In my last article I wrote about what you would need for your usability test. You can see that information here in part 1. Today, I’m going to write about the process of the usability test. How do you perform a usability test?
What Do You Do?
To perform the test set a time and location for the test. Find the 3-4 test participants and schedule them about an hour a part. It’s good to have a little time between test sessions to reset the test and discuss the previous session. Set up two rooms. One room will be for the test. In that room you will setup the camcorder (if you have one) and a computer with a webcam. In the second room you will have the observers of the test with their computer(s).
You will then need to setup a means of transmitting the test from the test room to the observation room. For that I recommend using livestream.com with Procaster. It’s free and can transmit both the webcam feed and broadcast the screen of test participant’s computer allowing the observers to view both the participant (through the webcam) and what they do on the website. Procastor can also record the session so people can review the tests again later. If you do not use livestream.com or something similar, then you will need to run a line from the camcorder in the test room to a TV in the observation room so the observers can view the test that way.
Once you have everything setup, you can start the tests. For the test you want to allow the test participant to view the various pages of the website that you are testing and give their impressions about the pages. You may have specific tasks you want to test that you can ask them to perform.
Here are some tips for the test:
- Start each session by introducing yourself and getting to know the test participant a little. This will help make the test more comfortable.
- Emphasize to the participant that you are not testing them, but rather you are testing the website. So, they can’t mess up. If fact, if they struggle with something, it will show you things that need to be improved, which is a good thing.
- Also, let the participant know that it’s important that they be very honest and reassure them that they don’t need to worry about hurting anyone’s feelings.
- As the volunteer goes through the test, encourage them to think out loud. They should say what they are thinking and why they are doing what they are doing.
- Be willing to answer some questions, but you also must keep the integrity of the test. If a test participant asks a question related to what you are testing, you may need to politely tell them that you aren’t able to answer that question right now, but if they still have the question at the end of the test, you will answer it then.
- Allow the participant to “make mistakes” and figure things out themselves. Remember, you want to see what it would be like if they were visiting your site while alone at home without you there to help them. At the same time, don’t let them get too stuck and get frustrated. If you sense they are really stuck, help them through that point so you can continue the test.
- If you have tasks for the participant to perform, as much as possible let the participant determine the specifics of the task so they have more personally invested in performing the task. For example, if the task is to find and buy a product. Ask the person what they might want to buy and have them find and buy that product.
Once the test session is completed, if you have time before the next session, have the test giver and the observers discuss the session briefly. The observers may be able to give advice to the test giver to help the following tests.
Debriefing After The Tests:
It is important for the staff members to discuss the tests immediately after you concluded the final test session. The tests are still fresh in your minds and you should be able to quickly identify key areas that need attention. If you do the tests in the morning, perhaps you can do the debriefing over lunch. The point of the debriefing three-fold:
- Identify any low hanging fruit
- Identify 2-4 big things that need to be fixed
- Determine action steps that need to be done.
Low Hanging Fruit
Low hanging fruit are things that are quick and easy to fix. This may be changing the wording for a link, changing the picture on a page, or removing some text. These are things that can be addressed without a significant investment of time and effort, so you might as well go ahead and fix them.
2-4 Big Things
These are things that need to be changed but are going to take some planning and a significant amount of work to correct. They may be things like redesigning the navigation structure of the site, redesigning the template of the site, or changing the entire content of a page.
Determine Action Steps
Meetings are generally useless if people don’t leave with action steps. So, don’t just end the debriefing have identified what needs to be done. Determine the action steps that need to happen to get those things done. Assign those tasks to people and set deadlines. This ensures that the issues you determine need to be addressed will be fixed in a timely manor.
Some Things To Consider When Debriefing
While debriefing there are some things to keep in mind:
- Don’t throw out the baby with the bath water. Most issues on websites don’t require you to completely through out what you’ve done and start over. Look for ways to fix the specific issues within the framework of what you already have.
- Don’t break something to fix something else. During the test note what works well as well as what isn’t working. Make sure you don’t change the things that are working well when trying to fix the things that aren’t working well.
- Know the difference between a preference and a usability issue. “I don’t like the color orange for the background” is a preference. “I don’t like the color orange for the background because it makes it so I can’t read the red text” is a usability issue. Don’t get caught up in people’s preferences or you’ll be changing everything on your website every time you do a test. Everyone has their own preferences. Focus on usability issues, things that prevent people from using the site they way it’s intended.
- One person does not make a majority. If you find that one person had an issue with something on the site, but the other 2 or 3 people did not, that may not been an issue you need to address. If there is a quick fix that will address the issue while not breaking what was working for the other 2 or 3 people, then feel free to fix it, but otherwise, focus on the issues that affected a majority of the people.
- You don’t need to resolve every issue in one test. Find 2-4 major issues that need to be resolved and address them. You can always do another test later.
Hopefully, the information from these past couple of articles has helped to demystify usability testing and show that any company or organization can and should perform usability testing. Another great resource about usability testing is the book, Don’t Make Me Think by Steve Krug. He also has a website, sensible.com that has additional information about usability testing as well as info about where to get the book. I highly recommend Don’t Make Me Think both as a resource for usability testing as well as practical information about website design.
Usability testing is a fantastic resource for any web designer (or designer of anything else). Any organization can do it. So, what are you waiting for? I’d love to hear back from a bunch of you over the next couple months about your usability tests and they affected your websites.