A while ago we had Brian, one of our user interface designers, write about the importance of user testing. He works very closely with people we invite to our user testing sessions, so he knows what he’s talking about. Here he is with a follow-up, this time with tips on preparing and conducting your own user tests – Rochelle
In my last post I mentioned how important it is to look at how “user-friendly” your website is. If it’s user-friendly, it’s far more likely that your visitors will come back and make visits to your site a regular thing.
User-friendliness is one of the most intangible and hard to define characteristics of any website. It can be really hard for someone who has built a site to make that determination alone. The very best way to find out how friendly your site is to your visitors is to ask the visitors themselves.
At Homestead, we bring in a group of users every month and just watch them use our website. It’s been invaluable to see where some of the design choices we make work out well, and where some fail. It gives us an opportunity to fix things where necessary, and make Homestead a better product to use. Several changes we’ve made over the past year have come about from observations made in user testing, such as the redesign of SiteBuilder Lite we did a few months ago.
Most software companies test their websites with real or representative users. Some companies, such as Google and Microsoft, have entire teams of people and advanced fancy equipment to test. However, you don’t need to be a large company with vast resources to test the usability your website. All you need is time and a little help from your friends.
Mapping out tasks
Last time I talked about deciding what your goals were for visitors coming to your site, like making a purchase, submitting a survey, finding your location etc. Once you’ve identified those, put yourself in the shoes of your visitors and map out the tasks you think they should do in order to accomplish each of those goals. Each task should be something a user on your site would typically do to accomplish one of your goals.
For instance, if one of your goals is for a visitor to make a purchase, the tasks they need to accomplish might be searching for a product, selecting a size, choosing shipping options and inputting payment information.
Some tasks should take less than a minute to finish, others will take more. It all depends on how complicated the task is. Come up with an ideal length of time for each task. For a simple five page information-based website, any task should not take more than a minute or two. For an ecommerce website with many products, tasks may take several minutes.
During your test you’ll have an opportunity to see how long it actually takes users to accomplish your goals. Don’t be discouraged if each task takes longer than you expected. Either you misjudged, or you just identified an area of your site that needs improvement (more on that in another post).
Who to test
Now that you’ve identified your tasks, you need to identify who your target users are. If you’ve made it this far you probably have a good sense of who they are. If you sell products, such as pet supplies, a target user might be dog owners. If you’re an electrician, a target user might be a homeowner.
To run a proper test you’ll need to observe at least five people using your site. Ideally these five people should be representative of your target user group. If you can’t find representative users, use family and/or friends.
However, make sure that that their technical proficiency is roughly the same as your target group. If you believe your target group is not tech-savvy, you probably don’t want to pick your nephew getting his degree in computer science.
For a good test you should schedule between 30 and 45 minutes of time with your users, either at your home or business, or at their residence. Remember, they’re doing you a favor so take them out to lunch or bake them some cookies for lending you a hand.
I hope this gives you enough tips to help you plan for your own tests. In subsequent posts I’ll discuss how to conduct your tests and how to interpret your results. If you have questions, let us know!
June 7, 2008 at 11:47 pm |
This is something I have never done since 2002. I am not really clear as to how to pick those to ‘test’ my website. If I use friends or relatives it would seem to me they would have a tendency to be biased toward myself.
I am going through my AOL to do this testing. I hope you all get this, as I don’t want to be left out of the loop.
June 8, 2008 at 12:50 am |
You know, I have been spending hours and hours for years and years going through my site stats trying to get exactly the same information that this project will do. The biggest thing for me, as far as information that I gleen from my stats is products people want that I don’t have.
Now, who has the miracle to getting others whom buy and sell the same type of products you do, they ask for reciprocal links, I put up one for them, and sadly remove it after 6 months of their not keeping their word?
This idea of your’s is truly grand, and I can’t wait to spy on my subjects! Teresa
November 10, 2008 at 2:04 pm |
I am a middle age male who has MS. I have been trying to do business with Go Daddy and all they did is loose $150.00 of my money in an account. I have MS, therefore I have ADD automatically! It is nothing to get ashamed of, it just comes with the Disease. I am on Homestead learning right now, Nov.10th, 2008. I can actually read your instructions and I am taking my time. I thank you. I type one finger,but do take notes. Go Daddy has so much stuff on their web that it makes it hard to concentrate. I shall see how it goes, BUT thank your web technitions for using the KISS method. I have been a business man but not usuing a computer. Learning at 60 is not impossible but with MS it becomes difficult. Thank You again, Sincerely Mike Craft (albanyconcerts@yahoo.com)
November 25, 2008 at 2:10 pm |
You might also want to have the users test using different browsers in case some code doesn’t work on a different one.
-Dean
http://www.preparedsociety.com/
December 26, 2008 at 5:16 pm |
I usually go through and test my sites links a lot. If I get lost and cannot get back to where I want to go, rather easily, I know that I need to change a few things.
If the navigation on any site that I visit looks complex and overwhelming I usually do not stay very long.
I have a tendency to shy away from many articles that are filled with links within the text because there are to many rabbit trails for me to lose track of where I started.
Generally my rule of thumb is to only include about three anchor links in my content and oftentimes I will make sure that they open in a new window.
When I design websites for customers, I encourage them to go through the entire site and click every link. I explain the importance of simple navigation implementation and ask them to report to me if there are any areas that need to be improved upon.