Designing with Data Insights: An Optimization Strategy
Web analytics tools are more robust and accessible than ever before, but there is still a customer data gap, with too much data and not enough actionable insight. As a result, the dashboards and monthly reports that analysts toil over can end up in inbox purgatory, instead of having a real impact on the business. The problem is that web analytics data is often used only to report on marketing ROI, which is important, but more or less just reporting the weather. You may get an even better “return on analysis” by using web analytics to optimize a digital experience by designing with data insights.
Data insights connect design to outcomes
Executives might not be able to act on web analytics every day, but designers can. Web analytics measure how people actually use your website, with the emphasis on “people” and “use,” which is precisely the kind of insight that designers need to deliver a great customer experience.
UX researchers often interview people before a system is built about how they think they will use it. This “voice of customer” data provides valuable insights for user flow design, helping designers to map the path users will take through your website and the actions they might take there. But that’s only part of the story.
Web analytics data confirms whether your user flow design actually works as expected—and highlights opportunities to improve it. Understanding the measurable outcomes of a user flow helps the design team deliver a great experience to both the business and the user.
Start with your hypothesis
By comparing the user flow design with the actual user flow a designer can begin making a hypothesis for testing and optimization. Whether an analyst or a designer takes the lead, the hypothesis states the problem and the expected solution to that problem through a new design approach. Because the realization of a problem is based in data, but the solution being tested is based in design, the best hypothesis is specific to a user behavior and a single expected measurable outcome. For example:
“The shopping cart abandonment rate on the shipping page of the check-out process is high because the shipping calculator is confusing to users. We see the evidence of the user confusion by analyzing the number of times the calculator is submitted with errors. By using a radio button selector to improve the calculator’s ZIP code and city options, we expect to decrease the abandonment rate on this page.”
Test your design hypothesis
After a new design launch there’s a tendency to step back and move on. But this is actually the best time to start optimizing your design with data insights. The design team still has their heads in the game, and their ideas are still top of mind. So celebrate the launch and then start to apply analytics and testing to your design hypothesis and principles. And encourage the original designers to riff on their work based on the data. They’ll still be fluent in the principles that guided their design work and can apply them in new ways to improve results.
This testing approach promotes consistency in the optimization process because it expands on the design vision instead of introducing piece-meal design. A designer who is part of this process will also grow by staying connected to the business outcomes and purpose of the experience they designed and honing their designs based on the results of their previous work.
Iterate and improve design
Using data insights to test and optimize design iteratively provides quantitative rationale to support recommendations for the larger redesign, and provides benchmarks to monitor and improve the experience afterwards. Making many drastic design changes in a test could lead to a positive lift but offer no lessons about why it improved. The process of iteration will generate a repeatable set of data-informed design principles and build credibility. And because any new design can be measured, designers are empowered to take more creative risks.
Analytics and testing should connect a design team to the outcomes of their work and plug them into a larger optimization team that can rally against quantifiable problems. Content, UX and Creative should all be at the table, using data to discover how design elements influence a user toward or away from the intended goal of that design.
Analysts may have never thought they’d be helping designers, and designers may have thought web analytics were just a way to count how well ads were performing. But great digital experience hinges on these two practices working side by side. Companies that deliver great experiences win, and that’s a KPI everyone can rally around.Tags: Data, Design