This article was originally published in Rapid TV News.

Developing an incredible OTT TV app that meets your business needs and also creates a real buzz with consumers can prove tricky. Doing this across 10 plus platforms (like AppleTV, Amazon Fire, Roku, connected TVs etc.) sounds impossible. But it can actually be done, if planned and executed correctly. Here are some of the lessons we learned when doing this.

Planning

Building apps for such a fragmented market requires vision, a well thought-out development plan, and relatively deep pockets. But get the app development right and, provided you have great content too, the positive results will speak for themselves.

Creating TV apps is really expensive so combining multiple business models – transactions, subscriptions, advertising, and freemium – to provide flexibility and maximize revenues is critical. Plus adding dynamic ad insertion, VOD and cloud DVR functionality can further increase revenue potential. Much like in the print magazine world, where consumers used to subscribe to more than one magazine to reflect their multiple interests and hobbies, most people will eventually subscribe to more than one OTT service too. But for this to happen subscription prices need to be squeezed.

In this scenario, having a TV app that supports multiple business models – and makes upselling and cross-selling as friction-free as possible – is the only way to realize revenues and profits.

All of this requires a very strong backend system that offers all of the API calls needed to manage the various business rules. In addition, the same platform must be very smart so it is able to distinguish between devices. With the massive fragmentation in the market, different devices require different technologies to perform the same task. For example, the backend will need to support at least three different DRM standards to deliver encrypted content to all devices;  the same backend will likely need to support two or three streaming technologies; and the same complexity exists for tracking analytics and running ads. Therefore, you will need a backend that is like a toolbox and can use different technologies in different scenarios.

Writing Code

In such a fragmented market it would be great to reap efficiencies by using the same code with minor variations. Unfortunately, from a technical point of view, this is not easy. Most of the development for smart TVs, Playstation, Wii U, and Amazon Fire is done with a combination of JavaScript, HTML, and CSS. Replacing the HTML and CSS with custom UI or libraries opens up Apple TV and Xbox One. And Xbox 360 requires C# and XAML. But other platforms use proprietary languages: Roku uses BrightScript, Android native is required for Android TV, phones, and tablets, and of course iOS devices require iOS native. So what’s the best approach?

The smartest thing to do is to use what works in the largest number of environments. Start with web development and build outwards from there. Each TV platform has its own design guidelines. Take the guidelines and focus on creating one great design that works across the overlapping 80%. Invest the time you have saved by creating only one design into making that one design fabulous. And don’t be afraid to divert from the design guidelines a little to make your design stand out from the rest.

The approach above is great for connected TVs, Android STBs and Amazon Fire device. For iOS and Roku, you can develop quickly by relying on their templates. Once you have launched on all devices, you can go back and build customized apps for iOS and Roku which will match the look and feel of your web-based apps.

Launching

There are two ways to launch.

Approach one: launch wide but lean. Develop lightly for as many platforms as possible as quickly as possible. The key is to learn quickly—from your growing experience with the platforms, from usage analytics, and from user responses. Then iterate just as quickly. Invest heavily where initial results show promise and reinforce there. As you gather data, you base later-stage strategy on results. Don’t spend too much money in places where it might not be spent effectively.

Approach two: do a Netflix and go all out to cover all platforms from day one. The fact that this is still a relatively immature market means that there’s still market share up for grabs. If you can move faster and stronger than the competition, you can own it. The downside is that this requires immense resources up front, to create tailored apps as well as compelling content. It’s a bigger gamble, but with a commensurately larger potential reward.

Summary

 Creating a true cross-device experience is hard unless, of course, you have an army of developers at your disposal. But as platforms like AppleTV, Roku, Amazon and connected TVs become more developer friendly, smaller teams can build such experiences. Since the fragmentation in the video space is here to stay, this endeavor is worth the effort.

This article was originally published in Videonet.

Although OTT TV is clearly taking off and competing with traditional cable and satellite almost everywhere around the globe, some still wonder what makes OTT the TV of tomorrow? What makes it so much better?

If you think about it, the television viewing experience hasn’t changed much since This is Your Life aired in the 50s and defined prime-time as we know it today. Why was prime-time so important? Much of it was because those were the hours for the entire family to watch TV together, which made the viewing experience so much more powerful. It took a passive individual experience (a person staring at a screen) and turned it into a social activity (people laughing together, commenting and sharing their thoughts on the show in the days after).

To understand why OTT is going to be the way millennials prefer to watch TV, we need to examine TV time’s biggest competitor for young viewers’ eyeballs – Social Networking. Companies like Facebook, Snapchat and Twitter can teach us how to design a user experience that every TV service aspires to create: engaging and sticky.

It’s worth going back to some of the major social networks’ founders and seeing the principles they emphasized early in their networks’ success. For example, a young Mark Zuckerberg believed:

“The attractive component of this was probably that the base was so localized…we have so much data about what people are doing on the site and how people are using it that it just makes it so that we can really enhance the experience and target stuff towards those people in ways that no one else has really been able to do before.”

That focus on localization and data allowed for personalization, which is what has helped Facebook grow into the $350 billion company it is today.

Snapchat is one of Facebook’s biggest competitors because it dominates the teen/college students demographic: 7 out of 10 snapchat users are millennials. These are the cord-nevers/cord-shavers of tomorrow. So why do they love Snapchat so much? Here’s how Snapchat CEO Evan Spiegel explained their early success:

“Now the mobile phone has really powered the idea of instant expression, which is really showing someone where you are and how you’re feeling in the moment. And this is important as it relates to identity, because really that’s one of the things that’s at the core of social media.”

One reason Snapchat has been so popular is that has allowed people to express their identity as it changes in real time.

So, personalization, identity, and real time—here we have the pillars of social networking. These are also some of the core values that will make OTT services a success.

Identity: OTT’s Big Differentiator

Let’s start with identity. Until recently, the best TV could hope for was to apply some very basic demographics to viewership. With OTT, however, identity is fundamental to the user experience. It begins at authentication. A good user experience makes this easy, perhaps even allowing users to import their online identity straight from Facebook itself. Once users log in, the provider knows who they are, what they watched, what subscription they used or one-time movie they bought.

But it needs to go further than that. A user’s OTT identity needs to feel consistent to them on all their devices. Cloud DVR is the best example of that. It’s a personalized video recording library in the cloud that allows users to enjoy their content anywhere – via mobile, the web, or through connected TVs. Similarly, users can create their own watch lists and see their viewing history.

Turning on your TV app on any device should be like looking in the mirror: you should see any shows that you are in the process of watching, some that you plan to watch next, and a few smart suggestions based on your previous behavior. At the same time, DRM and concurrency tracking are also aspects of identity in OTT, as they are crucial components required to prevent identity credentials from being stolen or shared with others.

Personalization: The Heart of User Experience

When an identity can be carried throughout the OTT experience, this makes personalization possible. Today’s viewers expect a highly personalized service, from the recommendations to even the ads themselves.

Users don’t only want to manage their own favorite lists; they want to be able compare their own viewing history to that of their friends. They expect to be able to set their own notifications so they can be informed about new content that they may enjoy as it becomes available. Their recommendations should be optimized not only based on what they’re watching, but what their social connections are watching as well.

User experience personalization should apply to business rules, as well. Users want to create their own deals, choosing whether to purchase additional services, like cloud DVR storage. They want to pay only for the content that they want to consume. They want to decide what level of viewership they want to engage at—whether free users, paying users, VIP users etc. Giving them flexibility allows them to choose the price point and viewer experience they prefer. They are even coming to expect advertising to be more targeted and relevant to them.

Real Time: Building Off the Advantages of TV

Since its early days, TV has been about live and real-time feeds. A good OTT user experience builds off that, even when offering on-demand services. A good start is incorporating real-time social feeds based off Facebook and Twitter hashtags. Again, we can go further. Users could receive real-time show reminders and notifications. When a user’s favorite show is about to begin or when a movie by their favorite director has been added to the service, notifications could be sent to users in-app or via email, depending on how they set their preferences.

Real-time reactions are even more important on the backend. Real-time analytics are key to making sure user experiences are positive. If something goes wrong for the user, it’s imperative that the service provider knows immediately so they can fix it. Today’s viewers have little patience, and are not shy about making their displeasure known!

Great User Experiences

Social networks have exploded over the past 15 years; OTT is just starting to explode now. By building off the fundamental social concepts of identity, personalization, and real time, OTT providers can develop a great user experience for their viewers, which will make the TV of tomorrow – OTT – resonate so much better with millennials.

As well as the rise of comedian and actor Amy Schumer, 2015 saw the continual rise, and industry acceptance, of over-the-top (OTT) internet TV services such asNetflix and HBO NOW.

One of Schumer’s biggest achievements last year was her Live at the Apollo HBO special. In October, a few days before the show premiered on HBO and HBO NOW (the OTT subscription service that doesn’t require a pay TV subscription), my friend said she was about to sign up to HBO NOW. Her thought process was that she didn’t have a pay TV provider, but she wanted to watch the special as soon as it became available. “It’s $15 per month but you get a seven day free trial,” she said. “I can drop it whenever I want.”

I work at Kaltura, where we design online video services, so through the companies we work with we’re acutely aware of churn – customers discontinuing a service – and the problem it poses for OTT services in particular.

While churn was always an issue for service providers, OTT made it much trickier to fight. With paid TV, it takes one phone call to the call centre, convincing several different representatives that you are definitely quitting. Then you have to take the set-top box to the nearest service centre. It takes determination to go through with it. With OTT video apps, all you need to do is tap the screen and you’re free. The consumer’s dream is the service provider’s worst nightmare.

Subscribers cancelling OTT services as a percentage of the current subscriber base, in US broadband households. Illustration: Parks Associates

A study from Parks Associates (see graphic) sums up the problem. Viewers sign up and get their fix, knowing that they can drop the service any time and subscribe again in the future when a new series or another Amy Schumer special becomes available. In order for OTT to retain its momentum in the market and offer a real alternative to cable and satellite, it must retain customers. A leaky bucket will never reach a tipping point. Today, three main strategies can reduce OTT churn.

A/B testing

The process of showing a modified version of the website to a subset of users and then comparing their behaviour to the regular version proved to be highly successful, especially for Netflix, which has embraced this method from its early days in 2005. Netflix has run more than 1,000 A/B tests with tens of millions of users. It helped the brand redesign its TV app interfaces, launch the personalised log-in pages, improve search and more. The main goal is to increase usage, which leads to increased retention and eventually lower churn.

Raising the quality of experience

Increased usage isn’t enough if the viewing experience isn’t good. Viewers today have the same level of expectations from their OTT provider as they have (or used to have) from paid TV. In a nutshell, it’s the spinning wheel moment. When you turn on a cable set-top box, you don’t ever expect to wait one to two seconds for the video to load. OTT must be on the same level. The problem is that the internet was never designed to stream HD video at scale – and things get even more complicated when it comes to 4K or live video. If a video takes two to five seconds to load, about 20% of users will drop, and after 10 seconds, around 40% will have abandoned it, found a new study (pdf) from the University of Massachusetts Amherst and Akamai Technologies. It was previously hard for OTT operators to notice service issues, but today there are a number of quality of service (QoS) technologies that constantly track buffer time for every user and the video quality of every stream. If there’s reason to believe something is wrong, the operations team will get notified immediately so that they can address the issue by fixing the video file or switching to a different content delivery network. QoS may sound to some like a minor issue, but in markets with increased OTT competition, most vendors offer the same content at the same price. The main competition is on the quality of the experience.

Personalised and social marketing

Yes, those emails, phone notifications and in-app alerts that let the users know there’s a new show available. In OTT, such marketing campaigns don’t have to annoy the viewers. The service marketers should have full access to each viewer’s history so that they can send personalised alerts that will drive engagement and retention. OTT services should also embrace social networks to market to new viewers and go beyond Facebook or YouTube. Many of the young people in Generation Z today spend more of their time elsewhere like on Snapchat and Vine. Strong social marketing will be key for the growth of new OTT services.