Tracking User Metrics #3: Gather User Feedbacks

User feedbacks come from many places: user surveys, in-depth user interviews, app store comments, social networks feedbacks, beta feedbacks, support tickets, … It can be overwhelming to naviguate this flow of data and after some time you might wonder: what the f*** do I do with all this data?

Learning how to use quantitative user feedbacks that goes beyond analytics can be complicated. I had to search a way to gather all those feedbacks, give them value and learn how to analyze and navigate through them. Thanksfully, after some tool comparison I ended up chosing Productboard.

I started by importing all the user feedbacks we already had through many sources (around 13.000 feedbacks) and started tagging them manually. While for English products you might resort on an automatic learning, usually it’s not happening with French (complexity of language I guess).

I created several categories of feedbacks: bugs inside the application, improvements about sections of the applications, demand for support outside the usual countries, … The aim was to be able to articulate the user feedbacks in a clear and efficient way, making it easy to naviguate for anyone with access to it.

Each feedback had several metadata: anonymized ID of the user, source of the feedback, related metrics (for example for an answer to a UMUX-L or NPS survey), date and the feedback itself.

Once all the data was sorted, I worked in two cycles with both Product Managers and Lead Developers to be able, for each feedback, to give ballpark estimates on some key indicators: complexity of development and impact on the business drivers.

Aside from ProductBoard, I tried to work on a Product timeline with the modifications deployed on our app to be able to spot if an increase of a certain kind of feedback was coming from something new in the application.

As ProductBoard allows you to filter on a timeframe, it was easy to see relevant feedbacks from last month or last quarter and to be able to pinpoint exactly the impact of some of our launches.

This system is working well to sort a lot of user data but comes with some caveats: you have to be able to tag feedbacks manually each week (I usually set aside 30 minutes a day to sort the feedbacks), you can’t really use visualisation on your data and you can’t include user interviews inside ProductBoard as this could break the analysis balance. However, if you have a lot of user feedbacks to navigate through and handle, it does the job quite well.

I’m currently working on making a switch to a full AirTable system to be able to aggregate, analyse and work with user feedbacks. This will allow us a better handling on our user feedbacks while giving us a lot of tools to sort, analyse and visualise them. I hope to be able to give you a bit more feedback in the coming months!

Pro of ProductBoard:
- Easily plug a variety of data sources
- Ability to highlight only parts of a feedback
- Ability to create personal categories
- Honest pricing

Cons of ProductBoard:
- Most of the new things are added only for the top plans
- Can’t divide from data sources once sorted
- You can’t create graphs about user feedbacks