Analytics Measurement Plan
Analytics can be easy to add but difficult to employ usefully. As with all Research, it's important to [[!Start with why]] and drill down to key questions and metrics from there. One method suggested by the Nielsen Norman Group is a measurement plan:
Create a measurement plan
- Goal - Macro conversions/big actions users need to complete for the product to be a success. Ex: X purchases completed.
- Desirable actions - Micro-conversioins/small actions that combine to support a goal. Ex: visiting a specific page/clicking a particular button.
- Metrics - Data that indicates if these actions occur.
I like this as it flips the script, forcing you to start with your goals rather than falling into the easy trap of what metrics you have access to. However, I think it's key to remember your overall research goals and 'whys' and fit this kind of plan in there. Analytics cannot measure everything or answer some, often crucial, questions and so this is really a very specific approach.
Also good to note that the goals described here are focused on product success, not customer success. Ideally those two should be intertwined and product success support customer success but how we know we're successful and how customers know they're successful can be two completely different metrics. The latter is likely better targeted with [[!Qualitative research]].