At last week's E-engage Your community event in Wellington, former Yahoo! and TradeMe user advocate Natasha Lampard talked about user testing of your website and how to do it well. One of the key steps is actually to have people outside your organisation actually test your actual website.

Then the question arises of how many users should you get to test? The good news is that Effective User Testing Doesn’t Have To Be Extensive

Jakob Nielsen’s study on the ideal number of test subjects in usability tests found that tests with just five users would reveal about 85% of all problems with your website, whereas 15 users would find pretty much all problems.

The biggest issues are usually discovered by the first one or two users, and the following testers confirm these issues and discover the remaining minor issues. Only two test users would likely find half the problems on your website. This means that testing doesn’t have to be extensive or expensive to yield good results. The biggest gains are achieved when going from 0 test users to 1, so don’t be afraid of doing too little: any testing is better than none.

A key point that both Natasha and Jakob make is that wa website is not so much a pu8blishing mediaum 9and therefore subject to the rules of news media etc) as it is a tool for getting t5asks done. So the logical thing to test is how long it takes and how easy it feels fro0r your users to compelte the tasks they come to your website for.

For those of us trying to run effective websites, the big questions are then, "what are those tasks?" If tyour organsiation has a website, could you, or any of your colleagues, write down in less than a minute the 5 main tasks that people come to your website to perform?

If not, that is where your next site review needs to start. Then test it.

If you're new here, you may want to subscribe to my RSS feed. Thanks for visiting!

Share this post

Hide Sites