Hopefully I'm not dragging the conversation back into the mud, but I understand where both sides are coming from regarding #5. I have seen fundamental changes in products that took them from dying in the market to a a leading example in the field thanks to user observations and usability tests. I have also seen projects weighted down by nonstop focus groups and groupthink, as well as poorly run usability tests that led to inaccurate solutions.
That said, I think each of your points is valid, and addressable. I've heard each of these from clients, and I wouldn't mind running through the conversation here to make sure my thoughts are in order. 1. Most products are fairly simple and most of the testing can be done in house. - Fair enough. Though I have mostly been working with large systems and evolving ecosystems of products and services lately, these can usually be broken down into their simpler components. It seems that doing testing in house (assuming this means only having other people in the company look at it) would mean that the same biases would be with everyone. Unless you are your own client, you probably won't unveil any new revelations, aside from bug reports. This is useful, but it's more QA than UX. 2. Most usability tests are not even close to reflect any realistic version of the environment your product will end up in. - Great point. Also easily solvable. Do your usability tests in the environment your product will end up in. Go to your users and test on their systems, in their environment. Have a conversation with them. It does take a little more time, but you will get much more useful information than pulling people into a little florescent interrogation room and drilling them with questions about why they clicked -there- while you stand over their shoulder. (No offense to anyone who has really enjoyed doing this.) 3. The mistakes that you might find are not going to be those that will determine the success of your company. - I've seen very poorly run usability tests that prove this (One in particular involved a woman who repeatedly asked "This makes you feel like X, right?"). However, I think you may be looking at a different question than I am. Usability testing, for me, isn't about finding mistakes. It's about uncovering opportunities for success you may not have thought of. I start very early in the design process, when only the initial concepts are thought through, and most likely sketched out on pen and paper or modeled out of foamcore. This makes sure we are addressing the right issues with the concepts. Finding mistakes is much later in the process. 4. Many usability tests consist of max 10 people which is simply not a significantly high enough number to make any decisions based on. The single best solution is to start simple simple and make sure you can measure how people use your product. If people are having problems you will find out soon enough and you will find out where it matters. -I think this may be the biggest discrepancy. Qualitative research (User observations, Usability testing) and Quantitative research (Focus groups, QA) are very different things and serve very different purposes. Qual is not meant to have statistical significance. There is simply too much information there, and it's only done until themes start emerging. Usually 10 or so rounds and they become very obvious. This is intended to be much earlier in the strategy phase and is much deeper than Quant. You get inside people's lives and really figure out what would be useful, how what you are developing will fit into their lives. Yes, you ask questions about the product, but what they say doesn't matter as much as how they say it. At the risk of sounding trite by quoting Steve Jobs, "You can%u2019t just ask a customer what they want and try to give that to them. By the time you get it built, they%u2019ll want something new." This is where the experience of a designer/researcher really shows. They know what to listen for, how to ask questions that indirectly reveal the customer's feelings, and they can convey it later. Quant is a very different animal. This happens much later in the process for me, when everything is nearly built. Everything is on track, and you have very specific questions that you need answers to. This is when you want statistical significance. People don't typically look forward to surveys as a highlight of their day, and they will only give you the quickest and most basic answers off the top of their head. This is not a dialog or conversation, this is feedback. Doing Quant too early in the process tends to give useless feedback on something that is changing anyway. Doing Qual too late in the process tends to unearth major issues that will only frustrate everyone who has invested so much in the product. Doing Qual early on to make sure you build the right thing, then Quant later to make sure the details are right, then QA at the end to bug test has been successful for me. I have seen many examples that prove your points, and typically they really only show that the wrong kind of testing is being done, or it is being done poorly. I have lost track of the number of companies who, when I asked if they have done user interviews, replied "Of course, we do focus groups all the time. Want to see the charts?" At which point I smile and start a conversation about what they have done so far, and the different methods we are going to use, as well as what kinds of results they should get. So far, people seem to be happy when we uncover new things, or put their findings in a new light. Chris Dame http://theusabilityofthings.com . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Posted from the new ixda.org http://www.ixda.org/discuss?post=45640 ________________________________________________________________ Welcome to the Interaction Design Association (IxDA)! To post to this list ....... disc...@ixda.org Unsubscribe ................ http://www.ixda.org/unsubscribe List Guidelines ............ http://www.ixda.org/guidelines List Help .................. http://www.ixda.org/help