Book review: Laws of UX (2nd ed.)

★★★☆☆ I was delighted by the Laws of UX website when it came out. The site was beautiful (it still is!) and it felt important to bring some of this stuff to designers earlier in their careers. But the book based on the website was largely a disappointment, and seems like a good case study of an unsuccessful adaptation – it felt this was pushed to become a book without editorial help and without thinking too much about what makes for a good book. The book lost a lot of what made the site great – it’s a pretty generic-looking production with flawed typesetting, an uninspired cover, and poorly sized and reproduced images. But chiefly, I also feel the book showed there is limited rigor behind the whole premise; the writing feels academic in the sense that it’s a little boring, but academic writing at least can be precise and follow process. Not here. The laws of UX are not “laws” in the traditional sense and the combination of “laws” presented, as well as examples of them in use, feel really arbitrary and sometimes at odds with the entire premise. I felt disagreeing with the book often. For example, I feel the chapter about Doherty Threshold feels is teaching the wrong lessons (100ms is not enough for a bunch of things!). Or the advice on gradually deploying changes (Jakob’s Law) is missing a core component of maintenance and how to approach the contingent of users who will never graduate to the new interface if given a chance to stick with the old one. I also started worrying that the book doesn’t fully understand how design works. From the very first page:
This project was somewhat unique in one specific way: I was being asked to justify a number of design decisions to project stakeholders, without any data to support them. Normally, when you have quantitative or qualitative data available to draw upon, this is pretty straightforward task – but in this case the data wasn’t available, so the process of justifying the decisions would have to be a little different.
This is… This is not what design is. This is never true. You rarely have the data – and if there’s data, it’s never netural, always at the mercy of people collecting it and people interpreting it. My friend summarized it well – “design is not mathematical” – but at various moments the book suggests it’s as simple as knowing a certain “law” and applying it. This is perhaps most visible in the Aesthetic-Usability Effect chapter, which touches upon craft without really understanding it. On the positive side: I think what the book is trying to do feels important and appreciated. Some of the stuff like Fitts’s Law or Doherty Threshold and Jakob’s Law are good to know about, they are still relevant, and can serve as useful tools in your toolkit as a designer. I also learned some new things from it! I have never heard of shape coding before (turns out I’ve been practicing it without knowing, so learning about it was validating), and never really thought about the equivalent of heat maps for mobile. Also: I don’t think this book is for me. I get a sense this is a volume for a very different group of UX designers, maybe even people at companies where UX design is not at all established as practice. There is a lot of stuff like explaining personas and basics of user testing and even ethics that feels somewhat out of place and like it’s padding the content – but I can see how that could be valuable. However, I still wish the book didn’t oversimplify a lot of things like I think it does. I believe there’s a way to do it while still keeping it accessible and not overwhelming. But today, I would rather recommend the beautiful poster that seems more true to what the website was trying to aim for. In terms of how I would improve the book:
  • Have it reviewed by someone who actually lives and breathes this stuff for living.
  • Invest in better writing and better storytelling. This of good stories and not just data. Ditch the random O’Reilly-esque callouts and integrate them into the stories.
  • Either get deeper into more specific and deeper examples for most of the stuff, or make it drastically shorter.
  • Don’t package it all as “laws”, or at least – if this title sells – contextualize it better inside. These are useful tools, but they are not laws like physics has laws. Also, all of them, like most of design, will be caveated with “it depends.”
  • Consider adding stuff about motor memory, Sturgeon’s Law, monotony, gestalt to flesh out the toolkit, and maybe group the chapters into a few bigger areas.