Digital Fundraiser: Nonprofit Web Design Process Part 2d: Usability Tests

Tuesday 22 October 2013

Nonprofit Web Design Process Part 2d: Usability Tests

Note: This is the fifth in a series of posts about the Nonprofit Web Design Process. See the end of this post for a linked index of other posts in the series.

Purposes

Usability Tests provide insight into how your users interact with your website. There are many approaches to usability testing, including card sort tests, 5-second tests, and moderated, task-based tests. The approach you choose for your project will depend on your timeline, budget and available resources.

Usability Testing of any sort is a good idea to incorporate into your design process to validate (or invalidate!) the design decisions you’ve made along the way. With any design, there is some guess-work involved on the part of the designer so getting a reaction from an end-user may reveal some different options for your design.

Our purposes for Usability Tests are:

To gauge users' reactions to a new design
To validate that users are able to access key content or complete tasks within your new website structure
To identify stumbling blocks with your new site structure so we can attempt to resolve them before launch

Methodology

Moderated or Unmoderated?

The fastest and cheapest route for usability testing is unmoderated testing. Here at Blackbaud, we use UsabilityHub to run unmoderated usability tests. We are big fans of their 5-second tests to elicit opinions about homepage designs at a glance. Their Click Tests and NavFlow tests are useful too but I’ve found that I prefer doing moderated tests when I want to verify that users can easily find specific information or complete specific tasks.

Moderated testing, conducted 1 on 1 with participants, takes longer and is more expensive but allows us to dive deeper with constituents to not only see how they interact with the website but also to find out why they make the decisions they do.

Recruiting

If we decide to do moderated usability tests, we’ll need to start by recruiting 6-8 participants that are representative of the site’s target audience groups. To find these participants, like with Card Sort Tests and User Interviews, you can rely on your internal stakeholders to identify folks that might want to participate and fit the profile. You can also start with a Survey and ask for volunteers that way. I’ve had success with an online recruiting tool called Ethnio that serves your website visitors a pop-up with a few screener questions. Those that meet your criteria are then asked for their contact information and you can follow-up with them for the card sort.

Since a moderated usability test can take about an hour, we suggest offering an incentive for participants. We like Amazon gift cards. Sometimes, the participants may prefer to donate their incentive to your organization, which is a great option to offer.

Research Questions

Before planning a usability test, it’s important to define what questions you're hoping to answer through the testing. For unmoderated tests, you should limit yourself to 1 or 2 questions to answer with each type of test. So, for a 5-second test, your question may be, "What adjectives would someone use to describe our new homepage?" For moderated tests, you can ask more questions. Typically, we're looking to find out, "Are users able to easily do or find X, Y and Z on the new website?" where X, Y, and Z represent important tasks or information.

Examples of the tasks we might test are:

Where would someone click first on the homepage?
How would someone make a donation?
What path would someone take to find a recent press release about the organization?
Where would someone look to register for an event?

Once we identify what questions we want to answer, I’ll fashion a script for a moderated usability test to guide the conversation during the course of the test. The script will include some introductory language to set up the test and will then list out the tasks with probing questions in between each task such as:

What would you expect to see on that page?
Do you see anything odd or out-of-place on this page?
Is anything missing from this page?

The Prototype

While you can get pertinent information by only testing a single wireframe or page (for example, the homepage), it’s sometimes useful to have other pages to show the participants to broaden the scope of your test. A prototype is a collection of pages that simulates the experience of browsing your website. As I work on the usability test script, it usually becomes clear to me what additional pages I should be prepared to show as part of my prototype.

Your prototype can be something as simple as paper sketches or wireframes or a more complex, interactive prototype. What we use depends, of course, on time and budget, but also on what stage of the design process we’re in. I think it’s ideal to do moderated usability tests once we have a solid set of wireframes finalized but before we begin the visual design process. That way, we can use the wireframes as our initial prototype and add pages here and there where it makes sense for what we’re testing.

Deliverables

For unmoderated usability tests, the deliverable is often just a report of the results of the test coupled with some recommendations for next steps. For moderated tests, I like to create a deck of recommendations for wireframe/design changes that are each supported by the test results or participant feedback. I also try to record each usability test session so I can share the recordings with the client. It’s worth noting that a round of revisions to the wireframes/designs is almost always required after a round of usability testing.

What's Next

My final user research post is on Constituent Personas so stay tuned for that next month! Then, I'll tackle what we do with all of our great user research data and how it influences Content Strategy, Information Architecture and Visual Design.

Digital Fundraiser