Main image
19th November
2012
written by Chris

Power users are important, or they ought to be.  Power users are the folks who can help you figure out what your product is really capable of doing; they encourage you to push the boundaries of your existing feature set and can really help to highlight aspects of your service that you didn’t know were important.

No one seems to have told this to the folks at Netflix.

Now, don’t get me wrong, Netflix streaming is an amazing product.  A huge library of TV shows and movies delivered over the internet with just a few mouse clicks still feels like something from the future, but apparently it’s a future in which user interface design exists to confound people like me.

I, you see, want knobs and twiddly bits.

I get that simple is better — that you don’t want to overwhelm your users with too many choices or too many ways to accomplish the same thing but I can’t shake the feeling that Netflix’s interface is overly simplified and, indeed, designed around user interface paradigms that just aren’t true for the majority of my use.  Recommendations are great, lists of shows by category are great, but Netflix seems to have decided that it serves exactly two kinds of users and I fall squarely outside those two camps 75% of the time that I use the service.

On my phone, Netflix’s mobile browsing experience is minimalistic and elegant.  On my Roku — a set-top box — Netflix’s browsing experience is a little slow, but otherwise fine.  The remote limits what it is that I can do and I am grateful for an interface that accommodates its use.

But on a desktop or laptop PC, Netflix’s user experience is terrible.  Despite having a browser that is happy to display arbitrarily long documents with automagical vertical scrolling, Netflix has opted for some sort of horizontal scrolling nightmare.  Despite having a mouse with which to drag scroll bars so I can read at my own pace, Netflix has chosen to fix my scroll speed for me.  Despite having the ability to rapidly transition between user interface elements to mold search queries, Netflix has chosen to dramatically limit the complexity of any search I might perform.

I might wish to see, for example, a list of every “Heist” style movie ever made.  I should be able to display this on a PC with no issues but, instead, am lucky to get a list of 20.  I might wish to further narrow this list to movies which are also comedies – but Netflix offers no meaningful tag combination search save that which is predefined by the interface.  I might further wish to narrow this to films starring Robert de Niro or made after 1972 but there’s no meaningful metadata filtering, much less in combination with tags.

Yet I am sure – sure beyond even the tiniest shred of a doubt – that Netflix has all of this data and stores it in an epic relational database.

I recognize that I am a power-user — most Netflix users don’t want the same granularity in search that I do — but I also recognize that some of the interface complaints I have must also frustrate less technically inclined users of the service.  More generally, however, Netflix’s failings in this regard offer a lesson for web application developers moving forward.

When I first entered the software development workforce the web was a fairly uniform medium; people came to the websites I build using computers with keyboards and mice.  Today, however, the web is becoming ubiquitous.  Users may interface with a website via a traditional desktop system, a laptop, a tablet, a phone, a set-top-box, or even an application which wraps and presents website functionality.

Building user interfaces for such a diverse ecosystem requires that we ensure that our design decisions never rob Peter to pay Paul — in other words, that we never sacrifice ease of use on one platform in order to improve it on another.

Leave a Reply