The Data-Driven User Experience: A Better Way to Approach UX DesignPOSTED ON 5 April 2012 BY
The best user experience is relative to the worst user experience. But how can we determine either unless we test?
Perhaps in your mind you’ve already solved this problem by hiring an entire in-house team of UX professionals and interface designers. I’m sure they always agree on everything. And without a doubt, everyone is completely clear on his or her role based on his or her expertise, right?
Or even better, you may have hired one of those super-smart UX firms (you know, the one with the guy who “wrote the book”) to consult with you on exactly how your site needs to be built. Now all you need is for your project managers, business owners, designers and developers to agree with the 99-page report and recommendations, successfully obtain the increased budget request, and make this thing happen!
That said, it’s probably safe to assume the following:
1. You’ve spent a lot of money on this “User Experience” thing
2. There’s a bit of overlap in the roles, which is causing confusion on direction
3. It’s taking time and resources to reach a consensus on how this project will take shape
4. The HiPPO (Highest Paid Person’s Opinion) is not on board, or vice versa, thus jeopardizing this whole project
5. The deadline has been missed, while the scope of the project continues to increase
6. You just want this process finished and performing to everyone’s expectations
The UX dilemma
As design and UX professionals, we all have our “educated” opinions on what the best user experience should be. Sure, we all have some great examples to point to; some logical science and methodologies, great success stories, books and references and awesome design instincts. However, we need to embrace the fact that every client site, every client application and every client interface that we design comes with its own set of scenarios, personas and user complexities relative to that client and their audience. Those, we simply cannot accurately predict – nor does it make sense to even try.
Let the data drive the design
Companies that understand the value of testing can and should be continuously leveraging their testing tools to determine the best user experience.
Many corporate designs begin in the creative side of the house, move to production, and only upon an apparent lack in performance will they be measured and analyzed. By choosing first to measure a series of strategized design tests, we allow the live test results to inform our UX decisions, resulting in top performance and maximum efficiency from the beginning.
Taking advantage of A/B split testing, we can gain true and accurate insights on successes and failures of the interfaces we design. By analyzing this data, we can then make our strongest UX recommendations based upon the goals and KPIs of the site.
Case in point
We noticed a high click volume on the widget of a client’s product page. This widget was originally designed to help buyers make a more informed decision, ultimately leading to an anxiety-free purchase. After some review and discussion, our team decided that the widget was a bit confusing and probably needed some extra work. Seeing that the click volume was very high, an assumption was made that it must be an important factor in the user’s decision-making process.
However, rather than beginning the hefty process for a typical UX redesign (discovery, personas, wireframes, designs, user testing, etc.), we immediately decided to run a test version of this page that completely removed this widget altogether. We all believed that this feature was quite important to the purchase – but we just wanted to make sure. Surprisingly, we discovered that removing this widget from the page increased the conversion rate by 17%! This was a quick discovery made with a fast, simple test that increased the company’s revenue. Not to mention the billable efficiency increasing and potential sales losses decreasing due to smarter, less design-intensive changes to the site.
Please note – this 17% increase in conversions was discovered quickly without the use of UX experts, usability testing sessions, expensive UX testing software, senior designers and developers, information architects, project managers, or a super-complicated business process. Not that those people aren’t essential to an organization, but recognizing the power of testing can save a lot of resources.
The unmatched benefits of testing for UX
One of the greatest advantages of testing UX is the ability to uncover the unique complexity of different users, segments and audiences that exist for each client. Our designs become data-driven, analyzed, and optimized. Testing allows us to put to rest the opinions and arguments of everyone involved, operate much more efficiently and cost effectively and gain the trust and respect of everyone.
But we don’t stop there – a true testing culture will continuously monitor, measure and optimize for the ultimate user experience.
Maximize the value of your testing tools
For enterprise-level organizations using powerful testing tools like Adobe® Test&TargetTM, testing for UX may not be something that you have put a lot of thought or energy into. Take the necessary steps to educate designers on testing strategies and the features of your testing tools, and how they can be leveraged to gain valuable insights.
A paradigm shift
Regardless of the amount of lift or revenue gained through testing, the real value in testing user experience is the shift that takes place in our thinking and in our approach. If we allow real data to inform real design decisions, we will reap the rewards of analysis and objectivity all the way from designing interfaces to making complex business decisions.
Brooks Bell is the industry's leading online conversion firm. We help large online retailers and subscriber-supported companies maximize their online sales funnel by testing and optimizing all points in the conversion path.