One of the services offered by our User Experience team is what we call an Expert Review. The end result of an Expert Review is an easy-to-digest list of must-fix issues, smaller issues, and positive attributes. In the field of usability, an Expert Review is also called a Heuristic Analysis or Heuristic Review. It’s an efficient way to identify usability issues with your website or product, because we do them in a day or two; inteviews or usability tests generally take a few days longer.
How to do an Expert Review
The basic idea to walk through most of the common tasks and identify potential stumbling blocks. But, instead of pointing out only the negative (which is usually easy to do!), Expert Reviews from Atlantic BT go one step further and call out specific features that work particularly well.
Issues are identified with a list of heuristics (i.e. good design principles) and the reviewer’s own expertise. Nielsen’s Ten Usability Heuristics is a good starting point for developing your own list. Each heuristic is backed up with research. For example, we know from cognitive psychology that people can only store a limited amount of information in working memory. Therefore, web pages should be designed to minimize or eliminate the need for a user to remember information from one page to another.
When conducting a review, it is important to remember “The Evaluator Effect”. Research has shown that when multiple usability specialists look at a website (or any product), they often identify different issues. To minimize this problem, it is important to involve multiple usability specialists in this process. It always necessary to include an experienced usability specialist if a junior-level usability specialist is involved. We usually have one usability specialist do a thorough review, and a different usability specialist do a quick review.
How to Format an Expert Review
An Expert Review should be easy to scan and fun to read. There is no reason to create a long, boring list of usability problems if nobody is going to read them. At Atlantic BT, we create a bulleted list and use a three dot severity system:
The green dot is used for positive findings. The red dot is used for must-fix severe usability problems. And the yellow dot is for less severe problems and nice-to-haves. If it’s a big site or there are a lot of problems, introduce a 4th dot (orange). For a lot of our clients with smaller sites, we have found that our three dot system works very well.
Don’t Forget Recommendations!
An Expert Review is not complete unless each item is actionable. Therefore, each issue should have recommendations on how to resolve it. The recommendations presented in the Expert Review should not be final, because a multi-disciplinary design team (including marketing and design) will also have potential solutions.
When to do an Expert Review
Since a website or product’s users are constantly changing, it’s necessary to constantly check-in to make sure their needs are being met. An Expert Review should always be done before site redesign efforts and any user research (e.g. usability testing or surveys). If major issues are discovered in an Expert Review, further user research on the existing design is often unnecessary and not cost-effective.
Why Expert Reviews Save Money (Maybe)
For a simple website, an expert review is much more cost-effective than a usability test. A standard usability test requires at least 5 participants, which need to be recruited and then paid for their time. Before that happens, a research plan needs to be developed with a list of tasks and other materials for the study. Depending on how hard it is to find participants, it takes at least a few days to complete a usability test.
Advanced user interfaces require more usability specialists looking at the site. In this case, it may actually be more cost-effective to conduct a usability test.
Usability tests are great at finding unanticipated issues that may go undetected in an expert review. But, if there are salient problems, then an expert review should be conducted and those problems should be fixed before any research is conducted. Participants in a usability test have a hard time looking past major issues, and valuable time will be wasted.