4 ms·

There needs to be a Big(O) table for UX. Number of user interactions and decisions to get to each functionality in each screen size. Then comes prioritization.

by dzink 4y ago

There needs to be a Big(O) table for UX. Number of user interactions and decisions to get to each functionality in each screen size. Then comes prioritization. You can only fit so many buttons and boxes close to the hottest zones of user interaction, especially on Mobile, and even less of them in advertising-powered products. When someone prioritizes the wrong use case on premium real estate and pushes back buttons that matter to you more, that only gets worse over time - your needs are not their priority. That is why on mobile at least the giants are not always giant enough to squish competition entirely - one can always provide a better UX than Facebook or Google with a small but focused app that works for your problem better. The real damage to competition comes from device apis that put UX overhead to use cases you serve bette than them, but then give their apps with the same use case UX priority in their next release. Just like data and code, complexity increases for UX with each added user story and use case, but the screen space remains the same.

wruza 4y ago
https://en.wikipedia.org/wiki/GOMS https://en.wikipedia.org/wiki/GOMS
wildrhythms 4y ago
I work in UX and I can tell you the people managing these design teams have never even heard the term 'Big-O' much less understand it; in my sphere I find they are mostly interested in boosting engagement metrics on whatever new feature will get them sooner promoted before they exit to the next company and do the same thing. I love this idea, but my outlook is grim.
- Spivak 4y ago
  Create maligned reward system, get maligned work. It's not like our field doesn't have resume driven development, and honestly why would you do anything else? Doing good work for your employer only pays if it literally pays. If your employer and the market rewards launching features, products, and making metrics go up then why toil away needlessly making money for someone else?
specialist 4y ago
I haven't done user interface design, in anger, for 20+ years. Before the kids renamed it "UX" (old wine, new bottles). Anywho. We used to consider Fitts' Law, Hick's Law, and so forth. Celebrity UI designers (ahem) like Bruce "Tog" Tognazzini and Jakob Nielsen would belabor these seemingly obvious design considerations ad nauseum. Your Big-O suggestion could be a nice heuristic for scoring and ranking design alternatives. Cool. Not that design intent ever mattered. The age old tale remains the same. Grind, iterate, validate. (Does anyone do usability testing any more?) Voilá! Then some PHB doing drive-by mgmt decrees "Those buttons should be cornflower blue. I like the old font better. Just change it all back." I eventually rage quit UI work. Preferring to have my good taste, experience, skill, and efforts denigrated in other domains. https://en.wikipedia.org/wiki/Fitts%27s_law https://en.wikipedia.org/wiki/Fitts%27s_law https://en.wikipedia.org/wiki/Hick%27s_law https://en.wikipedia.org/wiki/Hick%27s_law It now occurs to me the audience for all that UI design advice was our bosses, not us practitioners. For appeals to authority. For populating our bookshelves, to exude the facsimile of learnedness.
- marcosdumay 4y ago
  > Does anyone do usability testing any more? My impression is that it's all quantitative post-fact A/B testing nowadays. What is interesting, because it was widely known that quantitative usability research was mostly waste and you were much better doing 10 times the amount of it with only qualitative results. What was not widely discussed¹ is that post-fact testing is also almost useless. It can only tell you what solution is better, but the real gain comes from discovering what problems exist. 1 - My guess it's because it is too obvious.

by dzink 4y ago

wruza 4y ago

https://en.wikipedia.org/wiki/GOMS https://en.wikipedia.org/wiki/GOMS

wildrhythms 4y ago

I work in UX and I can tell you the people managing these design teams have never even heard the term 'Big-O' much less understand it; in my sphere I find they are mostly interested in boosting engagement metrics on whatever new feature will get them sooner promoted before they exit to the next company and do the same thing. I love this idea, but my outlook is grim.

Spivak 4y ago
Create maligned reward system, get maligned work. It's not like our field doesn't have resume driven development, and honestly why would you do anything else? Doing good work for your employer only pays if it literally pays. If your employer and the market rewards launching features, products, and making metrics go up then why toil away needlessly making money for someone else?

specialist 4y ago

I haven't done user interface design, in anger, for 20+ years. Before the kids renamed it "UX" (old wine, new bottles). Anywho. We used to consider Fitts' Law, Hick's Law, and so forth. Celebrity UI designers (ahem) like Bruce "Tog" Tognazzini and Jakob Nielsen would belabor these seemingly obvious design considerations ad nauseum. Your Big-O suggestion could be a nice heuristic for scoring and ranking design alternatives. Cool. Not that design intent ever mattered. The age old tale remains the same. Grind, iterate, validate. (Does anyone do usability testing any more?) Voilá! Then some PHB doing drive-by mgmt decrees "Those buttons should be cornflower blue. I like the old font better. Just change it all back." I eventually rage quit UI work. Preferring to have my good taste, experience, skill, and efforts denigrated in other domains. https://en.wikipedia.org/wiki/Fitts%27s_law https://en.wikipedia.org/wiki/Fitts%27s_law https://en.wikipedia.org/wiki/Hick%27s_law https://en.wikipedia.org/wiki/Hick%27s_law It now occurs to me the audience for all that UI design advice was our bosses, not us practitioners. For appeals to authority. For populating our bookshelves, to exude the facsimile of learnedness.

marcosdumay 4y ago
> Does anyone do usability testing any more? My impression is that it's all quantitative post-fact A/B testing nowadays. What is interesting, because it was widely known that quantitative usability research was mostly waste and you were much better doing 10 times the amount of it with only qualitative results. What was not widely discussed¹ is that post-fact testing is also almost useless. It can only tell you what solution is better, but the real gain comes from discovering what problems exist. 1 - My guess it's because it is too obvious.