On Numbers, Micro Design Decisions and Habit Loops

Since I’ve started posting here on Medium I’ve noticed something interesting about the way we choose to display larger numbers in UI design:

1k vs 1,000

Admittedly, I’ve never gotten enough response on any other platform to come into contact with numbers that large. But, I’ve had an opportunity on Medium to see how a seemingly simple micro design decision like how to display numbers can impact user engagement.

These numbers most often show up in what I like to think of holistically as “stat counts”. These are likes, followers, hearts, views, reads, stars, pins, repins, retweets, recommends etc etc. Playing off the human need for social validation, these simple game mechanics are brilliantly designed tools that drive some pretty strong habit loops in social applications.

If you aren’t familiar with the concept of a habit loop, InVision has a nice article series that lays out the details. There is also a good book, referenced below. But basically a habit loop consist of three parts: a cue, a routine and a reward. The cue triggers the routine, and once routine is completed it is reinforced by a reward.

In the example of social platforms, a cue happens (you find a great article to share, or your child/pet does something adorable), this triggers the routine of posting to your preferred social app, and then you are rewarded with likes.

These stat counts are one of the most successful examples of rewards in the design of digital habit loops. We all want to see our stats go up. Every post has the ability to trigger an obsessive loop of checking and rechecking those counts. And every time we check, we are more and more likely to post again. In this case, the act of checking the reward can itself become a cue to trigger the routine of posting.

In the world of the attention web, where many business models are predicated on the simple act of users opening and viewing content, this kind of habit looping can be the holy grail for success.

But, the way we choose to display those stats can impact the strength of the habit loop.

Let’s take Medium stats as an example, since that’s where I made my observation. Here is a shot of the stats for one of my stories that got some traction.

The Medium Staff account recommended this one and it reached the top 5 on the site for a day or two. It was my first post that really took off, so, as you can imagine, I was excited.

At first, I found myself checking Medium literally every few minutes to see what was happening. Almost every time I checked, the stats had gone up. 100 views. 110. 140. 145. 190. 262. So I checked more and more often. I was spending the majority of my time looking at the app. But then something happened. I hit 1,000 views. Then 1,000 reads. Now when I checked back all I saw was “1k”. It took 100 views or reads for the numbers to change. 1.1k. The feedback wasn’t immediate anymore. I found myself checking less. My loop slowed down. When I hit 10k, I was barely checking. (At the time the display for 5 digit numbers was only 10k, 11k etc. It only incremented after 1,000 additional views. Medium has since updated it to be xx.xK or 10.1k.)

The choice to truncate large numbers in the user stats caused my engagement with the product to drop significantly at a time when I was highly primed to engage.

As a partial solution Medium does provide a hover tooltip to get the full number.

I call this a partial solution because for most of the day I was checking on my phone. No hover on mobile.

A similar issue shows up with Follower counts as well.

This is not an issue that is unique to Medium, lots of services truncate their numbers. There are a lot of reasons to make that decision. It could be due to layout constraints, or an aesthetic choice to maintain a clean UI. But, as YouTube shows, it is possible to show your numbers in all their glory — even when it’s nearly 2.5 billion.

Now, despite this issue I’m still posting on Medium regularly and I still have fun watching my stats go up. But for me it’s a good reminder that even micro design decisions, like whether or not to truncate a number, have the potential to have a big impact on how users engage with your product.