User Testing: Designing something that works
What is User Testing?
User Testing is a way for us to understand user behaviours, habits and methods of using a system. This information gives us insights into what users think about a site, an application, a product, or a process. They can point out what site content is memorable, what people feel is important on the site, and what ideas for improvement they may have. Good design isn't something we can eyeball into existence. It's something we test, learn from, and refine, over and over, until what we've built actually works for the people using it.
For one of our clients user testing sessions, we ran A/B tests on our designs to find out which layouts made the most sense to everyday users. Rather than relying on opinion, we put the options in front of real people and watched how they used each one.
The testing showed us where users expected key content and actions to sit on the page, and how that differed between designs. It also showed us what people remembered from each version once they'd finished: which products, messages and features stuck with them, and which ones they missed entirely.
This gave us clear evidence for which designs to take forward. It helped us settle any doubts about which layout would work best and let us make those decisions with confidence, knowing they were backed by how people actually behave. users. The results showed us which designs to go with and helped us settle any doubts about which layout would work best.
When should you be using User Testing?
What is good design without functionality?
We don't want to create things that just look good, we want to create things that people will actually use, things that will actually help. We use the data found in these user testing scenarios to back up our design systems as what might seem intuitive to people who use it every day, may not be to new users. User testing helps to uncover the pain points that people might encounter before anything goes live.
User testing can be done in a variety of situations:
- Before you have a design, to inform personas, journey maps, feature ideas, workflow ideas
- When updating an outdated design system, to see where actually needs the most improvement overall and ensure that this is where the most resources are spent
- Once we have finalised our design to back up our thinking, and validate the course chosen
Moderated vs. Unmoderated Testing
How do we do it?
Neither approach is inherently better, the right choice depends on what stage the project is at, how many participants we need, and how much nuance the questions require. Often we'll use a mix of both across a single project: unmoderated testing to validate at scale, moderated sessions to understand the details behind the numbers.
What is Moderated Testing?
Human to human, real time testing. A facilitator sits with the participant (in person or over a call) while they work through tasks, asking follow-up questions, probing hesitation, and digging into the "why" behind a click. Moderated sessions are slower to run and harder to scale, but they give us the richest insight, we can adapt on the fly, chase down an unexpected reaction, and see body language and tone that a recording alone won't capture.
What is Unmoderated Testing?
The participant completes tasks on their own time, usually through a testing platform, with no facilitator watching in real time. It's faster to organise, easier to run across larger groups, and removes the social pressure some people feel performing in front of an observer. The trade-off is the depth of responses we get. With unmoderated testing we can watch what someone did, but we lose the chance to ask "what were you expecting to happen there?" in the moment.
Benefits
There is a long list of reasons that User Testing can benefit your product, including saving you time, money and resources, all before you’ve even launched. Here is a list of the reasons that we recommend user testing:
- Low barrier to entry
- Easy collection of data
- It’s easy to add on
- Doesn't take time to spin up
- Lowers risk of bad design, and mitigates features that are not useful
- Replaces assumption with evidence which means that decisions get made on what users actually do, not what the team thinks they'll do
- Builds internal confidence as you know its a design backed by testing is easier to sign off, defend, and move forward with
- Surfaces issues early, when they're still cheap and easy to fix
- Keeps the user at the centre of the conversation