“Let’s put a pin in it”: Intuitive Content Saving on YouTube’s Mobile App
With over a billion hours of video watched daily, YouTube is one of the largest and most influential content platforms of the modern day. It’s algorithm, while flawed, is regarded as one of the best at determining what its users want to see, consistently delivering new and interesting videos.
However, one thing Google doesn’t (yet) consider is the habits, mood, and free time of their users. It’s quite a common occurrence for a long video to be recommended when users may only have a few minutes on hand. A podcast that shows up while in the mood for spectacle, a movie trailer in the feed when scrolling for background music. These are all cases where the recommended content is interesting and relevant to the user, just at the wrong place or time. They’d love to see it later, and in fact their finger is just hovering over the clip, tempted to throw their responsibilities to the wind and indulge in YouTube’s endless flow of entertainment. But when the finger reaches the screen, it flicks up, sending the unfortunate video careening off the screen into the virtual void, relegated alongside that which is “spam”, “not interested”, and “unappealing”, to sink back beneath the vast ocean of videos in Google’s servers, unlikely to surface again.
A Relic of a Different Era
Credit where credit is due, the problem of saving a video to watch later is something YouTube has thought of long ago, adding the feature under the apt name: “Watch Later”. In fact, this feature has been around since late 2010, over a decade ago at the time of writing. It created an automatically generated private playlist in the user’s account called “Watch Later”, in which videos could be added or removed via a button in the video player. It got translated along to the YouTube official mobile app, which landed on Android at around the same time and onto iOS in late 2012. It put the action of saving to Watch Later into a little three dot menu in the video card, and buried the playlist under the user’s account page, where it remains to this day.
As you might imagine, a system that was never designed for mobile, adapted in an age when the rules of mobile design were still in the midst of being written, ends up falling a bit short. Don’t get me wrong, it’s perfectly functional, and I myself use it on the daily. However, with the passage of time and how deeply buried the interface is, it’s fallen out of common use, replaced by individualized solutions that bypass the provided option.
Breaking Down the Problem
Users have one of two choices when stumbling across a video they can’t watch right now. They can either let it slide by, hoping for it to stay in the feed until they do, leaving their content at the tender mercies of the Algorithm. Alternatively, they could record it in some way to come back to, either with the built-in solution or a homemade one.
In talking to users, I arrived at several key insights.
- It’s remarkable how little people use the inbuilt option.
- Several opted to use primarily the Desktop version, allowing them to open many videos in new tabs, coming back to the whole tab at a later date.
- Several others took it upon themselves to organize the videos in custom playlists, ignoring the default one provided.
- More still opted for the most ancient of solutions, remember the title or channel and search for it later.
- It was almost always a time constraint issue.
Nearly every time, the feature YouTube engineers spent many hours building in the early 2010s was replaced in favor of more convenient options, ones that were fewer taps and clicks away.
Building it up from scratch
Given the improvements in app design paradigms over the last decade, what can we do now to implement a mobile first, accessible solution to take the place of YouTube’s Water Later?
Right from the start, one particular solution stood out: pinning. A paradigm popularized by forums and chat clients, whereby a particular piece of content is selected to always show up at the top of the feed. YouTube uses it themselves in their comments section and live chat, used by creators to bring attention to a particular post.
In this case, it wouldn’t quite be the traditional use of pinned posts, as those are usually for others to see, rather than for oneself. However, it provides a good basis for which we can build an interaction for keeping videos you want in your feed.
Alright, we have an idea. Now let’s make it look good. The initial criteria for these entry points is that they’re unobtrusive, easily accessible, and fits in with the existing design language.
Of these, I initially preferred the long press action the most, however it seems that the iOS version already has the existing context menu bound to long press.
Alternatively, the pin button best replicates how YouTube Desktop’s Watch Later button works, but without being able to respond to the hover of a mouse, it clutters each and every video card.
The Swipe to Pin action was inspired by Gmail’s swipe actions, and is as simple and discrete as long press, if very brightly colored when doing so.
The menu option represented a baseline, an action equivalent to the current system.
Narrowing it Down
As a few iterations went by and the design was refined, a key insight allowed us to settle upon “Swipe to Pin” as the preferred method of interaction: Users didn’t need to see the video in their feed anymore after they’ve saved it, so swiping it away makes sense. It’s much the same reason why the archive action in Gmail is built the same way, where swiping means it’s no longer in your feed, cluttering up the space.
Now that the video has been saved, we need to provide a convenient way to access it again. This can be thought of as the largest flaw with the existing Watch Later system, as it requires digging into the user’s Library tab to find the location of the Watch Later playlist. That’s three taps just to see what’s they’ve saved, nevermind watching it.
Continuing with the paradigm of pinned content, the designs are all variants of floating pinned content to the top of the recommendation feed.
There are three categories of users we have to consider here.
- Users that don’t save videos
> We have to keep the new feature unobtrusive so that something they don’t use doesn’t take up extra room on their screen.
- Users that save a small set of videos
> Videos must be easily accessible, enough to justify the existence of this feature over Water Later.
- Users that hoard videos
> Videos must not take up too much space when the list gets large, as it would crowd out the new recommendations.
The first and second types of users are served well by simply floating pinned videos to the top, however, the third might make it unusable.
To solve the issue of excessively large lists of saved videos, we can segregate the pinned videos to their own section, with an button to expand and collapse it at the top.
Bring it all together, and we have a sliding action to pin a video to the recommended page, removing it from the feed. These videos can then be accessed with just one tap at the top of the page, expanding the list of videos onscreen.
Compared to the existing system, it’s one fewer interaction to save a video, and three fewer to watch a saved video. It provides a framework to add more swipe actions to YouTube in the future in the same vein as Gmail’s mobile app. Finally, it remains discrete, not impacting the user experience for those who don’t bother with archiving content for later.
Now, I’d be remiss if I didn’t mention the downsides to this new design as well. It’s bright, single-colored action is quite jarring against the general look of YouTube’s app. The discrete nature of the feature means that, like the current “Watch Later” feature, it’s hard to discover and might not ever be used. It doesn’t translate well to the Desktop experience, unless merged with Watch Later, something people already might have a flow for. Finally, there are questions about the implementation that remain unanswered.
If a user watches a video, does it get removed from the list? If someone pins more than 20 videos, do all of them show up in the feed when the section is expanded? Do we implement this feature to the recommendation feeds below a playing video? What about the horizontal feeds for news, how would this interaction fare then?
In hindsight, the easiest fix would simply be to keep the current two tap interaction for Watch Later, and move that playlist into the recommendation feed along one of those horizontally scrolling lists that the Top News carousel currently uses.
But then that wouldn’t quite warrant an 8 minute article. :D
This is a case study for a project in Intro to Digital Product Design. I am in no way affiliated with Google or YouTube.
I am a junior at Cornell University, pursuing a major in Computer Science, with minors in Game Design and Information Science. While I’m typically on the software engineering side, this was my first venture into a formalized design case study. I hope you enjoyed it.