To add a little context, I think that there are two sides to technical performance: measurement and perception. Regarding measurement, depending on the variability of your app, you may be able to test and record response times yourself. Try it from multiple computers on multiple networks, at different times of day (ideally in scenarios close to what users might experience). This can be difficult to test with users unless you go to where they work (remote testing methods like webex can slow down the computer/internet). Once you know the times, you can map that a prioritized list of tasks to see how it affects them.
Regarding perception, Chauncey makes some great points. To test perception, I would test users in context, record the time, and watch the user closely for signs that they are frustrated with waiting. You may also want to followup with them after each task (e.g., What did you think? Did it load fast enough for you?) On Wed, Mar 25, 2009 at 9:11 AM, Chauncey Wilson <[email protected]>wrote: > Performance or system responsiveness comes up often in usability > surveys as one of the top 3 usability problems. There are a few > issues here: > > 1. Consider both performance and performance variability. If a > system has a great deal of variance in performance that lends > unpredictibility to the user experience, that might be worse than a > long, but predictable experience where you can do other tasks. > 2. There are many factors that affect perceived responsiveness. For > example, good progress feedback can influence the perception of > responsiveness. > 3. In your testing, consider realistic databases. You might have a > small testing database, but gather data on the sizes of your customers > databases. > 4. Consider that performance can be too fast. There is some research > showing that performance that is too fast or too slow will result in > more errors than a mid-range response. > 5. Direct manipulation and immediate control responses need to be > responsive or people will feel like they are working in a tub of > jello. > 6. Perceived time will be influenced by the user's perception of the > size and complexity of an operation. So, if you are loading gigabytes > of photos, you will accept fairly slow response, but the same may not > be true of operations that are judged as not so complex. > 7. If you are looking at performance of your system, consider response > time, response variability, scalability, and startup or launch time as > three major buckets to consider. > > Chauncey > > > On Wed, Mar 25, 2009 at 7:00 AM, Baruch Sachs <[email protected]> wrote: > > I am talking more on the technical performance side, keeping in mind the > UX > > ramifications. For instance, a lot of UIs might be designed well from the > > design side but perform poorly (slow loading, refreshing a whole page vs > a > > section on the page, etc) which ultimately negatively affects the UI. > > > > I have been asked to assist with the development of this methodology, > using > > the UX to drive what we should be testing for. I have my own experience, > but > > wants to get a broader sense of the topic. I sense that we do not often > have > > the opportunity to influence this area.. > > > > On Thu, Mar 19, 2009 at 9:24 PM, Chauncey Wilson < > [email protected]> > > wrote: > >> > >> Hello Baruch, > >> > >> Could you describe a bit more what you mean by performance testing of > >> the user interface? There is benchmark usability testing, but also > >> performance testing of widgets under different loads and system > >> testing of various types of responsiveness. > >> > >> Thanks, > >> Chauncey > >> > >> On Thu, Mar 19, 2009 at 6:19 PM, Baruch Sachs <[email protected]> > wrote: > >> > Lookng for any resources available on methodologies for performance > >> > testing, > >> > focusing specifically on the User Interface...preliminary searches > >> > bringing > >> > up very little. Hoping someone can assist...thanks in advance! > >> > ________________________________________________________________ > >> > Welcome to the Interaction Design Association (IxDA)! > >> > To post to this list ....... [email protected] > >> > Unsubscribe ................ http://www.ixda.org/unsubscribe > >> > List Guidelines ............ http://www.ixda.org/guidelines > >> > List Help .................. http://www.ixda.org/help > >> > > > > > > ________________________________________________________________ > Welcome to the Interaction Design Association (IxDA)! > To post to this list ....... [email protected] > Unsubscribe ................ http://www.ixda.org/unsubscribe > List Guidelines ............ http://www.ixda.org/guidelines > List Help .................. http://www.ixda.org/help > -- Paul Nuschke Principal, Research & Strategy ELECTRONIC INK© www.electronicink.com ________________________________________________________________ Welcome to the Interaction Design Association (IxDA)! To post to this list ....... [email protected] Unsubscribe ................ http://www.ixda.org/unsubscribe List Guidelines ............ http://www.ixda.org/guidelines List Help .................. http://www.ixda.org/help
