Why Website Redesigns Should be Avoided (And a Proven Process)

At a certain point, companies can feel stuck: traffic is increasing, but signups or sales are plateauing.

Does that sound familiar?Probably. I mean… almost everyone is struggling to increase conversions.

But here’s the deal:

For many, the default solution after a while often ends up being a complete website overhaul. Unfortunately, website redesigns are highly unlikely to solve the problem (and can be astonishingly expensive).In this blog post, you’ll learn why a complete website redesign is one of the worst solutions to increase conversions, in addition to what experts say is the best solution.

The problem with website redesigns

A redesign might lead to an improved user experience, but its ability to convert visitors into customers is never guaranteed to improve.

A Website Redesign Does Not Guarantee ROI

A website redesign is just like buying new clothes: you feel good, fresh, and renewed with confidence… and seriously, who doesn’t like new stuff?

However, changing your look doesn’t mean you have to buy entirely new clothes. Adding accessories and tweaking certain aspects of how you dress can make a major difference.

The same principle applies to your website. A complete overhaul is expensive, but not always necessary. Including time spent for planning, designing and deploying, your pockets will likely be empty by a quite few thousands of dollars at the end of the journey.

Then the problem goes even deeper…

Just imagine this: You spend months working with your team on a new website. Your goal is to grow the company and increase revenue by getting more conversions. Then it’s time to get it live - and you notice after a few weeks, that your conversion rate stays the same. That would suck.

Redesign did not improve conversions

So let’s avoid this and start by diving into even more details with 7 reasons why you shouldn’t redesign your site.

1. It's expensive

First things first, going all out with a full website redesign will set you back by a few thousand dollars. Outsourcing the work won’t be cheap, and if you have staff, the same applies. Initially, the costs of doing the work in-house can be less obvious, but the man hours of your employees will definitely add up. And that’s money - with no return on investment (ROI) that’s guaranteed (I’ll talk about how to increase that ROI later in the post)

2. It takes too much time

Completely redesigning a website can take months of work. It’ll require time from many members of your team - from marketing, design and engineering, constantly, for an extended period of time. Remember, this is time that could be spent developing other features or on other things that are more likely to grow your company.

Thorren Koomans director of strategy at a digital web studio, mentioned the undertaking of such project could last up to 6 months or longer. “Most web projects should allow for 12 to 16 weeks from the time that the project kicks off to the time that the website launches. Where complexity is higher or the scope of the project is particularly large, projects can take 6 months or longer.”

3. Can hurt SEO

If you decide to go with a full redesign, once live, your site will be completely different; therefore, this means it can affect your rankings pretty dramatically, as shown by this chart from Search Engine Journal:

Website Redesigns and SEO

A complete article could be written on the topic.Your rank in search engines is determined by over 200 factors; however, SEO troubles caused by a redesign are often the result of improper technical planning: Your existing URLs will most likely have been crawled by Google and other search engines a handful of times.

If some of your links change or are broken and redirects to an error page, the page authority of the affected pages will decrease. Almost automatically, a lower page or domain authority is likely to sink your ranking into the deepness of Google.You wouldn’t want to lose all those hard earned backlinks, wouldn’t you?

Redesigns often results in broken links

Without proper planning, a redesign can definitely be risky for your ranking and traffic. Use tools like Google Webmaster Tools to detect potential or ongoing URL problems.

4. You redesign for the wrong reasons

Why exactly are we redesigning our website? is an important question to ask.

Bored of your website's design?

If the main motive is simply for your company to keep up with design trends and refresh your image. You’ve got the wrong reason.If it’s because your competitors got itself a new site, that’s another bad reason.

If it’s because your bounce rate is too high or sales are too low, a full redesign is rarely the solution.

Websites are built to be improved over time based on continuous learnings from key measurements. Not to be re-built every two years!

5. Designers are rarely conversion experts

Designers are artists, they make things pretty. But when it comes to optimizing your conversion rate, that’s science.

Although I know some that do, not all designers focus on how to design for conversions. More often than not, it’s the digital marketer's job to analyze data, create A/B tests and perform website improvements to increase conversions. The skill sets are completely different, but also complementary. Designers and marketers have to work together.

An underperforming redesign often involves a carte blanche to visual creativity. In other words, a designer can build you pretty site, but there’s no guarantee it’ll actually make you money.

6. Regular visitors will get confused

Every time Facebook, Twitter and other major sites change a minor feature, there’s an outcry of people complaining they’re confused and don’t like the changes.

Twitter users not happy of new feature

It’s normal: people generally don’t like change. Facebook’s redesigns, for example usually follows best practices: only small changes are made and rolled out progressively (and tested extensively). It makes it easy for users to get quickly get used to it because of the familiarity.

The real problems generally come out when you make a complete redesign and start from scratch or when you redo the visitor’s familiar path to micro and macro conversions.

Change destabilizes; people that are used to your user experience and navigation flow will have to adapt and change their behaviors. This means that they’ll have to put some “work” to re-learn. They have to think.

And just like web usability expert Steve Krug’s famous book: Don’t Make Me Think.

Steve Krug's Don't Make Me Think

In 2012, popular link shortening service bit.ly learned it the hard way. After a completely new website and renaming some popular features, the new website angered and confused many of their loyal users (including me).

I literally do not know how to shorten a link on the new bit.ly, nice job guys! — Anthony De Rosa (@AntDeRosa) May 29, 2012
bit.ly flooded with howls of complaint after incomprehensible and catastrophic redesign. http://t.co/lePP5rVY #bit.ly — Roger Ebert (@ebertchicago) May 29, 2012

No matter how big or small your UX changes are, always aim for familiarity in your changes and paths to conversion.

7. You won’t know what really works

If you change absolutely everything on your website, how do you know what really caused the increase or decrease in conversions?

Website Redesign Effect on Conversion Rate

Let’s say your new website, redesigned from scratch as many companies do, went live a month ago and you notice that your sales are down. How is it possible? It’s a new design!Well, let’s look at it this way: everything’s new. You probably have a new layout, new navigation, new images, new headline, new call to actions, new checkout process, and so on… Maybe you realized that some of the site’s new elements would improve conversions, but unfortunately, they’re outweighed by other changes that decrease it, as shown by the above graphic.

But how do you know which ones are causing the decrease? After all, you might have hundreds of new things...

At this point, your only solution will be to start testing everything based on new hypothesis. Yes, from the bottom-up. You won’t be able to apply any learning from your previous tests, remember, this new website has been re-built from scratch.

But what if you absolutely need a major redesign?

As you’ve learned above, redesigns can bring a whole lot of troubles and challenges, but sometimes you just have no choice.

Let’s say your bounce rate is always 65%+, your conversion rates are all under 1%, the site looks old and outdated, and the experience is confusing... How can you aim to improve a specific goal when all of them are underperforming and affect each other? When small improvements and day-to-day optimization is just not enough?

It can be very tricky. If the site has a bunch of problems at once and makes optimizing for the improvement of one goal almost impossible, you’ve got a sign that a redesign is probably the best option.

“But, didn’t you say I shouldn't do a complete overhaul?” you might ask.

Yes and no. The traditional process of a website redesign looks like this (too often, unfortunately):

  1. You find inspirations of sexy web designs
  2. Brainstorm how you want the site to look
  3. Make a wish list of “features” you want to add
  4. Create content
  5. New website is wireframed and built
  6. New website goes live (and possibly performs worst than its predecessor)

The problem with this is that you completely ignored what actually worked on your website, and what improved it during previous optimization work.

You might set goals around what you want to achieve around your site, the messages you want to convey, the audience who you want to appeal, and so on… Making it look like you’ve done your homework.

The truth is, ignoring past learnings and improvements, how your visitors behaved on the site, and failing to create an exhaustive list of data-driven hypothesis won’t make the new site better.

The missing stage? Yep, deep data analysis.

A Successful Approach: Planning and Testing

A successful website redesign is definitely not impossible.

Although some designers will find it less fun as they can’t go all out on the new stuff, we can agree we’d rather make more money than winning web design awards for pretty, but unprofitable designs.

Pretty but unusable website
It's a pretty website... but how do I navigate? What are you about? What am I supposed to do? Where do I start? I'm confused! "clicks browser back button

In a recent article, Adobe explained how they approach their website redesigns and avoid falling in traps like the ones I described above. No only do they say they always start by planning their redesigns 6 to 12 months ahead of the expected launch, but they also mention “testing like crazy” is key.

“We test every little thing that’s working and not working now as well as every little thing we’d like to incorporate in the future.” - John Fuhriman, Adobe

Analyzing your best conversion paths and understanding how your visitors converted on the previous design (even if conversions were low) will give you key insights on what to slightly re-build and what to completely redo.

Evolutionary Web Design: The Solution to Avoiding Redesigns Gone Wrong

Redesigning using evolutionary web design is the complete opposite of handing the redesign file to your graphic designer or creative director.

At it’s core, evolutionary web design is a simple concept.

As Chris Goward, a well-known advocate of evolutionary web redesign puts it:

“Rather than relying on the gut-feeling and potentially flawed intuition of an art director, your website decisions should be made against the crucible of customer actions. Once you’ve defined your website’s goals clearly, you can test and continuously optimize to improve on them.”

In other words, marketers are the ones creating the site, and then designers will assist on building it. Not the other way around.You see, evolutionary website redesign is logical: You keep your best performers, try to improve them by emphasizing on what made them work (which you can discover through testing and analysis), and scrap or improve what doesn’t work.

Also easily forgotten but always a key aspect to evolutionary website redesign is knowing your website goals: your KPIs, target metrics, key segments, and so on. Create a solid analytics measurement plan to map out what you want to achieve and how you’re going to measure success. By focusing on improving these goals, it will be easier to create logical hypotheses which you can then test. Always, always be testing.  

Google Analytics Measurement Plan Framework

The ideal structure of an analytics measurement planThink of sites like Amazon and Google. They’ve never had full major redesigns, and never has their user experience and familiarity been hugely compromised, yet, they have changed and improved many things over the years.

Evolution of Google and Amazon's Website

Google is famous for going as far as testing which shade of blue performs better (warning: don’t do this unless you’re Google, you have much better things to test. Read this post I wrote a few months back on creating successful A/B split tests).

As a side note, in case you’re overthinking how you’re going to test your changes, remember that A/B testing is still the preferred method. If you have a large site with a considerable amount of traffic, consider performing multivariate tests in order to identify what combination of changes produces the best results.So finally, what’s the difference between evolutionary website redesign and regular website optimization and testing?With evolutionary website redesign, you’ll often be testing big, and important changes. Whereas day-to-day optimization will more likely involve smaller tests and changes such as headlines, copy, images and forms.


Redesigning a website is hard work. Some companies have set cycles where every 3 or 4 years they’ll redesign to “stay with the trends”. The file is then handed to a team of designers to paint their canvas; user experience and visual design being their biggest focus. The process is long, costly, and once live, might generate less sales than the previous version.

It’s now time for companies companies to stop thinking of redesigns as “that thing we do once in a while hoping it’ll make our business better”. This article makes it clear: this approach probably won’t work.

Evolutionary web redesign, on the other hand, where data-driven decisions trumps visual artistry, and where gradual improvements through continuous testing rules the game, companies are able to achieve their goals much more efficiently without risking their sales, cash flow and months of work.

This is not guesswork, and no gut feelings are involved. It’s science, and you better get on it.

Insights from Splitbase

Sign up to get short weekly emails about ecommerce CRO.

Get emails about conversion optimization, user experience, A/B testing & ecommerce growth each week. We'll keep them short & to the point.

Thank you! Be on a lookout for an email from us!
Oops! Something went wrong while submitting the form.