When Japanese sportswear brand Asics announced its $85 million purchase of fitness tracking app Runkeeper back in February 2016, Runkeeper CEO Jason Jacobs had the following to say about the general market trend behind the deal:
The fitness brands of the future will not just make physical products, but will be embedded in the consumer journey, in ways that will help keep people motivated and maximize their enjoyment of sport.
And also in ways, he might have added, that maximize their purchases of those physical products. That, after all, is the real goal behind the Asics/Runkeeper deal, just as it motivated the 2015 acquisitions of Runtastic by Adidas and MyFitnessPal by Under Armour.
In short, sportswear companies that have traditionally sold physical products are looking to build ongoing, even daily, relationships with customers, by connecting with them every time they exercise, not just when they purchase a pair of running shoes.
Asics may have lagged some way behind its rivals in the rush to snap up an app, but the race to get full value from these acquisitions is ongoing between the sports apparel makers. After all, these apps all came with considerable price tags, so it’s hardly surprising that their buyers need them to boost overall revenues, either through helping them to shift more exercise gear, or through the sale of app subscriptions - but preferably, through both.
Analytics will play a vital role here, because the work of detecting seasonal, geographical and demographic trends and spotting new sales opportunities will rely on the mixing and matching of data from a wide variety of systems. At Asics Digital (the new name for the former Runkeeper business), much of that data is held in a cloud-based data warehouse from Snowflake Computing, according to Advanced Analytics Manager, Chris Drouin.
This data warehouse, says Drouin, hosts somewhere between 10TB and 20TB of data, drawn from Asic’s e-commerce operations, its fitness apps (including Runkeeper and Asics Studio), and its OneAsics customer loyalty programme. As he explains:
Our goal is to build a single view of the customer, to build a clearer picture of their individual fitness goals and journeys, and see what works for them and what doesn’t. And a particular focus for us is using analytics to personalize customer interactions, so that we’re not treating a long-term customer like this is their first interaction with Asics.
The decision to use Snowflake was originally made by Runkeeper, before its takeover by Asics, and its implementation pre-dates the acquisition. It replaced an AWS Redshift deployment that kept running into concurrency issues and what drew Runkeeper to Snowflake was the way it splits compute from storage. This means it can spin up ‘virtual data warehouses’ for specific workloads, which can be resized according to need, or paused from time to time, enabling concurrent workloads to run without impacting each other. As Drouin explains:
So we’ve got ongoing ETL [extraction, transformation and load] jobs going on the whole time that can’t be disrupted and we can split those out. We have data scientists like me using an ad hoc, business intelligence query warehouse, where we can go to at any time to get answers to some of the bigger questions we like to ask; and we can also split out resources for the reporting folks, who are wanting to connect Tableau [to Snowflake] for their reports. Being able to split out workloads into different warehouses was a huge use case for us.
Today, there are a number of new ways that Drouin wants to use Snowflake - and they tie closely to this idea of using a blend of physical products and digital approaches to generate more revenue for the company.
The first is an integration between Snowflake and Salesforce Marketing Cloud, to take Asics:
...from a blanket approach to a more personalized touch when it comes to the messages we send out to customers.
Take, for example, the ‘Shoe Tracking’ feature in Runkeeper; this enables users to monitor how many miles or kilometres they’ve while wearing a particular pair of shoes, so that they have some idea of when they should replace them. (General advice points to between 300 and 500 miles.)
While runners will be notified within the app that a worn-out pair of running shoes is nearing retirement, this integration with Salesforce Marketing Cloud could hypothetically enable Asics to back up that notification with an email that includes a money-off voucher for the user’s favourite model of shoe.
Second, Drouin and his team are looking for ways they can build out more data-driven features for the company’s fitness apps, as a way to attract new users and keep existing ones loyal.
This might involve taking third-party data from the recently launched Snowflake Data Exchange - such as weather data, perhaps - and blending it with Runkeeper data, so that users can see how their running performance varies in different weather conditions.
This is similar to work done by Strava, one of the few fitness tracking apps to remain independent and also a Snowflake customer. As diginomica reported earlier this year, Strava’s ‘relative effort’ score, underpinned by Snowflake, enables users to compare and contrast heart rate metrics from their wearable devices across different activities.
In a highly competitive market, fitness apps are increasingly battling it out on the richness of apps that they offer, and this functional richness frequently comes down to the richness of the data that underpins them. The more users the apps owners can recruit - especially if these are ‘elite’ users who pay for the privilege of extra features and functions - the more money they can make. And, if they’re selling sports equipment on top of all this, the greater their opportunity is to sell those products to loyal users already tied into that app.