I started going through my activity history to see what did I do to gain that many steps. Was it from my Starbucks afternoon break (that chocolate chip muffin was really good)? Or from walking back and forth between the devs and my CEO trying to reproduce the new bug he found on our latest feature?
Not too long after I started looking, I found it. I got almost a thousand steps from a 17-minute motorbike taxi ride home. Mi Band 3 tagged the activity as slow walking, which was pretty weird.
I started to research online and found out that many people have the same problem with me both from Mi Band 3 and a high-end competitor, Fitbit. Reading through the Reddit posts and many community forums, I discovered that the technology around counting steps might not be able to differentiate between each activity as accurately as we all want (although Apple Watch users do not seem to have this issue).
Mi Band 3 offered a behavior tagging feature in 2017 to solve this problem. You need to open the app, go to the tagging menu, choose the activity you are doing, and click start every time before doing anything. And do not forget that you also need to end the tracking manually after you finish.
I am happy for you if you can remember to do that, but I have to admit that I am not capable of completing all these steps throughout my hectic day.
Believing that the time spent, engagement and user retention are one of many metrics to be optimized; I know that a few changes to improve the behavior tagging aspect of Mi Band 3 could make a significant impact on key metrics.
These are some possible product improvements Mi can experiment with:
Edit or Delete Activities
People that use the fitness tracker generally understand that it can’t be 100% accurate. So they would not care if the steps are a bit off. But if your goal is 8000 steps a day, gaining almost a thousand from 17-minute motorbike ride is a bit much. Having an option for users to edit steps, calories, or periods of any activities would allow users to control their own activity history. I believe that just having a delete option would be the best MVP since it would require fewer resources, especially on the Information Architecture side.
Allowing users to label activity might be more tricky to achieve but could expand the feature use case a lot further. For example, users can label some action as at work, with friends, or party. Tagging past activity feature should also allow users to see the activity dashboard by each label, which means users could include or exclude any activity from the dashboard as they wish. This feature will also make the app more embedded into users daily life since they can log any event in the day in Mi app, not just workout-related stuff. Being able to integrate with users’ life events might also result in longer time spent on the app and higher frequency usage as well.
Well, you have the tagging feature already, right? Each behavior has its’ own calorie burn rate. At the moment, users can go to the tagging feature and start the timer before they eat, sit, stand, or drive to tag each activity manually.
But during the day, everyone is so busy with things in their life already. Users will not have time to complete a 6-step tagging process every time they do an action such as driving or eating.
But what about at the end of the day, when you come home, chill, and look at your fitness tracking app? You want to see how good you have been with staying active today. And that’s the most ideal time for users to make small edits to their tags and activities.
Users can spend this time tagging activities they have done to understand their behavior more. For example, the app is showing a 20-minute walk, 1000 steps from 3 pm to 3.20pm which is equal to 15 calories burned. If the user tags this activity as eating, it could go down to 0 steps and five calories instead. The longer time spent during this editing process could be a minor metric that leads to long-term user retention as well.
After letting users tag or label activities manually, Mi can also use machine learning technology in order to perform these actions automatically in the future. If a user tags his activity at 7 pm as driving every weekday, Mi should be able to auto-tag or suggest the driving tag for this user.
Pause the Tracker
Pausing the tracker could be the easiest to communicate option. Mi can let users pause the tracker via the app or the watch screen. Simply put the pause button on the steps counter page or let users tap three times to complete this action quickly. Mi Band 3 users are well-aware of how incredibly cheap it is already. Mi doesn’t need a fancy feature to please users, just a functional one.