What YouTube's 'Something Different' Test Gets Right and Wrong
+20 XPCase study
What YouTube's 'Something Different' Test Gets Right and Wrong
Late at night, YouTube notices a viewer scrolling without finding anything and offers to show something different. You will see why the first version made the problem worse, how a smaller set of choices fixes it, and how heavy ads undo the goodwill.
+20 XP when you reach the endWhat you will learn
- Recognize the paradox of choice in a content feed
- Give people clear feedback and an easy way back after a mode change
- Use fewer, comparable options to make a decision easier
Scroll to follow the story. The screen changes with you.
Based on the learnings from Growth.Design: YouTube's Attempt To Solve The Paradox of Choice, presented in our own format.
Step 1 of 10
- New to you
01
The setup
After a long scroll, a card asks if you are looking for something different.
Deep into a late night session, a card appears in the home feed: 'Looking for something different?' with a 'New to you' button. It shows up exactly when the viewer has stopped finding anything worth watching.
TakeawayListen to behavior in real time, and offer help at the moment it is needed.
Real time personalizationA great experience provides value aligned with users' needs at the right time. Nudging people while a behavior is still happening, sometimes called behavior profiling, makes the help feel timely.
Play Something02
Context
Netflix tried something similar with a single shuffle button.
Netflix offers 'Not Sure What to Watch?' with one 'Play Something' button that picks a title based on your tastes. Both companies are trying to solve the same problem: people who cannot decide, and leave.
TakeawayWhen a catalog is huge, deciding becomes the product's biggest barrier.
Another sea of videos03
The problem
The 'new' feed is another wall of the same kind of videos.
Clicking 'New to you' loads a full grid of thumbnails about tech, celebrities and fitness, much like what the viewer already skipped. Instead of help deciding, they get another sea of options.
TakeawayA different set of options does not help if there are just as many of them.
Paradox of choiceAccording to Barry Schwartz, the more choices people have, the more likely they are to make none. It is why people often open Netflix and end up watching trailers.
- Tiny 'New to you'
04
The problem
The only sign of the new mode is a tiny chip at the end of a row.
The page looks almost identical to the normal home feed. The only hint is a small highlighted 'New to you' chip at the far right of the filter bar, so the viewer cannot tell how to get their usual recommendations back.
TakeawayAfter a mode change, make it obvious where people are and how to return.
Discoverability and feedbackAccording to Don Norman, the first rule when moving people between screens is to reassure them they are in the right place, or show them a way back if not.
Current Too many optionsRedesign 
Back to normal view05
A better version
Show two related videos, a clear title and a way back.
The redesign titles the view 'Try something new', offers just two architecture videos, adds 'back to normal view' at the top, and a 'Watch something random' button for those who feel lucky. The topic is new but still close to what the viewer likes.
What changed
- Two options instead of a full grid
- 'back to normal view' link at the top
- New but related topic, architecture
- 'Watch something random' for a variable reward
TakeawayNarrow the choice, keep it relevant, and always leave an exit.
- Fewer views, longer
06
Why it works
The second video is similar but a bit less attractive.
Both are architecture videos, but the one on the right has fewer views and is longer, on a slightly harder topic. Comparing two similar things is quick, and the weaker one makes the other look like the obvious pick.
TakeawayComparable options are easier to choose between than very different ones.
Decoy effectAccording to Dan Ariely, choosing is much easier when we compare similar things, because the brain can judge the differences quickly. Very different options are harder to weigh.
Before 
Pre-roll adDuring 
Banner over video07
The other problem
Once a video starts, ads appear before it and over it.
A pre-roll ad plays with a sponsored card beside it, then a banner ad covers the lower part of the video while it plays. Good discovery design saves people time, but the viewing experience keeps interrupting them.
TakeawayHelping people find content is wasted if watching it feels like a fight.
Ad after the video08
The other problem
When the video ends, another ad takes over the screen.
A television ad plays full width after the video, with a matching sponsored card in the sidebar. By now the viewer has seen ads before, during and after a single video.
TakeawayEvery extra interruption spends trust the product worked to build.
Premium banner09
The result
Then the home page offers Premium: everything you love, ad free.
A large banner for YouTube Premium sits on top of the home feed. Growth.Design notes that YouTube had recently increased the number of ads sharply, and when people sense they are being pushed into paying, many resist.
TakeawayMake the paid tier attractive on its own, not by making the free one painful.
ReactanceWhen people feel their freedom is threatened, they react strongly against the pressure. If viewers work out that ads are meant to push them to Premium, it can hurt conversions rather than help.
Good start Right momentBetter path Simpler choice10
The takeaway
A smart idea that needed fewer options and a clear exit.
The viewer was delighted that YouTube noticed they were stuck, then confused by the same crowded feed with no way back, then worn down by ads. Growth.Design gave the experience a D for being humane.
What changed
- Detect disengagement in real time
- Offer a small, relevant set of options
- Show where the user is and how to go back
- Keep interruptions in check
TakeawayTimely help works only when the next screen is simpler than the one before.
Every takeaway, in one place
- The setupListen to behavior in real time, and offer help at the moment it is needed.
- ContextWhen a catalog is huge, deciding becomes the product's biggest barrier.
- The problemA different set of options does not help if there are just as many of them.
- The problemAfter a mode change, make it obvious where people are and how to return.
- A better versionNarrow the choice, keep it relevant, and always leave an exit.
- Why it worksComparable options are easier to choose between than very different ones.
- The other problemHelping people find content is wasted if watching it feels like a fight.
- The other problemEvery extra interruption spends trust the product worked to build.
- The resultMake the paid tier attractive on its own, not by making the free one painful.
- The takeawayTimely help works only when the next screen is simpler than the one before.
Source: Growth.Design: YouTube's Attempt To Solve The Paradox of Choice
Help people choose, instead of showing them everything.
Large catalogs and feeds can leave people scrolling and leaving with nothing. Zyxware can study how your users browse, find where choice turns into overload, and design focused discovery moments with clear feedback and a way back.