Bummer! This is just a preview. You need to be signed in with a Treehouse account to view the entire video.
Results and Recommendations2:38 with Dan Gorgone
In the end, the key deliverable from your usability tests should be a set of recommendations you deliver to the people responsible for the website. They should be based on your test results as well as personal observation.
Things to Consider
- How will you review the tests after they have been completed? Do you have recordings, notes, or other information you can refer to?
- How can you best report on the results of each test? Do you have qualitative analysis to explain what happened and how problems arose?
- Are you able to provide recommendations for each problem you've found? Which ones have clear solutions? Which ones are more troublesome? Is more testing needed?
- Who is ultimately responsible for making the changes you suggest? Have you given them enough information to compel them to act?
- When will you conduct testing again, and what might you test?
Interpreting tests results in some cases will be very easy.
Simple observation can provide many clues about what website changes should be made.
Also, many users will tell you directly that something is wrong, bad,
it's not working, or that they don't like something.
These comments often relate to usability issues but it is important to separate facts from opinion.
When providing results, consider the experience each user had.
Did they successfully accomplished each task,
and in each case, where they satisfied with the results?
It's very possible for a user to eventually find what they were looking for after a long period of time.
This doesn't necessarily mean the website is in good shape.
Other times, tests just don't yield clear results.
If you're unsure what some of your data means,
make a note of it so you can review it later and then move on.
When all is said and done, the key deliverable from your test should be
a set of recommendations you deliver to the people responsible for your website.
It should be a clear list of directives based on your research,
the test you conducted, and personal observation.
The data you collect will help stir you in certain directions
when generating recommendations for improvement.
For instance, if every test user had trouble finding our return policy
on the Shirts 4 Mike's site, there must be a problem.
Sometimes the fix is obvious like adding a link in an expected location but not always.
More testing can often reveal the best way to rule out different areas
and decide upon the best solution.
Your recommendations don't have to be produced in the form of one long check list.
And in fact, a list like that can be misleading.
You should be able to provide background on each item and rank them in order of importance.
Fixing typos on a page isn't as mission critical as fixing a shopping cart that crashes every time.
Therefore, you should organize these proposed fixes into a format that make sense
and that's easy to implement especially if you are the one that has to do it.
Lastly, understand that while findings can show areas for improvement,
it can also confirm that something is working.
Neither case knowing something works or doesn't work can be equally important.
If you can run these tests with three to five people each month or two,
you'll have plenty of data to keep yourself busy with changes and improvements
especially if you end up implementing them immediately.
And if that's the case, you'll have something new to test the following month.
You need to sign up for Treehouse in order to download course files.Sign up