New platform VI: media landscape means different technical demands
27 July 2017
The new digital environment of Voetbal International makes one thing crystal clear: changes in the media landscape mean that the platform and CMS must comply with new technical demands. It’s about speed and being ready for new revenue models.
At the end of April Voetbal International launched a new digital environment: the website and the apps have undergone a complete overhaul. In terms of content, opportunities have been created to use new revenue models – through branded content, for instance. The new VI PRO also has a key role. This new type of subscription intertwines articles from the magazine with public and exclusive content. Now that it has become easier to integrate data, new possibilities exist for personalisation and matching content. This data gives the publisher the possibility to develop new revenue models in the future.
Under the bonnet: the demands of a new media platform
But what are the technical demands that such a platform has to fulfil nowadays? During a pretty intensive ‘track zero’ we examined with VI the demands of a media product and came up with a technical vision for this.
As in sport itself, a key requirement is the ability to peak. “It’s a race against the clock,” says VI’s Product Owner Mathijs Kraai. “A bit exaggerated, but the publisher that is first to publish the news gets all the traffic.” First and foremost, this means that a CMS is an editorial issue. Ultimately, that boosts the speed with which articles, videos and other content can reach the visitors.
Those more extreme peaks also have unavoidable technical consequences for a publisher. If there is really big news and the editor sends a push message from his CMS screen to all the app users, then in one fell swoop the number of active users increases from four to fifty thousand. A new platform therefore has to take such a peak load – and a future multiple of this - into account. If it doesn’t, and the content id slow or delayed, then the user experience plummets to really bad.
Not an unimportant factor was the wish to also arrive at an extensive data strategy that can allow VI to personalise the online experience. By exploiting content preferences, reader loyalty and engagement will only increase.
The Feyenoord champion- scenario
So it was quickly decided to use the Magnolia CMS: an extendable integration platform, or headless system, that enables all kinds of distribution channels to be linked using APIs. The VI apps developed simultaneously are thus directly connected to the CMS. And if any new channels appear, the CMS is ready for them.
This makes the system future-proof and, moreover, user-friendly. The CMS screens are designed to fit in with how the VI editors prefer to make, curate and distribute content in the fastest way. In contrast to many other CMS variants, the editor has the freedom himself to change settings depending on the situation: the article structure is just as easy to adapt as is the spot where the paywall appears in the piece. It’s all about these details if you want to entice readers to continue reading.
A ‘walking skeleton’ was co-developed with the people at VI. In this very realistic test environment the performance of the model at peak load could be tested after each sprint. Feyenoord’s championship win was regarded as one of the biggest conceivable peak moments. Every new function or change in the platform had to be capable of handling this calculated increase in traffic.
That posed a major challenge, due in part to the extensive form of personalisation. Because the addition of extra server capacity has its limits. By working with fragmented caching and by integrating the open source authentication solution Keycloak, not only was a secure solution created but also a scalable one capable of handling any conceivable situation.
A change in the VI team and the change in strategic direction gave the project a completely new dimension. Up till that point VI worked with a paywall that excluded (in part) non-paying readers. In VI PRO that black-and-white paywall has been swapped for the introduction of new forms of subscription. This has prompted greater attention for online revenue models: everyone can now read five PRO articles free of charge.
The data strategy thus proves its worth: using previous surfing patterns, the platform now makes personalised content suggestions. This brings visitors into contact with the premium content and the idea is that some will be converted via a ‘trial’ to a paid PRO subscription.
Shortly after the launch of the site and apps in April this year, the first successes came. For example, of all the news sites VI was quickest to publish the report of the Ajax-Lyon match and push it in the direction of the app. The platform has also coped with all kinds of peak periods: the highs and lows during the days when Feyenoord stuttered towards the championship of the Dutch Eredivisie were handled without any problems.
But perhaps even more interesting is the commercial effect. The PRO content receives one and a half times as many users that read more and spend more time reading or looking at other content. The biggest success was achieved by VI just after the article about the appointment and then non-appointment of Henk ten Cate as coach of the Dutch national team (‘Waarom Henk ten Cate bondscoach was en daarna weer niet’). The article behind the paywall was so well shared via social media that it was read 130,000 times. By ‘kicking down’ the paywall that enticed readers to register, that single publication led to 2600 new people signing up.
It is still difficult to say how many extra subscriptions have been taken out or whether users are easily enticed but Product Owner Kraai considers the initial figures very encouraging. The technical basis for speed, performance and data has to allow the publisher to grow. In-depth content and the combination of statistics and live events give the new source of income even more possibilities to expand.
This article was published earlier in Emerce.