4 ms·
Let's all remember that UX (& IxD!) should be based on research and testing, not on assumed patterns. The statement that "Make your website more scannable by u
by kagamine 11y ago
Let's all remember that UX (& IxD!) should be based on research and testing, not on assumed patterns.
The statement that "Make your website more scannable by using less text and adding more imagery (as humans are 90% visual creatures)" sounds as much like dumbing-down as anything I have ever heard, resulting in useless pages that I don't need. In fact at http://uxmyths.com/ http://uxmyths.com/ myths 4, 7 & 8 counter this idea of more visual being better. The mantra I like more is "content is king". Think man pages. Although a specific use case, man pages benefit from concise, precise information, and pictures would be a detriment. Same goes for most news articles.
Similarly with error messages, the article uses specific errors to make a case and then uses Skype's login to illustrate the "right way". These are two very different use-case shapes being forced into a square hole. As many of us here should know, there is nothing worse than a fuzzy, warm error message that can't be searched online or in documentation. Sure, errors should not be "error: 0x15654213!" but should include enough information for the user to solve the problem (and please do not use exclamation marks in errors!). Often, some technical info, even if just an error code and a short description are enough to resolve the issue. If you can't log in, as in the Skype example, well, it's one of two things 99% of the time. You probably don't need to google that.
I like the rest of what the author has to say, especially that part about developers getting to do it over and over while UX is expected to be right first time (assuming the place you work actually does UX, mine doesn't and the results vary greatly). My workplace doesn't test either which I think is a huge mistake (so much so I almost couldn't write this sentence without swearing). That's something I am trying to change.
- paulojreis 11y ago> My workplace doesn't test either which I think is a huge mistake (so much so I almost couldn't write this sentence without swearing). That's something I am trying to change. This is the biggest problem, I think - the conception of UX without "user" work. This doesn't mean only testing per se, but any kind of empirical or data collection work (from using anthropological methods early to strict usability testing). If a given company is not doing any of this, then they're not doing UX at all. If they're not in contact with the user in any way, then it's not UX. Having someone creating UIs in Photoshop is not UX and this is, I think, the biggest and most detrimental misconception of the field.
- ArekDymalski 11y ago>man pages benefit from concise, precise information, and pictures would be a detriment. Same goes for most news articles Assuming specific audience of course. In case of news it's possible to find readers craving for big pictures. Or pictures only.