Why Engineers Can’t do Your UX Work


April 17, 2017 by SDesign

There are a lot of times in my career when I have been asked about ways to continue UX projects without the use of UX resources.  Generally what is requested is a style guide that would allow the engineering team to continue building off of work completed by me or my team in an initial engagement.  Unfortunately, this will just not work.

The basic premise behind a request like this is that UX Design is a recipe: all you have to do is have a list of the ingredients, and then the end result can be replicated.  I absolutely understand why project stakeholders think this way: they want to contain costs.  Unfortunately, this type of thinking is wishful thinking, and it does not represent a true path to getting a usable end result.

In UX we may have contributed unwittingly to this misunderstanding by providing lists of heuristics and style guides.  Apple provides detailed HIGs (Human Interface Guides) for iOS, Google has a great deal of documentation on Material Design, and Microsoft has also followed suit with its Universal Windows Platform.  Adding to that, type in ‘usability heuristics’ in your search bar and you will immediately run into Jakob Nielsen’s 10 heuristics.  I think his list is great; but I think he has unintentionally done us a disservice by making design seem so easy.  After all, how hard can design be it be if all you have to do is follow 10 rules?

My immediate counterargument is to point to a foundational text in the HCI literature, a set of comprehensive guidelines written by Smith and Mosier.  These guidelines contain hundreds, if not over a thousand of design heuristics.  Moreover, since it was written in 1986, it does not take into account the rich graphical capabilities that we have today, as well as new senses such as haptics (the vibe on your phone) and rich sound.  The US Department of Health and Human Services has also produced an excellent and robust research-based set of guidelines which is more current; Jakob Nielsen himself now references 2,397 guidelines (!?).

So, the first point is that UX Design is not about following a small number of easy-to-digest heuristics that can be contained within one source.  The second point is even more crucial: Design is not fundamentally about memorizing a set of rules, it is about arbitrating between conflicts within these rules. The rules all intersect in complicated ways, requiring the designer to trade-off different guidelines that often are in opposition to each other. (For more detail, see this youtube video I recorded.)  A simple example is the interaction between two primary guidelines: 1) Keep lists short, and 2) provide multiple exits.  If it is important to for the user to be able to  exit your product, then it would make sense that you provide an opportunity in multiple places. However, if you are constantly providing the capability to exit, then all of your interaction points will be cluttered with the same exit opportunity.

There is no simple resolution to a conflict such as this.  Only expert human judgment can resolve it.  If you are working on a product where exiting is both important and difficult, then you might tend to over-represent the exit function. On the other hand if you are working on a website where all users have access via the browser to the Back button and to the window close (‘X’) function, then you might even choose to totally ignore the Exit heuristic. There are also heuristics that caution against redundant functions.  How do you square that rule with our multiple exit mandate? 

Only practiced judgment and trade-off analysis can resolve these inherent conflicts.  UX Designers spend all day every day studying and analyzing design patterns. They may make errors and get feedback, whether it comes via usability testing, customer feedback, or a more senior designer.  Over time the designer builds up the expertise to make these difficult trade-offs. The vast majority of software engineers do not have the expertise, time, or the appetite to conduct the analysis, weigh the options, and make a careful selection.

A closely related point is that design is….well, Design! It is a creative endeavor.  Would you ask an engineer to be your architect, make your a sculpture, or take a high-quality portrait?  The reason that good design is so sought after is that it is non-obvious.  Sometimes the answer to the vexing guideline contradictions is a clever new idea.  My mentor, Tony Tsoi, used to say that he would be stuck on a problem for days and then suddenly the answer would come to him in a dream or in the shower. (A poignant example of the well-known incubation period in the creative process).

Let me close by summarizing my arguments.  For cost reasons, it is not uncommon for management or stakeholders to want to produce a recipe book to produce a user interface, thus saving them the cost of a UX designer on staff.  This does not work. UX Design is not a recipe: it is a practice cultivated by dedicated professionals who are experts in their field.  For one thing, doing good design work requires understanding many guidelines and minutiae which are not easily represented by well-meaning but ultimately glib heuristics. Contained within these minutiae are multiple conflicts and ambiguities which require expert human judgement to resolve.  And finally, good design solutions are creative in nature; guidelines and libraries are inherently mechanistic, and will never produce novel ideas on their own.

1 comment »

  1. Bablofil says:

    Thanks, great article.

Leave a Reply

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