. On top of that, the Czech Radio broadcasts on several different platforms: radio frequencies, websites, mobile apps, podcast services, and smart TVs.
Time for App+Web
We have found ourselves in an environment where various content is being consumed through various channels, some of which have limited or no possibilities for direct audience data collection.
When Google dropped the beta version of Google Analytics App+Web we knew that this session and event-based tool would be the perfect solution for data collection (and utilization) of this kind.
What, Who for, and Why?
To design the best data collection possible, we wanted to know who are the people looking at all of this data. That is why we have interviewed CRo staff members from all of the key departments, such as content editors, board members, members of the New Media department, HBBTV team members, and employees in charge of podcast broadcasting. This has not only helped us to learn their needs but also to find out which specific data insights they are lacking in the current setup.
Time for the Google Tag Manager
Following up, we have redesigned and unified the data layer and the Google Tag Manager setup. One of the crucial steps was to create comprehensive tracking documentation consisting of unified terminology and compatible event structure, especially for the metrics from 3rd party platforms, such as YouTube, Spotify, Apple Podcast, Smart TVs, etc.
For the solution to be as autonomous and independent as possible, we have come up with a steady reporting strategy – reports created in Data Studio.
Right now we are able to stream data straight to the BigQuery through the Firebase Blaze program for regular data ingestion cost, together with all 3rd party data using custom collection APIs. The result is easy-to-use reports with all necessary information about the platforms, stations, and program/episode IDs.
And what if we can’t utilize APIs? Yes, that happened with some portion of the data, mostly from podcast services. For these particular cases, we have built the so-called App Engine Instance which uses a script to extract the data from external services and then sends it to Bigquery.
All in One Place
Thanks to the new App+web solution and all of the connectors, data from various sources are now in one place. Further data integration will be much easier because the data structure is already created and another integration is just a matter of transferring the data from place A to place B.
The design of the solution not only enables further data integration if needed but is also applicable and scalable for any radio station or broadcasting service.