At ustwo, we know the most difficult products to create are those which bridge the gap between art and science. With adidas go, we knew we had to merge the hard data from sensors with the very subjective and personal concept of music taste. We had to understand what runners want from their music when they run, and figure out how we could deliver those needs in an app that would not only meet our standards but at the same time be technically feasible.
In the end of 2014, adidas had – together with their creative agency Animal and Spotify – identified that 70% of people were running regularly with headphones, listening to music or podcasts while making their effort towards a healthier lifestyle. Without a doubt, there is a strong connection between music and moving – they belong together. Most running apps on the market are based on numbers and collecting data such as distance, pace, steps and time and only using music as a driver. With running and music having such a tight connection, we figured there must be something else we could do.
Testing, testing, testing
We set off asking people who normally run with music to specifically think about their music experience on the go. We needed them to think about and reflect on things like the tempo of the songs and which genre they enjoyed the most. Over a couple of weeks, we did rigorous testing, changing songs based on pace – only to find out that what is fast for one person might be too slow for another. We tried changing songs more often to boost energy – only to realise that you were really looking forward to a specific part of a song. We experimented with all possibilities of available sensors for heart-rate, weather, body temperature, terrain etc.
Two things became very clear. By serving music in your stride rate you improve your running style as it helps the body to keep a steady tempo, this means you can endure longer runs. The second thing is – by increasing your strides per minute, you actually lower the risk of getting injured, since it lowers the impact of each step. The novice runner has somewhere between 150-160 strides whereas the professional has around 180 strides. Matching the runner’s stride-rate with a song’s BPM also creates an absolutely magical feeling, making your run a whole different experience.
Creating a game-changer
adidas go set a new standard for how running apps in the future will be constructed, not only registering the data users produced while running but using this data to deliver a completely new running experience. After launching adidas go in April 2015, it quickly became adidas fastest growing app to date. Soon after, the other big brands started using the same idea with combining running and music in a new fashion.
Through a real-time algorithm, we used the runners step count and their Spotify music profiles to create real-time playlists that listened to you and your movement. There’s something truly magical that happens to your experience when you match your stride to the beat and music becomes a true motivator rather than a background sidekick.
Future of running apps
With adidas go we wanted to create an experience that was more emotionally connected to the user than facts and data. And it started something there we are curious to explore. What if you could listen to the same music as your friends at the same time but pitched to your own stride-rate? Could we create real-time synthesized music based on data sets such as pace, stride-rate, heart-rate, weather and so on? Would it be better? Would your running be easier, faster, longer or less injured? The only thing we know for sure is that music and running belong together, and they just started a whole new relationship.
Are you interested in using machine learning to improve your product or service? Contact us on firstname.lastname@example.org