Halftank Fuel: How object audits spark innovations and a YouTube redo đź“ş
Published 5 days ago • 2 min read
If you have worked on any redesign project in any capacity, you have probably seen some kind of audit done on the current state of a system. Usually, these audits focus on components as a tool to promote design consistency.
I wrote recently about the benefits of rethinking the auditing process to focus on auditing for objects in place of components. By promoting consistent visuals for a system’s objects, that system becomes much more intuitive and easier for users to learn. It becomes easier to learn because by thinking about how different objects appear in your system, giving your objects their own visuals better mirrors how we learn systems in the real world.
So instead of getting into the tactical weeds of how certain buttons look with a typical design audit, you are taking a more strategic view of your system and focusing on usability by looking at your objects with an auditor’s lens.
Audits are a powerful tool, but I wanted to go deeper into how I define objects in an existing system. For a thought experiment, I wanted to take a look at YouTube.
I find YouTube’s interface overwhelming. I believe that this is because the video thumbnail cards that you see on the page have multiple minor visual treatments depending on what type of video it is. For example:
Most videos have basic metadata at the bottom: titles, video duration times, creator information, etc.
Live videos have “live” tags and some small visuals to indicate that it is a live event
Movies have a “free with ads” or “rent or buy” tag
If you have saved a video to a playlist or “watch later”, there is no apparent indication on the video card that you have saved it - you have to go to the playlist or try to find it somewhere on your home page
In other words, different types of objects are mixed within the interface, which gets confusing quickly, as those object types and categorizations are difficult to distinguish at a glance.
I’ve been thinking of ways to organize this interface differently using an existing mental model. One mental model I considered was channel guides on most TV interfaces. These interfaces can also get overwhelming, but they have a clear, timeline-based organization that is easy to understand. Is there a way to incorporate a time-based navigation into YouTube?
With this in mind, I took a YouTube template from Figma’s community (thanks to @berkaymsert) and did a quick and dirty reorganizing in a wireframe:
I put the columns in rough timing order from shortest to longest and I have the rows as “channels” which could cover interests, subscriptions, etc. If this were my project, I would do some more research into different active states for the channels and if changing the sizes of the video cards would help with understanding, since those rows are very thick.
I wanted to show with this exercise that considering a system’s objects can give you opportunities to try new things and innovate. By taking a step back to look for what’s important in a system, you can’t help but think of different ways to reference existing mental models and make the most important objects in your existing system more prominent.