User Experience Stack Exchange is a question and answer site for user experience researchers and experts. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

UX Myth #11 examines the perception that websites need to be redesigned periodically.

As the article explains, there's quite a lot of evidence to suggest users are naturally opposed to major changes to a familiar site.

As UX consultants, how should we manage this risk when addressing major new business objectives and over-enthusiastic marketing depts?

How can we roll out new designs without driving away loyal visitors?

Do you regularly redesign your sites? Why?

share|improve this question

We redesign sites for non-profits to add usability, accessibility, standards and train their people to maintain using these. We work hard to keep a similar Look and Feel, keeping any colors that pass contrast, not redoing logos, if they have top/left/right navigation, keeping the navigation in same spot, just improving it's use and structure.

The sites don't look the same, and may not lay out the same, but there's continuity in the design and elements. If we're going to something completely different, we always do it in steps and stages over a planned period, and ask for user responses between each phase.

Dealing with NP sites and their users though is different from dealing with a designers blog, where I don't think these concerns apply, but for NP's they're a must in my opinion.

share|improve this answer

For some organsiations - specifically two that I work for, gradual enhancement is the practical reality of managing their websites, while a complete rebuild might be a long term vision.

Take organisation number 1. Huge site, vast number of users across various segments, millions of visitors each month. Given the complexity of the business, internal and external stakeholders, pressures, egos etc gradual enhancement is the only way to get things done. BUT -- its on an overall roadmap. There is a "grand plan" albeit at a high level, that helps prioritise work, and specific changes are usually tied to work else where in the organisation. Its the only way to get funding.

Now take organisation number 2. Vast number of sites - 59 in total. No consistency in segmentation, and traffic varies from site to site. They are desperate to consolidate, rationalise, simplify and create consistency, not just from a UI/UX perspective, but infrastructure and content management aspects as well. They don't have the budget, skills, technical capabilities, design resources etc to deal with a large rebuild programme. Even tactical changes take a while to implement. The only way to keep any improvements happening is to take small incremental steps and a gradual release of new designs and functionality.

share|improve this answer

In my experience, alot of redesigns are spurred by new hires. "I am the new sheriff in town and I want to redesign the website!" Sometimes it's based on a new product launch like, "We are changing all those other pages, why not refresh the look." Personally, I change my blog theme once in a while because I get bored with it.

Summary: Most redesigns have nothing to do with what the user wants.

share|improve this answer
I think this is right, but I wonder if it's just territorialism. A lot of the champions for site redesigns simply didn't know how much could be gained by small improvements and testing. – Todd Moy Jun 25 '10 at 21:17

Re-designs and product reiterations should really all be decided on by A/B testing. In some cases a huge site re-design is necessary (like updating a 2000s website), but in most cases, small periodic design changes that are supported by data is your best bet.

A/B testing will help tell you if your new design actually is turning away loyal visitors or reducing sales.

On all of the sites I work on, we are constantly adding new features, changing designs, and removing things. And it is all supported by our A/B test data.

This is a really great article over at 37Signals about their A/B testing with Highrise:

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.