TV UX: Spotlight for TV
There is a massive discrepancy between how important television is to our lives and how underwhelming its experience is. For a screen that is the centerpiece of our homes and content at the center of pop culture, the TV experience pales in comparison.
The biggest struggle every TV watcher faces is what to watch. When they turn the TV on, the watcher deals with too many choices, with hundreds of live channels, tons of apps, and shows and movies. It is too much of a good thing, and it overwhelms the user, who either spends close to ten minutes figuring out what they want to watch or retreats in the comfort of a show they already know.
To figure out ways to improve the TV experience for our OEM partners, we sought to find answers directly from the watchers. Reaching out through a Facebook campaign, we sent out a questionnaire to more than 10,000 persons about their TV experience, how they choose to watch their TV shows, their habits around TV streaming, live sports, recommendations, appointment watching.
We also built a proper living room in our office, fitted with cameras, mics and one-way mirrors, and invited more than 100 users over the course of three months to spend time in our living room watching TV, undisturbed. This way, we were able to capture candid remarks, thoughts, and choices.
The TV User Lab also allowed us to dive deep into 10-feet usability and refine our interfaces accordingly. For example:
- Because we found out that users had issues understanding where their cursor was in most TV interfaces we tested, we decided to overemphasize the focused element in every screen, by adding a size change, an outline and glowing effect on the focus element.
- We tweaked the font size, hierarchy and contrast between elements to properly highlights the deciding factors of the interface.
- We tested out the optimum item density on screen, and found out that 4 visible items per row was the combination that combined image recognizability, logo and copy readability, while allowing the user to process the amount of choice the best.
Some of the key takeaways were that users don’t necessarily trust or comprehend the recommendations they are given by algorithms: only 55% of the participants say they have started a show recommended to them by an algorithm. In comparison, 73% started a show on the recommendation of a friend.
In the same vein, time was an essential factor: 90% of the participants decided on watching a specific show depending on the time of the day. 90% say they don’t choose a particular show because it is running too long: for example, 60 minutes can be a considerable commitment for two parents who do not have much time in the evenings.
Lastly, while 95% of people commit channel-surfing on live TV, meaning they wander from channel to channel hoping to find a show they will enjoy, only 10% do not mind the experience. Interestingly, the time measured spent channel-surfing was roughly the same as the time spent looking at streaming recommendations. The random act of going from one channel to another seemed as effective as the intelligent algorithms.
We retooled our discovery layer with these insights in mind: instead of a user score that feels too abstract for users, we surface a different array of positive signals for shows we think a user would enjoy: how many of their friends have recommended it, if a surge of watchers recommended a specific show, recognized and understood scores like Rotten Tomatoes, awards, similarities between shows.
Privacy is an essential factor for watchers who do not want their watching history to be known for 91% of participants. For recommendations, the mechanic of watching is not necessarily a good signal of enjoyment:
- Some people watch a show but don’t love it.
- Some people leave things in the background.
- Some people have guilty pleasures that they don’t want broadcast.
One of our key research takeaways was that while people don’t want their entire viewing history in the open, they sure love to recommend and be recommended shows: 88% started a show that friends recommended. To take advantage of that fact, we introduced a different mechanic that put people in charge. Instead of using their passive watching history as a signal, we put users in control, made them active participants in the process, and introduced a “Recommend” mechanic.
The ACR technology that allows identifying all content on-screen also provides the opportunity for quality-of-life improvements: notifications when a followed show has a new episode airing, when your team scores a touchdown, when a commercial break ends, when a surge of viewers make a show trending. It also allows brands to reward users that stick through their commercials with offers directly.
