insights

How LinkedIn’s algorithm taught me how to redesign an improved Covid test

Thomas Verhoef

Sometimes the truth lies in the threads, conversation threads, that is. That is also the case on Linkedin, where content gets driven by interaction - not of just anybody - but of professionals who actively engage on specific posts that generally interest them.

In this ecosystem where professionals meet, you sometimes get the best out of your browse time when you dive into some conversation threads. This way, you can discover a gold mine of good advice, multiple analyses, links to articles, solutions, substantiated opinions, links to other interesting persons regarding the topic, and so on.

It can be, by far (for me at least), the best time spent on public forums looking for new insights. Let me explain by a recent example, which is a current topic as well: a suggestion of a newly designed covid test that aims to reduce the mental effort of comprehending if your test is positive or not.

Design of the Covid test

And as every good story begins: it all started when Andréas A. suggested a possible redesign of the by now well-known Covid test.

It is a grateful approach: something gets suggested that could make sense of the bat, in addition to the familiarity we experience when we see this test. By now, everybody knows the fairly unpleasant experience of doing the test and can also relate to the somewhat obscurity of the test itself, where you, just to make sure, grab that little paper manual that explains the result.

Without pretending to be a product design expert, I can directly relate to a few suggestions Andréas made:

  • The removal of all abbreviations (S, T, C)
  • Removing the unfortunate coincidence that the ‘C’ in this test also brings Covid and not Control to mind, so getting rid of it makes sense
  • Writing in plain language most of the time simply works better
  • Although I’m becoming a fan of QR codes, the placement on the test itself is redundant considering the context
  • The addition of an explanation of the results strip
  • ...

A good idea ≠ a great design

However, via the ‘most relevant’ filter, LinkedIn opened my eyes and made me realize and elaborate on the practice of creating a great design. Without needing to read every comment, LinkedIn directly pointed me to those that seemed to get a lot of interaction, such as the following:

This comment immediately pointed me to a bigger picture where design - as in the graphical setup and the complementary user experience - will not suffice. Instead, we need to think about the broader picture:

  • Will everybody be able to read this? readability
  • Will adding the explanation be of added value when getting printed out? manufacturing processes
  • If we go on and choose a written out description, will it work in all languages? localization

Soon it becomes evident that, indeed, a good idea does not make it a great design per se.

‘A good idea does not make it a great design’ - Every product designer ever

Serhii made a great reply there, which, in a typical design thinking fashion, leads you to investigate further how to tackle the problems posed. Other replies entail:

  • The viability of the removal of the QR code from the Covid test? Isn’t this used to inform other people more quickly? Some tests may do so; others just point you to a digital manual, which in that case, you could argue, justifies the removal (and replacement) of the QR code.
  • Could we remove any language barrier? Let’s use emojis instead? But are emojis universally known and carry the same meaning?
  • What about blind users? How do they ever discover their results?
  • ...

Another Covid test iteration

Another iteration brought the discussion to the following Covid test design:

At first, this looks like a well-thought-out solution, tackling most of the problems mentioned above.

  • ✅ The removal of all abbreviations (S, T, C)
  • ✅ Removing the unfortunate coincidence that the ‘C’ in this test also brings Covid and not Control to mind, so getting rid of it makes sense
  • ✅ Writing in plain language most of the time simply works better
  • ✅ Although I’m becoming a fan of QR codes, the placement on the test itself is redundant considering the context
  • ✅ The addition of an explanation of the results strip
  • ✅ Will everybody be able to read this?
  • ✅ Will adding the explanation be of added value when printed out?
  • ✅ If we go on and choose a written out description, will it work in all languages?
  • ✅ The viability of the removal of the QR code?
  • ✅ Could we remove any language barrier?
  • ❌ What about blind users?

Main takeaway

Almost there? Getting there, at least. And for me, even better, I got led to this solution reasonably quickly just by digging into the threads suggested by Linkedin’s algorithm. In essence, I learned a lot in a few minutes.

Key takeaway: Whether it is Linkedin, Reddit, or for that matter, any social platform where discussion is key, take some extra time to discover what certain communities think of a ‘featured’ post. You’ll quickly learn there are more than two sides to any story.

Source: https://www.linkedin.com/posts/activity-6889528518447824896-oSUk/#

more insights

Close Cookie Preference Manager
Cookie Settings
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage and assist in our marketing efforts. More info
Strictly Necessary (Always Active)
Cookies required to enable basic website functionality.
Made by Flinch 77
Oops! Something went wrong while submitting the form.