Sooner: Focus on the user to conquer Europe “sooner” or later
Sooner is a relatively young OTT that is growing at a high pace. It’s available in the Benelux, Germany, Austria and Switzerland and it’s looking forward to expand its service to other European countries, amplifying its content to more than cinema and series. With a simple and direct navigation system, it focuses its aim on giving the user just what it wants. Daniel Siqueira, CTO at the company, reveals us how they are doing this.
What is the origin of Sooner? How did Content Scope, EYZ Media and Metropolitan Filmexport get along together and decide to create Sooner?
Sooner coalesced from the realization that market segmentation in Europe limits the growth and viability of VOD platforms. With Sooner we aim for an Europe wide brand that can serve local needs and mutualize costs both for technical infrastructure but also for content licensing and marketing. Metropolitan approached EYZ to develop a concept which could unite multiple operators under the same umbrella but also keep them independent.
With such a big market of streaming platforms in Europe, what makes Sooner different?
One key feature is the mix between Films, Series, subscription and single purchase content. We partner with film festivals to bring users the latest trends of the film industry and have streamlined the rental process making it more seamless where users won’t have the typical painful shopping cart checkout process. We noticed most platforms used very similar navigation and layout, which relies heavily on quantity, bombarding users with an endless wall of thumbnails and scrollers. In our assessment this can lead to choice paralysis, as formalized in the Paradox of Choice. We chose to simplify, centering the main selection covers and giving users a more focused approach when looking for titles. This exercise in decluttering the UI also led us to design a single title page that brings the content forward blending it into the UI, using a background video loop extracted from the film that’s harmoniously tinted with the dominant colors of the title cover. It makes you instantly able to judge key aspects of the content without distracting from the browsing experience, watching a trailer that might reveal too many plot points.
Which was the biggest technological challenge for Sooner?
There are many of course, in the beginning it was architecting a comprehensive system encompassing scalable serverless backend, database abstractions for multiple payment processors, a frontend design that could handle web, smartTV and mobile together, and an integrated workflow for the content ingest and marketing team. I’m glad to say our strategy paid off and is facilitating our expansion into more platforms. The challenge now is understanding the friction points and edge cases from our growing userbase and streamlining those rough edges. In parallel, something that emerged is managing the expectations of VOD users. We’re now used to larger, more mature platforms, it’s a challenge to keep the pace and innovate, given our brand has existed for less than a year with a fraction of the development team size of other players in this market.
What are the technological challenges of being in 5 different countries?
We’ve mostly solved scalability, and distribution across territories. But because we operate with different corporate structures on each region, we need to manage each region’s accounts across many services, like payment processors and app stores independently. This means setting up, configuring and practicing release management for each separately, which of course creates an overhead but allows each region to operate independently simplifying other business processes that would have to be developed were they merged under a single account.
Which image format and quality is available on Sooner? Why did you choose that format and are you expecting an upgrade?
We have a mix of new content and legacy content from over years of operation. Everything that’s new is always available in the highest quality we can get from licensors. This usually means an HD ProRes master streamed at h264. We already can deliver 2k and 4K content but these are still rare from the distribution side. As soon as the companies up in that chain can deliver in larger formats we’ll be able to make them more available to our users.
In which platforms are you available and what is the challenge of being in different platforms?
We are currently available on the web, iOS and Android and soon launching apps for SmartTV on Samsung, LG, FireTV and AppleTV. Being a small team, the main challenge is being able to reach as many devices as possible utilizing hybrid stacks to leverage web standards when available and implementing native modules when they’re not. The main challenge here is testing, acquiring devices and setups that represent edge cases of our subscribers can become a daunting quest. This is especially hard for SmartTVs as users expect the apps to work even on older, less powerful devices which have outdated OS, while on the other end, the ever evolving OS refresh rate and deprecation of APIs means at every new update, many changes have to be made and a lot of testing needs to occur.
How has the encoding of your videos evolved over the years?
Tremendously. I actually started working in EYZ as a video encoder about 5 years ago, back then we were ripping DVDs and Blurays and using quite primitive scripting to generate FFMPEG commands for encoding multiple video sizes. As time went on, connections got faster and licensors started more and more sending us iTunes packages or single ProRes files via Aspera, ftp and Dropbox. It was seeing the limitations of that workflow and the increase in online deliveries that sparked the development of a mostly automated encoding system that could recognize formats and file patterns from different providers and normalize their outputs. This system could recognize covers, stills, convert subtitle formats, automatically crop black bars and select the correct audio tracks from assets with many tracks and multiple languages while integrated via API with remote storage and both our content management system and streaming servers. With this, our encoding and editorial teams could have an overview of where in the pipeline each asset was and was automatically update, deprecating our previous method for managing using cumbersome google sheets where statuses would get stale as the management overhead was too big. Today we have a new system that follows that same principles and workflow but has been refactored with templating and as docker containers that can be deployed to many different OS both on dedicated remote servers but also on premises utilizing computing power that would have been idle otherwise.
How do you manage all the amount of assets that you have? Do you collaborate with any company?
No, everything is inhouse. Each region has a dedicated encoding department comprising highly skilled people able to manage and process an astound amount of data per day. We try to build as much automation as possible in the ingest pipeline, but because we receive content from such diverse range or licensors there are many edge cases that need to be handled manually. We also have image editors for image covers and marketing material and contract subtitlers to localize exclusive content, that can include simple translations or the creation of full subtitles from scratch.
Which content and streaming server do you use and why?
We utilize Kaltura SaaS solution, it’s a very mature system which provides stream encoding and serving, cdn hosting, image resizing and DRM, they also have an integrated player which offloads some of the more critical playback components and allows us to focus on the broader user experience. We made the decision to use a third party platform that specializes on streaming and are working around the clock dedicated to that task because owning those components didn’t really match our core strategy to focus on content and user experience, it required very specialized knowledge.
With such and specific content, how do you manage to offer a personalized experience?
Each region has their own particularities regarding availability of offerings. The Benelux region has more of a focus on french cinema and studio films while DACH focuses on independent film makers and european series. Personalization plays a big role in the discoverability and user retention.
You opted for curated content on you experts, so does this exclude AI developments focused on personal recommendations in base on customer experience?
We use a mix of curators with a constantly evolving recommendation system. Our editors are much better reading the time sensitive cultural context of each market and react to them by prioritizing the content the majority of users will enjoy. But this approach is way too broad and we think a recommendation system offers the extra edge which can react to an individual’s content journey and surface content that’s likely to peak their interest but might not be in the current zeitgeist. We give the users the option to sort the content based on their own usage and utilize many different processes to match their history with other similar type of content.
What is the future of Sooner? What is the next step for the company?
We’re currently bringing the Sooner content to more environments and will continue to expand and reach more users including new territories. In the near future I can see Sooner increasing its market share in key regions and bringing more local content including livestream for concerts, theater and partnerships with cinemas.
For our users, next in the roadmap will be integrating a social networking dynamic where users can create custom playlists, share and follow each other’s activities and favorites. It won’t be a full fledged social network, but we believe having social integration will allow the cinephiles and power users to build an audience by curating their own content journey and making that available for other users to experience the platform as they do.
Rating the content as well, we feel the current star rating or link/dislike logic of other services doesn’t really capture what attracts people to the content, so we are developing ratings based on a few criteria like story, character development and so on.
As much as our curators and recommendation system can help users discover our huge catalogue, we believe it’s the users who can tell us and each other what are the highlights they are most excited about in our collection, and that is very exciting for us!