Sometimes Breaking the Rules Is Good

I encountered an example recently where a UI Golden Rule was broken to great effect.  I promise that the idea of “breaking rules” has nothing to do with my last post that was all about not knowing what the rules are.

So back to the UI Golden Rules – there is no shortage of rules out there that have been concocted and created through a combination of common sense, user studies, and solid media persuasion.  I say persuasion because there are no real “UX Golden Rules” that everyone agrees on – design is too subjective for that.  I would argue that there are Golden Rules around best practices for gathering information and analysing it, but when it comes to actual design, the rules tend to be more about who has the loudest and most-listened-to voice at the moment.  What’s this all mean?  It means most designers tend to form their own Golden Rules and work from those.  Myself I tend to model my thinking along the lines of Ben Shneiderman‘s 8 Golden Rules.  But this post is about breaking rules for the betterment of design – let me tell you a story.

I was recently booking a rental car for travel.  I checked the Googles for local rental car shops and visited their websites to check out prices.  Now, White Rock is not a terribly large city so I only ended up looking at two websites.  During the research into car availability and pricing I encountered an example where one site followed the Golden Rule of Consistency, and the other one broke it.  And frankly the site that broke it was easier to use.

Go ahead, see which you find easier – try to specify a pickup time of 12 noon (don’t get me started on if noon is 12am or 12pm, because both are wrong)

Site 1

Screenshot of site1's time picker

Site 1 focuses on using consistency by ensuring that all the time entries are uniform in appearance and format.  This is to allow quick scanning since you can mentally ignore some parts of the data, and jump around from top to bottom without worrying that the data is ordered differently.  The challenge with this design is that it is not easy to pick out any particular time – like noon.  I performed a kind of binary search by starting at the top time, skipping 3/4 of the way down and reading that time, then skipping half way back up, and then reading carefully to find the 12:00 PM entry.  This was a little slow because I had to scan over 12:30 and 11:30 to find 12:00.  Doing so allowed to me confirm that this was noon, and not midnight, since the previous timeslot is 11:30AM – but this might just be me with my overly-sensitive take on if noon is AM, PM, or neither.

So this dropdown used consistency to allow me to skip around to find a particular time, but once I started to narrow in on the particular data the consistency actually became an impedance as I then was forced to read each entry carefully to find the exact time I was looking for.

Site 2

Screenshot of site2's time picker

Site 2 uses the same dropdown with one important difference: it spells out “noon” instead of 12:00.  This lack of consistency makes noon really stick out when scanning the times.  This makes it almost instant to find noon.  In fact it also would help if I were looking for other times as “noon” becomes a quick divider of the data – either the morning or the afternoon.  In this case then the consistency is broken but with greater usability being the result.

One could argue that this is an artefact of my specific use case – looking for the noon time.  That’s very true.  But it appears that this might be a fairly common use case, as one site has broken the rule of consistency to spell it out, making it quick to identify.  So the site has possibly optimized this dropdown for people finding noon.  I can’t imagine Budget customers pick up at noon more often than Avis customers (or vice versa), so without knowing more I suspect that my use case is a fairly common one, which means optimizing for it makes sense.

There is no single set of design rules that everyone should follow in UI design.  Site 1 used consistency to some success, but it served me less well when I started to actually read the data.  Site 2 mildly broke the rule for greater usability.  I still believe that having these basic tenets or Golden Rules in my design toolbelt results in greater usability for the designs I make.  However they are just tools, and ones that can be left in the toolbelt to great success.  The ultimate Golden Rule remains “test, test, test with real users, users, users.”

Leave a Comment

Your email address will not be published. Required fields are marked *