Yle reached two significant milestones last week. The first was the fact that we published a new browser version of our Web TV and Radio player Yle Areena; users are now greeted with a more modern service, but perhaps more importantly they will also encounter the first steps towards an actual personalised service. This moment was also of huge symbolic significance for the organisation since it meant that practically all of our applications and clients for Yle Areena are now running on top of our new API architecture.
The second milestone was what we are announcing here today, that the leadership team of Yle has confirmed the decision to open these APIs to the public initially as a trial and for experimentation. The leadership team is led by CEO Lauri Kivinen and includes people like the Director of Media Ismo Silvo and Janne Yli-Äyhö the COO who have all been crucial in making this strategy happen.
This week was the result of many years of preparation. In particular, the change is based on a vision for our digital platforms that was formulated in late 2012 and early 2013. The vision was radical for us as a traditional media and broadcaster organisation, and gaining buy in throughout the organisation was not easy. As I now look at the blog post I wrote about the vision at the time, I see that “our goals are that in 2015 the Yle API will be open for our partners, developer and user communities and anyone interested”. After a couple of years of skilled work behind the scenes by highly talented software architects such as Kalle Ylä-Anttila and Juha-Matti Perttunen and our brilliant dev teams, the API team in particular, as well as the relentless efforts of Aleksi Rossi, who as the head of API services was tasked with opening them up, I am happy to be able to deliver on that promise as we are today announcing the first phase of opening up our APIs to the public.
Wohoo!
Why is this important?
Back in 2012 the initial vision was prompted by a couple of important realisations. The first of these involved the changing habits of our users and how media consumption is becoming ever more atomised, personalised and mobile. This means we needed to be able to quickly experiment with various solutions combining the diverse forms of content that we produce. It also means in particular that we needed to create, collect and utilise data in a much smarter and finely grained way than ever before. The latter would justify a whole series of blog posts about profiles, bubbles, recommendations, competences, privacy and much more, but I’ll leave that for another time.
The other realisation involved the rising mobility and platform fragmentation; we understood that it is in no way possible to keep up with the development of new platforms in terms of devices, operating systems and clients. We calculate thousands of different platforms using our services each month. The pace is constantly intensifying.
In order to more easily meet both of these challenges we decided to rebuild our web platform into an API structure, which makes it easier and faster to build and develop solutions whilst also establishing multiple independent projects. An API structure also makes it possible for us to open our products to outside talent.
We know that this year alone companies like Netflix are spending 500 million dollars on developing their product. This is a good illustration of the platform challenges we are facing, no doubt a lot of this investment are used to keep pace with the platform fragmentation. Obviously, they are also known for building an extremely data centric product, and that doesn’t come cheap either.
Netflix is much less complex as an offering than what a public service media company ever will be, yet the product development budgets of the latter are hundreds of times smaller.
However, the users do not care about budgets. The bar is set by the global players, and that is the level our users expect us to offer them as well. As they should.
We needed to rethink how we think about product development. When we now look at the Areena product, there are three layers of developers. In the centre you’ll find the extremely gifted Areena development and design teams headed by Kari Haakana. Their responsibility is to establish applications and experiences that are the most strategic and important for our users. They also create the Areena experience both in terms of content and service.
The second layer consists of our partners. Several are the makers of big platforms or companies that we work closely with, who build apps for Areena; these partners can create Areena applications for their platforms effectively and independently by copying the logic of the apps that our internal teams have built. We always work closely with them to guarantee the quality that our users have come to expect from Areena.
The third layer is the independent developer, the start ups, the students, the tinkerers, etc. This is the messy undergrowth where experimentation and innovation usually thrives. We hope that future builders of Areena like services will be found here, and we also hope seeds for future international start ups might be grown here. This is also where solutions for very specific needs can be met, for instance for small and vibrant communities that still aren’t large enough to remit a centrally built solution.
This messy undergrowth is not entirely without risk, but there is always a balance between risk and utility. I at least am a believer in the famous law – attributed to one of Sun Microsystems founders Bill Joy – which states that "No matter who you are, most of the smartest people work for someone else”. As an enterprise we need to focus on building our organisation as a system including the systems that power it so that these extremely smart people working for someone else can easily get in when they want to.
The API is by no means the only project where Yle is embracing openness; we have a specific strategic focus on building new partnerships, developing our campuses to create more open hubs for creativity with features such as co-working spaces and sharing the spaces with other organisations and institutions. Furthermore, we are currently running our second accelerator for media start ups to highlight a few examples of the current state of development.
This is part of a much larger transformation where Yle, like so many “legacy” organisations, is learning more network based ways of thinking and organising as the old powerful know-it-all and hierarchical structure just doesn’t cut it anymore.
In the end, though, it all boils down to this: We are a publicly owned and funded organisation, and therefore it is imperative that whatever can be opened should be opened.
More info about Yle API here: developer.yle.fi
Sami Kallinen
Head of Web & Mobile Development
I'm a humanist that dreams of 8-bit sheep. Work @ Yle, The Finnish Broadcasting Company. Head of Web & Mobile Development. Three main responsibilities: 1. digital platform vision. 2. Heading the team that prioritises and operates the Yle digital product development portfolio and 3. the team that is does insights, data and analytics, recommendation, discovery, findability and SEO plus Yle ID and Yle API development.
