The Museum tracks data from all digital experiences and this capability must be factored into the experience development. It is critical to understand in detail how visitors are interacting with digital experiences in order to assess their effectiveness, inform ongoing improvements and guide the design and development of new digital experiences.
The basic analytics requirements are:
- Google Tag Manager set up on AM account (AM will provide access)
- Google Analytics property set up
- Standard events set up for sessions, CTA, page views
- Video analytics on playback, time spent, pause, scrubbing, volume adjustment.
These requirements may vary between projects. The digital lead on your project will discuss analytics requirements with you in more detail during the contract and brief phases. Custom data layers may need to be set up and applied to the front-end code depending on what insights are being tracked.
Insights, vanity metrics, what is a session?
When setting up analytics it's important to always first ask what metrics you want to track and why. Ensure these relate to clear success measures and learning outcomes of the interactive and avoiding ‘tracking for the sake of tracking’. Avoid reports that show large ‘dumb’ numbers (e.g. “300,000 page views”) with no context as to how that relates to a clear insight for the experience, a learning, a component to optimise or tweak, etc.
When reviewing analytics it is also important to look at the physical environment of where the interactive is located in relation to the visitor journey, location in-gallery and proximity to objects.
Finally tracking user ‘sessions’ is often difficult as visitors can simply walk away mid experience and another visitor picks up where they left off. There are creative ways of closing off a session but typically reporting on sessions is largely an observational task that AM’s VMR team can inform.
If Google Tag Manager/Google Analytics is not possible then logging similar data to a CSV is possible. Requirements for CSV Analtyics:
- Rotated weekly.
- Data older than 3 months purged to avoid space issues.
- One event per line.
- Standard ASCII (not compressed)
- Contain a single header in each output file. Delimited in the same manner as each data row.
- Text strings should be in double quotation marks.
- Delimiter could be a pipe symbol or other less common symbol to avoid a delimiter requiring escaping in a string.