Kyle Soucy, of Usable Interface, is presenting this talk, again, a subject near and dear to my heart.
She started about talking about caring. If you don't care about the people, you can't build a usable product. When people interact with a product, they're interacting with you. Everything we do creates an experience for the end user.
Why care about usability? It's survival of the easiest. If you're not creating usable products, your competitors are. Sometimes you don't have the opportunity for a relaunch.
How to do it? User-centered design. Geting users involved in the design.
HOw do you do it? User research, focus groups, log file analysis, surveys, eye tracking, card sorting, usability testing, but not heuristic evaluations. The latter is a usability person's opinion, not a user opinion. This session will focus on user research, card sorting, and usability testing.
User research is NOT market research. Market research is quantitative. You need qualitative. Qualitative research is direct observation. The researcher can ask follow-up questions, probe on behavior. Really want to know why users do particular things. Focus groups just give you opinions, what people say they do often isn't what they really do.
Never use stock photography for personas. Make it real. Find a candid photo. Designers and stakeholders can't relate to stock photos.
The usability testing process starts with defining what you are testing and who you are testing. If people are to buy in to your test finding, they have to buy in to the people you tested. You can't argue with the data if they are real users. Define the tasks you are testing. Recruiting users is next, and the hardest part of usability testing. Then you prepare prototypes and conduct a dry run. If you don't. your first test is a dry run. Finally, you facilitate the test and run the analysis of the data.
Formal usability labs are not needed, and they can be a barrier to testing. You can just use a conference room and have people watch with screen sharing software.
Information architecture is organizing, labeling, and classifying content so it can be easily found. To create an effective IA, you have to understand business objective, content, user objectives, and content of use. IA methods include card sorting, free listing, and mental models.
For card sorting, put all your content on index cards, labeling content, give them all to a user and ask them to categorize them that make sense to them, then label the piles. Do this with enough people, you start to see trends on how people categorize information.