Imagine an app that consistently offers you the news and articles you personally find most compelling and rewarding. Imagine the app providing you with a view over a vast media sphere. Imagine never missing a single novelty or insight. You’re imagining Sanoma Flou.
Sanoma Flou is an iOS and Android app, designed to offer users what they desire from a large selection of Finnish newspapers, magazines, radio stations and other media outlets. In our experience, Flou offers users so much interesting material, taking a break from reading and perusing what's on offer is actually difficult. User feedback has been overwhelmingly positive.
Flou provides an almost addictive and deeply personalized experience. The customers love it.
Reshaping the Media Industry
Flou's value offer has the potential to both capture customers’ hearts and reshape entire industries. Like Reddit, Netflix and Spotify, it offers customers all the content they love in a single place. These services are global distribution points, bringing together customers and global content suppliers. As a market position, it emerges naturally to reshape industry, thanks to a superior customer experience.
We believe personalized aggregators like Flou will dominate the digital media sphere within 5 years.
This is a good thing. The shift to digital has reduced the number of paid subscribers all across the media sphere, but independent quality journalism needs paying readers. Flou can capture readers’ hearts and it could introduce paid digital journalism to a much wider audience. Flou makes funding diverse media an automated process, like Netflix and Spotify. For small magazines this is a matter of life and death. Following and funding infrequently updated digital content would be inconvenient without Flou.
Services like Flou are an important and positive thing for the future of independent quality journalism, one of the cornerstones of a free society.
The emergence of personalized aggregation services has been a slow process. Aggregation is easy - personalization is not.
In Flou, diverse content meet an equally diverse group of readers. This is both an opportunity and a challenge. Articles about games can be incredibly interesting for a limited audience. At the same time, you don’t want to promote women's magazines to male readers. Content must find its own audience and good personalization is absolutely critical for Flou’s success.
Flou's talented team was driven by a strong vision: they wanted to reshape the Finnish media sphere. Matti Härö is a middle-aged Sanoma engineer who first recognized the concept’s worth and created Flou as unmandated skunkworks project. It was later included in the Sanoma strategy and business manager Tuomas Pättö took the lead. Tuomas was a business manager with the dream of building something big. He was driven to make Flou a huge success. At Futurice we used some of our best engineers to craft the bones and the brains of the solution, but the team lacked the necessary expertise to create a recommendation engine.
So there we were: meeting with Tuomas, the assigned leader of the cause, on an Autumn day. We had presented our proposal for the recommendation engine work earlier and he'd decided to assign us a role in Sanoma’s grand designs. It was never going to be an easy role. Expectations were high as the engine was critical for Flou and Flou itself was of strategic importance. And time was short. We had to productize the bleeding-edge algorithms in mere months. The work had to be ready well before the scheduled release.
Fortunately, we knew what we were doing, and - in fact - our ambitions were even higher.
We had built a recommendation engine before and we'd started to see recommendation engines more as a piece in the greater puzzle involving customer desires. Recommendations are useful, but customers' desires are the reason companies exist. So we wanted to extract the full benefit of knowing the customers’ desires. We wanted to personalize every aspect of the service and support things like analytics, advertising and marketing on the side.
We did not want to create yet another recommendation engine. We wanted to create a preference engine - a model for the desires themselves.
The most exciting thing was that by getting certain design fundamentals right, it all seemed achievable within the given timeframe. We pitched the benefits of this approach in our offer, it seemed to perfectly fit Flou’s needs and we were committed to deliver it.
The results were impressive. The Flou Preference Engine could provide 2-5 times more interesting content compared to the traditional time-based sort. Services like Flou are all about content and doubling the density of interesting content doubles the service’s visible value. This change significantly improved the user experience, too. Suddenly, interesting content was everywhere and it became difficult to stop reading. The user feedback reflected this. Truly, the little app had the charm to capture a customer's heart and it was clear we had been part of something amazing.
The power of providing delightful content was complemented by its amazing performance. We could personalize 10000 articles in 13 milliseconds. It also made it real-time personalization of all views, searches, more-like-this recommendations and notifications possible. Although personalizing every aspect of the service was technically possible, we only had time to integrate the algorithm with the ‘Sinulle’/’For you’ view. In Flou, this view acts as the main view and it provides a personalized a window over the full content.
We helped our customer move forward in their grand plans to reshape Finnish media. We see Flou playing a critical and positive role in the Finnish media’s transformation from paper to digital.
We also welcome you to contact us, if you need help in reshaping your industries through data.
- Antti Rauhala, Christoffer Ventus and Futurice
Futurice crafted the iPhone and iOS Apps, was involved in the backend development, and created the preference engine pre-release.
Interested in data? Read more about ABB and big support data, or how predictive interfaces make software more intelligent or how automatic learning of customer preferences improve your business.