10Jun2010

Website Redesign: Big Bang, Gradual or Evolutionary Approach?

I'd like to look at three approaches to redesigning a website:

Big bang — move completely from an old to a new design in one go
Gradual — implement the redesign by sections over time
Evolutionary — avoid redesigning altogether and focus on improving the existing site

These approaches have their advantages and disadvantages. Let's consider them in reference to a large, content-rich website.

Big Bang Approach

Advantages

Keeps your site looking consistent. On one day your whole site simply changes from one design to another.

Easily understood by your visitors. Whether or not they like it, your visitors will be familiar with this traditional method of redesigning. Messaging around the redesign is not about what is going on but more about why it was undertaken.

Easy to sell this approach internally. It's easier for stakeholders to understand the concept of moving completely from one design to another than an approach that involves interim steps where more than one theme may be viewable to visitors.

Keeps your web analytics clean. Having a single cutover date makes it easy to immediately start measuring the effectiveness of the new design and comparing it to the previous one.

Disadvantages

It can take a long time. You need to migrate all of your pages over to the new design in order to launch and cannot gain any of the benefits of the new design until you do. While you're doing the project, you also have to maintain the live site and so are likely managing content in two places.

It can create a jarring user experience. No matter how well you advertise the upcoming redesign in advance, the changeover is still going to catch many visitors by surprise. Suddenly they have to relearn how to use the website, and many of them won't be happy about it.

Keeps your site static. Redesigning a website in this way is such a large endeavor that it is only undertaken infrequently. This can lead to your site design being static for longer than it should be based on changes in business conditions, your knowledge of your visitors, etc.

Gradual Approach

An increasingly popular approach is to launch a site redesign in stages, starting with the most popular pages or even just the home page. This approach has become quite common with news sites, which may just start with a home page redesign and go from there.

Advantages

Easier to implement the redesign. For large, content-rich sites, it's easier to move smaller sections of the site over to a new design than to do it all at once. You can also control the pace of the redesign so that the web team can keep up with maintaining the live site as well.

Shorter time to launch. If you're starting with, say, your home page, you can be up-and-running with a site redesign in a much shorter timeframe than if you were going to switch over the whole site. This provides an opportunity to gather feedback from your visitors much earlier and to apply it as you redesign the rest of the website.

More gentle introduction to the new design. Most people dislike change, and so introducing it at a more gradual rate will make it easier for your visitors to get used to it and adapt the way they use your site.

Disadvantages

Can be complex to do. If the redesign involves information architecture changes, it can be difficult to introduce the new IA gradually while still maintaining the old one. Extensive use of redirects may be necessary in order to avoid broken internal links. The same challenge is true for sitewide features like a new search tool.

Harder to sell to stakeholders. This approach is harder to sell internally as there will be concern that the site looks broken. Most stakeholders don't understand how task-focused web users are and instead focus on keeping the look of the site consistent above all else.

Evolutionary Approach

A third method, the 'evolutionary' approach, tries to avoid the need for a redesign altogether. Rather than change the overall design of the site, the pages, content, features and functionality within it evolve according to changing visitor and organizational needs.

Netflix is a good example of this. Although the overall theme of the site has not changed substantially for years, the pages within it (particularly the home page) are constantly changing as the team tests out new ideas with their users.

Advantages

Avoids the pain of redesigning. Large website redesigns are very challenging to do and will tax the resources of most web teams. They are often very political (especially what goes on the home page) and can be subject to the whims of higher level executives.

Frees up time to work on something useful. Time spent not redesigning can be spent on something more relevant to the bottom line, such as A/B testing of checkout flows, refreshing old content, or improving the conversion rate of a signup page.

Keeps your visitors happy. As I mentioned before, people hate change and website redesigns are generally received negatively. Unless your website is in desperate need of an overhaul, you may be better off leaving it as is for as long as possible. Better to direct your focus instead on ensuring that visitor tasks can be easily completed.

Disadvantages

We're bored with the old site. The teams responsible for the website will become bored with the current design and will want to update it. Redesigning a website is also a very obvious way of showing 'progress' to internal decision makers when it comes to annual review time. It's often harder to sell less tangible improvements made to a website unless they have more obvious visible components.

24 Comments Posted

I've been trying to do the big bang approach on my site since February. So far, all I've managed to do is reassess the content and rewrite some of it. Other work keeps getting in the way. :(

1. Posted by John Faulds on June 10, 2010

Im all about the big bang! However i think its important to retain some of the integrity of the old site so that users can still identify with the brand. Good post.

2. Posted by Web Design Nottingham on June 11, 2010

Im all about the big bang! However i think its important to retain some of the integrity of the old site so that users can still identify with the brand. Good post.

3. Posted by Web Design Nottingham on June 11, 2010

I just did a Big Bang on my site and left the original content in place. But the biggest disadvantage so far is the immediate Google changes in my search term ranking. Everything competitive that I was on Page One for went away in 12 hours. Well I guess I had to do it and every journey begins with a first step... even if that step is a Big Bang !! Regards - Mal

4. Posted by Mal Milligan on June 12, 2010

This is a really great blog, it's so very informative! Paul Alexander www.wowwebsitedesigners.com

5. Posted by Paul Alexnder on June 14, 2010

Each strategy definitely has it's advantages and disadvantages, sometimes I think it is easier just to start from scratch and move important content over after up-dating it. Change is good!

6. Posted by Albany Web Design on June 15, 2010

I don't agree: Yes, the Big Bang method takes some time until everythings is new. But all other methods waist time and money due to the complexity of updating just parts og a hp...

7. Posted by ff-webdesigner on June 17, 2010

You know, I think there's no need to worry so much about it - just get your new site up - your customers will love it!

8. Posted by Bristol web design on June 18, 2010

While I'd love to put new design and coding skills to use in re-designing / re-building my website (which is a few years old now) I'm loathed to do so because of the likely impact on SERPs. It took me a year or more to get #1 (or close) for most of my target terms - a long slog that I'd rather not repeat.

9. Posted by Perth Web Designer on June 18, 2010

Great effort. the blog looks informative & well explained. we are planning to implement big bang in our site. thanks you...

10. Posted by webdesign chennai on June 19, 2010

Great effort. the blog looks informative & well explained. we are planning to implement big bang in our site. thank you...

11. Posted by webdesign chennai on June 19, 2010

It really depends on the website... If you feel that the design of the site has gone out of time aesthetics-wise, code-wise, usability-wise and functionality-wise then you should use the big bang theory but if you do have certain parts of the website that associates to your brand like logo and theme then leave it be or try making use of another color combination...

Now on some levels its acceptable to change gradually specially if you are using a theme design which has parts that does not necessarily depend on each other, because some complicatedly designed websites have oneness on all the parts of the website which means that if you change one part the unity of the design becomes disoriented.

As for the Evolutionary approach it really depends on the subject or the purpose of the website.. Let's say google. . . google has not changed there design layout because it is very effective and simple yet functional, it is a no brainer how to use google, one of the major concepts of web deisign is "not to make your visitors think too much on how to navigate or use the website", although at some times, seasonally they add a subtle design which goes with the occasion like christmas.

But as for blogs having a main subject like; web design then that's a different story. Probably a yearly makeover would be enough, showcasing your skills as a web designer and at the same time gaining more reputation and of course you can discuss whether your readers like the new design or not. . .(instant article :). . .

12. Posted by John Tanedo on June 20, 2010

The catchy website with the interesting posts. You give the nice information that many people don't know before. most of your contents are make me have more knowledge. it is very different. I was impressed with your blog. Never be bored to visit your blog again. Have the nice your time.Keep enjoyed your blogging.

13. Posted by naples webdesign on June 20, 2010

I'm for the gradual approach if you can pull it off. Look at high-traffic sites like amazon and ebay. They've never done a big-bang overhaul, but more like a continuous change, test, evaluate, change again. If you pull the rug out from under your users too quickly, they'll get confused and leave. (they might even hunt you down with torches and an angry mob) Remember how facebook made some huge changes all at once? Lots of users had serious problems with that.
Think of the users and how they will be impacted- that should help you decide your gameplan.

14. Posted by Dave Vogler on June 20, 2010

I like it somewhere between gradual and big bang.

I like to work in cycles, and try to work in short burst with a beginning and in end. In between cycles is where retweaking and assessments take place.

The projects can very well be comprehensive, but it'll need to take place in parts to get it done. A complete overhaul takes a lot of time and really hard to back out of if not done properly or the proper resources dedicated .

Additionally, I like to thing frequent users like to see a site change incrementally from time to time

15. Posted by tobiaswright on June 20, 2010

Great Post!

Big Bangs will always be neccessity in many cases due to lack of interest in evolving a website many companies have!

16. Posted by DSM Design on June 21, 2010

I think the article says it all: 3 different techniques for use in different cases. I've led all 3 approaches and I think the keys to success are:
* Big Bang - have another set of servers (VMs) for the new site. Do NOT build on the same platform as the production site. It makes the cut over EZ (with a fallback just in case).
* Gradual - Be careful to avoid upgrade fatigue in the middle of the rolling upgrade. You must be quick about it or you end up living in a partially painted room for an embarassing length of time.
* Evolutionary - Start-ups need to evolve often to meet the market; Established companies need to evolve features to meet the consumer expectations like mobile versions & social media links.

17. Posted by Steve Heibein on June 21, 2010

I like the big bang followed by the evolutionary. I site will need to be altered to cater for the audience and you don't need to redesign each time this happens...

18. Posted by Evan Skuthorpe on June 22, 2010

I really like the big bang webstyle but i also like websites done under www.instacommerce.net

Regards,
anne

19. Posted by anne on June 22, 2010

Cool article.

I'm interested in more information about sales related to these different approaches.

In what ways can a designer acquire new customers based on their need for a new design, rather than a new website all together? It's tough to contact a potential client saying something along these lines: "your website sucks, you need a new one". Must a designer always wait for someone to contact them? Is there a respectful approach to contacting people about their poor website designs?

For the record, I understand 'beauty is in the eye of the beholder".

20. Posted by Lucas Larson on June 22, 2010

Cool article.

I'm interested in more information about sales related to these different approaches.

For the record, I understand 'beauty is in the eye of the beholder".

21. Posted by Lucas Larson on June 22, 2010

Just came across your blog and I like your approach. Anyway, from my experience almost 70% of my customers prefer the Big Bang web design pack :) It's easier to create something new, rather than modifying a old design.

22. Posted by Webservice on June 23, 2010

I think it depends on what you are trying to do. At the moment Im still working a lot on my own personal site, so its all very gradual.

23. Posted by Winston Muller on July 01, 2010

If you can hire a web design company, or have the time and the energy to do it yourself, a big bang redesign is the way to go. Otherwise, stick with the evolutionary. The 'gradual' method never works... it ends up making your website into a 2 or 3 headed design monster and your site's visitors will only be confused by the inconsistencies between sections or pages.

24. Posted by The Web Pro on July 12, 2010