“If we’ve got data, let’s look at data. If all we have are opinions, let’s go with mine.”
– 
Jim Barksdale, former CEO of Netscape

Everyone who’s ever been through a complete site redesign knows they are, at best, frustrating and difficult processes. With seemingly endless meetings, constant efforts to please all the stakeholders, and the need to juggle schedules with your designer and your IT department, the efforts generally take far more time than they are worth.

Redesigns also are expensive and often lack clear outcomes or goals. Worse, they are a distraction from your team’s ultimate goals. Finally, the redesign process often brings to the surface so many conflicting opinions between agencies, marketers, CEOs, and other stakeholders that the end result can be farther from the goal than the original.

Fortunately, with today’s proven tools and methods such as crowdsourcing and A/B testing, there is no excuse for ignoring the data-driven path to redesign.

The most efficient way to overcome some of the challenges posed by a redesign is to incorporate data from users, stakeholders, and employees. When we’re tackling a home page here at Cro Metrics and we face a number of different opinions, we ask everyone to design their own layouts. Then we test them against each other. The data leads us to the truth.

Truth-telling data also helps you identify which specific elements of various designs work best. Once you’ve identified winners, combine them into a single design and test that. Will that quote from The New York Times make a difference or not? Will a Facebook signup option increase the signup rate? Data-driven decisions, not management opinions, should determine how a design will evolve.

An A/B testing tool such as Optimizely allows the marketing department to control almost the entire website redesign process with minimal involvement from any other stakeholders. This data-driven approach minimizes disagreements and maximizes efficient decisions.

A/B testing also makes it easier to capitalize on higher-converting design elements and copywriting. Once the home page design is complete, A/B testing also facilitates the remainder of the process to drive overall design for the rest of the site so each page operates efficiently and with purpose.

Even if you opt for a big (and bombastic) redesign, incorporate data at every turn. Test the elements that are contentious. Test the copywriting, images, flow, signup forms, and call to action buttons.

  • Tools such as fivesecondtest.com will let you know if that icon or button label really means what you think it means to real people.
  • Services such as usertesting.com will enable you to see how real users react to the site, which elements grab them, and how they interpret the language you use.
  • I recommend undergoing this process before you code your site by creating a non-functional mockup using a tool like Axure (you even can link Photoshopped pages together to simulate the real site).
  • If you’re using an agency, choose one that will be your partner in an iterative approach, and let them know you expect at least two or three rounds of revisions along the way.

However you approach the process, remember this: Data is your friend. Use it to make sure the arduous task of redesigning your website delivers the value you are seeking.