Saved by testing, again

Feedback always adds value. Here, some late stage user testing transformed a team's design in small and simple ways.

It’s never too late to learn. I was reminded of this small, simple fact on a recent project. By working in collaboration with a strong designer and developers who were happy to tweak plans if it made things better, we made meaningful improvements to a complex product only weeks before a go-live.

The work saw us nest a financial services product inside an existing retail app. This introduced heightened security features and added complexity to a handful of pre-existing screens. Mindful that too much change could confuse existing users, we worked hard to reuse and recycle existing patterns and labels, while keeping new flows clean and simple. We were happy with the designs and prepared for usability testing and stakeholder showcases confident that our work would be well received and that learnings would be minimal.

---

Finding the problem

And we were right, almost. A series of unmoderated user tests, focused on the usability of the new security features, confirmed that the UX was straightforward and the copywriting clear. Users moved from A-to-B quickly: some even liked the journey. Similarly, when we demoed the work to stakeholders, they were impressed with what they saw. But, in both user testing and the playbacks it became clear that people didn’t actually understand the core functionality. In fact, even colleagues close to the project were confused. Again and again, we were asked, ‘How does this work?’ or, ‘Does this mean that…?’ We clearly had a problem.

---

A simple solution

Time and budget was tight - but we found a solution. Working as a classic PM-designer-tech lead Product Trio, we quickly honed in on the fundamentals: users understood how to use the feature, but were unclear on what this would mean for their money. The problem wasn’t about usability - the challenge centred on education. We decided that a short onboarding flow plus some minor tweaks to labelling would provide users with the context they’d need to progress with confidence. We designed the journey to preemptively answer the questions we’d heard during testing. Existing assets were repurposed wherever possible.

We went back and tested the new flow. Users found the onboarding journey easy to move through, so usability was unaffected. And the messaging was a success: not only did the questions we’d previously encountered vanish, users understood the product’s value proposition more clearly too. Stakeholders were also happier and conversations soon shifted from design details to future plans.

We moved towards launch knowing that our product had a solid design - and reminded that continuous discovery was the surest path to further improvements.

---

Takeaways

I’ve taken four key lessons from this experience:

• Always get feedback

  Testing really does work. Showing - or explaining - your work to those outside of the product bubble will blast away assumptions.

• Don’t validate

  Testing shouldn’t be about validation or verification. Work hard to find the wrinkles, the kinks and the gripe-enducers in your designs. We were lucky that we asked, ‘What don’t you like?’

• Problems first, solutions second

  We didn’t over-engineer things because we’d listened. This saved time and money without sacrificing quality.

• Friction isn’t always a bad things

  It’s important to keep things simple. But sometimes users will feel more confident of their experience, and more sold on your product’s value, if you take a little time to explain it.

---

Inspired by: Teresa Torres, Nielsen Norman

---