Never Redesign Your Website Without This Strategy!

by Chris Goward

Last updated on July 11th, 2018

Since the dawn of the web age, companies have been trapped in the website redesign cycle.

Every five years or so, a CXO decides the site design is looking “tired” and needs to be re-thought. Often the company pays little consideration to whether the new design will improve results. “It’s new,” they’ll think, “so it must be better, right!?”

Many have been fooled into believing that a website redesign will improve conversions and revenue. It’s common to assume that a slick new design that follows the accepted “best practices” of the day will increase customer trust and your sales rates.

That is a false hope.

And that’s not the only problem with website redesign projects.

The Risky Website Redesign Approach

Unfortunately, the creative process used by most agencies and marketing departments don’t consider risk mitigation.

To understand your risk exposure, think about the number of individual changes that are made during a redesign. Multiply that by the depth of change for each element. Imagine for a moment the laundry list of changes proposed during a creative meeting.

You’ll change the home page headline, imagery, site-wide template layout, navigation bar design, fonts, shopping cart or form layouts, and many more.

When do you discuss the risk of all these changes? Maybe some of those changes help conversions, and some probably hurt. How do you know which have positive or negative effect?

Website redesign masks the effects

In most cases that discussion about risk-mitigation doesn’t happen.

Marketers usually go into a redesign without a process in place to test the page templates and landing pages that are being changed. There’s no system put in place to monitor and justify those changes against key conversion metrics.

Mitigating these risks can only be done with a rigorous conversion optimization strategy. It requires a process that includes understanding the target audience, prioritizing test hypotheses to solve issues obstructing conversions, setting up controlled split tests, and analyzing insights from data to make informed changes.

Companies that are using a structured process that include A/B/n split testing as part of a conversion optimization strategy, like WineExpress, Iron Mountain, Electronic Arts and, are getting significant sales lift while reducing risk.

In fact, eConsultancy reports that companies with a structured approach for conversion optimization are twice as likely to have seen a large increase in sales as others.

Structured approach doubles conversion optimization success A structured approach doubles conversion optimization success

The Better Approach: Evolutionary Site Redesign

Testing, and a proven system to execute testing for design changes, is critical for today’s online marketer. The risk of making substantial website changes without it is too great not to.

This approach to website redesign is something I call “Evolutionary Site Redesign” or ESR.

The truth is that a dramatic, “revolutionary” redesign is dangerous for most companies.

It’s not that you don’t need a redesign. You probably do. But a better, and less risky, approach involves a process of testing with incremental (and often dramatic) improvements. This ESR approach gives a better visitor experience and results versus the traditional throw-the-baby-out-with-the-bathwater method.

ESR Gives Continuous Results Improvement

There are two major differences between ESR and the traditional “Revolutionary Site Redesign” (RSR) approach:

  1. It’s Faster
    After a traditional website redesign, marketing departments are usually so fed-up with the process that they’ll gladly wait another 5 years before trying again. Or else, they may spend the next 6 months scrambling to fix the conversion rate drop with their new site. ESR, in contrast, creates a system of continuous improvement so your website is always leading the pack. The traditional website redesign cycle of under-performance The traditional website redesign cycle of under-performance
  2. Success criteria
    “Gut feeling” and reliance on the so-called “best practices” of designers and UX practitioners rules in a traditional redesign. While the team may be talented, no batter hits .1000 and many of their changes are likely to hurt website results. With ESR, every change is measured in controlled A/B/n split tests against its effect on business goals.

The RSR approach leaves your website lacking and continuously falling behind in the intervals between major redesigns. But, with ESR, your website will continuously keep up and surpass the success of the rest of the web.

ESR essentially uses conversion rate optimization principles to redesign your site.

How ESR Eliminates Epic Website Redesigns

By adopting the evolutionary site redesign approach you can guard against website redesign risks while dramatically improving your website every day.

ESR continuously improves your website ESR continuously improves your website

With ESR, your website will continuously keep up and surpass the success of the rest of the web. Once you’ve defined your websites goals clearly, you can test and continuously optimize to improve on them.

ESR works by implementing a system of continuous A/B split testing throughout your entire website and digital marketing. Rather than relying on the gut-feeling and flawed intuition of an art director, your website decisions should be made against the crucible of customer actions.

You should test everything in your marketing:

  • Site-wide design styles
  • Logo, header and tagline
  • Product page templates
  • Landing page design & content
  • Your product or service value proposition statements
  • Lead generation forms, shopping cart and checkout
  • Home page design, eyeflow, merchandising
  • Imagery, copywriting, ads, calls to action, and offers
  • And everything in-between!, for example, should be continuously updating their design by testing the major site-wide elements that combine to create their look & feel: header, nav, PCTA, headlines, repeating listing areas, etc. ESR opportunities’s ESR opportunities

At WiderFunnel, we’ve developed a continuous improvement system to drive the ESR process for companies. It leverages great thinkers who have gone before us and adapts the scientific method for website practicalities.

The system ensures that website test iterations are driven by solving customer problems identified in the heuristic LIFT Analysis, voice of customer data and web analytics data (like Crazy Egg click heatmaps), rather than just creative aesthetics.

The Top 5 Reasons to Use ESR

Here’s why ESR is the future of website un-redesign:

  1. You get a new site “look and feel” and conversion rate lift at the same time
  2. You learn which elements actually improve results
  3. You maintain your team’s focus on the important business metrics rather than “aesthetic” redesign
  4. Your website never faces lags in results in-between redesigns
  5. You avoid the risks of a “Revolutionary” site redesign

What do you think?

Have you experienced the post-website-redesign conversion rate drop? What barriers does your organization face to redesign by testing?

You can learn all about the process for conversion optimization and ESR in my new book by Wiley called “You Should Test That!” It’s highly recommended by people who recommend stuff.



Get updates on new articles, webinars and other opportunities:

Chris Goward

founded WiderFunnel, the conversion rate optimization company, with the belief that digital agencies should prove the value they bring. Be sure to check out his bestselling book, "You Should Test That," and follow him on Twitter.


Comment Policy

Please join the conversation! We like long and thoughtful communication.
Abrupt comments and gibberish will not be approved. Please, only use your real name, not your business name or keywords. We rarely allow links in your comment.
Finally, please use your favorite personal social media profile for the website field.


Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  1. Anonymous says:
    June 28, 2016 at 2:28 am

    Very descriptive blog, I liked tthat bit. Will there be a part 2?

  2. Anonymous says:
    December 16, 2015 at 12:16 am

    It is grateful post. Thank you so much for sharing good information

  3. Steve says:
    October 14, 2015 at 10:28 am

    Enjoyed the article, but as an editor, had to say, MANY batters bat .1000 and then never play again! That’s one hit in 10 at-bats. I think you meant 1.000 🙂

  4. debangshu says:
    February 20, 2015 at 3:59 am

    I love the new look! and as a new follower, It was really fun to see the old designs as well. Thanks you for doing what you do. You are awesome!

    • Kathryn Aragon says:
      February 20, 2015 at 9:04 am

      Thanks, Debangshu. It’s great to have you following.

  5. Ravi Janardhan says:
    June 25, 2014 at 7:46 am

    Hi Chris,
    It’s an awesome post, with valid illustrations.
    I’m going to adopt (yours ?) acronyms: ESR, and RSR. The distinction makes so much sense.

    I see ESR is going to be niche area, that’s going to replace website maintenance, occassional redesign & conversation rate optimization, simply because it’s all that combined continuous process and very much needed.

    Thanks again, enjoyed your post!

    • Chris Goward says:
      June 25, 2014 at 8:32 am

      Good to hear, Ravi. Thanks!

    • Neil Patel says:
      June 25, 2014 at 2:32 pm

      Ravi, glad you liked it. Thanks for sharing 🙂

  6. D Thomas says:
    March 3, 2014 at 7:51 am

    Superb article Chris which points out some of the fundamental points and the pitfalls, it’s never an easy choise redesigning and with regards to my own, I try not to do it too often! Thanks for sharing look forward to reading more of your articles!

  7. John Francis says:
    February 25, 2014 at 8:02 pm

    Seems to me that you have missed one of the most important change agents in the web design business – the rapid advancement of technology. It is fine to criticize the 5 year approach. However, my experience with smaller firms is that they are very resistant to redesign; even that rarely.

    In my opinion, the technology changes driven by mobile requirements and other new communication/marketing/sales capabilities will force new designs continually on those companies that want to remain competitive. Therefore some methodology for researching the impact of these technologies and having a process for incorporating and testing those that are successful is more important than puttering with the visual elements. Function – then form.


    • Chris Goward says:
      February 26, 2014 at 10:00 am

      Re: “puttering with the visual elements”

      You’re a funny guy, John. (Though, it might be unintentional, unfortunately.)

      The repeated double and triple digit sales lifts we are delivering with this methodology encourages me to keep “puttering” all day every day. A few more examples:

      Of course, we recommend testing new functionality and we do that for our clients too.

      What you seem to misunderstand, John, is that testing design, copy and imagery does not preclude you from testing functionality, features, marketing strategy and everything else. The evolutionary site redesign strategy was simply the topic for this article.

  8. tom jackson says:
    January 22, 2014 at 9:05 am

    Great article, thanks. My site has very low volume, and the market is seasonal (1/3 of the year). What if my sample size does not allow for statistical significance?

    Also, what can I use to A/B test a WordPress site?

    • Tema Frank says:
      February 19, 2014 at 11:38 am


      In addition to Crazy Egg’s software (which I believe is quite good), there are many other A/B testing services out there that will work on a WordPress site (eg. Optimizely), there are WordPress plugins that do it, and now even Google has a tool you can use for A/B testing, called Google Content Experiments.

      I agree with many of the points in the article about the importance of continuous testing and optimization, but also with some of the other commenters who know that sometimes a site is so bad (and getting so few visitors) that a radical overhaul (with lots of research baked into it) is a better way to start.

  9. Phil King says:
    January 18, 2014 at 3:52 pm

    There’s a lot to agree with here and helpful to have some of the arguments in favor of evolutionary design work documented. I will say that I don’t think that evolution is the only or even the best way to approach redesigns 100% of the time. Combining the two approaches together is probably the optimal strategy, if your team has the capacity and long-term focus to pull it off. Many smaller iterations can absolutely help keep your experiences optimized and somewhat fresh, but it can be a big challenge to pivot towards new opportunities or create really innovative new experiences. I haven’t heard of many organization that have done a great job at both constantly tweaking and also making periodic large changes stay on par or ahead of the market.

    And that’s the thing the charts are really showing – the ability to exceed the state of the art. I don’t think that revolutionary redesigns can only achieve results on par with the industry – in fact I think they have a much larger potential to exceed industry best practices than an evolutionary approach. It’s true that the time between larger redesigns must be larger, but I think they are very much worth considering in a variety of cases – especially when new competition enters your market space, for instance.

    Another thought – it’s very difficult to market smaller incremental changes to users or to the media. Talking about what you’re shipping, and having the media talking about it, is important in some cases both for PR and for recruiting talent (just for instance). Incremental changes can have the desired results in terms of conversion, but tend to not support PR and thus may not help attract new users.

    Basically I think the answer is that both strategies are required and, ideally, should be part of longer term product planning in a high functioning organization. At minimum you should be doing evolutionary work and testing, but the risk of larger redesigns should not prevent teams from creating a design vision and considering the opportunity to disrupt their market.

    • Joel says:
      February 3, 2015 at 2:14 pm

      This is a great article by Chris that brings up a lot of the considerations that companies need to make when considering a redesign. Whether your company approches ESR or RSR, there needs to be a strategy and plan behind it. Too many times companies launch a revolutionary new product that fails (most products fail) because a failure to understand the market and a failure to test the product with the right people.

      Overall a business needs to take into account a whole range of metrics when it comes to optimizing their products for their target audience. RSR from a certain perspective makes sense when there is the opportunity to break into a new market or product. Whereas, some companies that use ESR still fail to understand wide market changes since they are focused on certain numbers.

      For example, let’s take Adobe Creative Cloud which has been the culmination of over 20 years of market domination by a company selling creative software. That software has incrementally improved since its inception but you get the feeling that Adobe CC is beginning to lose a large amount of influence in the interactive and interface fields because it has become an incrementally improved dinosaur. Adobe is a company that needs a revolutionary product if it hopes to continue to compete for the hearts, minds, and dollars of UI designers who are quickly converting to Sketch.

      • Kathryn Aragon says:
        February 3, 2015 at 9:36 pm

        Nice insights, Joel. I completely agree that you need to know and understand your market. And that’s going to influence everything, from your website to the products you create and the improvements you make.

  10. Mike says:
    January 18, 2014 at 5:54 am

    Good article. We need to follow this as we are on the verge of doing a complete site redesign without having done any testing or metrics checks. The word “tired” has been chicked around a lot!

  11. Blair de Jong says:
    January 17, 2014 at 2:16 pm

    Great post Chris! Timely post I will share with my team, as we continue to have this debate internally and struggle with the direction to head for a much needed redesign.

  12. January 17, 2014 at 2:09 am

    Great post, Chris and I completely agree.

    One of my favorites quotes is from Gerry McGovern (I’m paraphrasing) “Successful websites are all about continuous improvement.”

    He wrote a great piece here:

  13. Jay says:
    January 15, 2014 at 7:25 pm

    Small, iterative, calcuated, measurable changes. Absolutely the right approach. Couldn’t agree more. But as always, it’s not always the right solution. For instance, if your site is truly 5 years old, chances are it needs much more than simple changes. Your site could actually be jeopardizing your brand and online marketing efforts, at which point you need to consider the opportunity loss of slugging along with small changes over the course of months vs. a complete redesign.

    All of that said, your evaluation of ESR over RSR doesn’t add up.
    1. It’s Faster. After a complete redesign, you’re at a perfect point to begin incremental updates aka ESR. Even in the most archaic industries, I’ve never once heard of an approach that calls for complete redesigns every 5 years (just for the hell of it) with nothing measured or optimized in between.

    2. Success Criteria. Not sure why you assumed that redesigns are based on gut checks and managed solely by an art director? Even an average agency is performing discovery across their market, learning about customer segments and behaviors, reviewing all available data from pre-existing online channels and then designing a new site based on all of the above. RSR does not equal, “let’s just wing it.”

    It’s a shame you distorted reality so much for some extra book sales and subscriptions to your product. There’s a huge market for conversion rate optimization, regardless of which acronyms you’re adding to the discussion.

    • January 16, 2014 at 8:31 am

      Hey, Jay. I appreciate your thoughts, but can’t help but disagree. You’re probably right about agencies. But I’ve worked with businesses that did exactly what Chris says: redesign regularly just to say they did it–and make changes based on gut feelings and opinions, not testing and not even accepted heuristics. “Distored reality” is a harsh judgment. And, by the way, the book is well worth the investment. There’s no need to try to push extra book sales when it’s likely going to sell itself.

    • January 16, 2014 at 1:22 pm

      Thanks for your comment, Jay! I know I’m getting the message across when the trolls come out to play!

      I could add lots of experiences with typical web redesign agencies making huge website redeisgn investments only to see conversion rate and revenue drops, Jay. It’s interesting in itself that your agency doesn’t believe in your work enough to even design your own website!

      But I digress.

      To answer a few of your questions.
      ▪ Yes, I am promoting an iterative, calculated, measurable approach, but I didn’t say anything about “small.” Often our ESR variations are dramatic redesigns, when that’s needed. A/B testing works for dramatic changes as well as for incremental improvements.
      ▪ If a website has gotten to the point that it’s jeopardizing your brand then, yes, something dramatic is likely needed. But, making that dramatic change without testing can get you in even more trouble. A couple recent public examples:,
      ▪ In contrast, when was the last time Amazon redesigned their website? They don’t ever do redesigns without testing, but they’re also contstantly redesigning the experience.
      ▪ In the system we use, ESR is *faster* than a complete website redesign. Results start immediately upon getting the first, high-impact test live, which can take only a few short weeks. There’s a lot changed in an RSR that doesn’t nothing but take time and add complication.
      ▪ The data you mention in your “average agency” process doesn’t include any conversion rate or revenue-production metrics. Reviewing pre-existing online channels is a fine starting point, but after you launch this beautiful new website, will it *actually* help or hurt business results? And please don’t tell me you compare analytics data after the launch with data from before. That’s a “Pre & Post Test,” which is one of the 13 ways people screw up their conversion optimization efforts. More:
      ▪ Unfortunately, reality from within a traditional agency bubble does seem distorted, but I’m talking about reality. The traditional digital agency emperors are wearing no clothes!
      ▪ Also, selling copies of my book is the least of my worries. You may not be aware of how little authors make from book sales. It’s a negligible part of my business revenue. The purpose of the book is to educate the market so companies seek out what they really need rather than continuing to hire flawed website redesign processes from dinosaur agencies. Also, it gives step-by-step instructions so companies can do it on their own if they want to.

      • Justin Roff-Marsh says:
        May 9, 2014 at 1:22 pm


        I read Jay’s post. It seems clear to me that he’s a critic, not a troll.

        This is an important distinction. If you treat the former like the latter, you start to resemble a d’bag!

        MORE critics, not fewer, is a sign that you’re becomming a thought leader. If I were you, I’d be begging for more.


        • Chris Goward says:
          May 9, 2014 at 2:06 pm

          I disagree, Justin. I actually welcome critics and encourage debate. But, when he questions my motives and throws unfounded personal attacks, it’s no longer a debate.

          He says, “you distorted reality so much for some extra book sales and subscriptions to your product.”

          That’s an unfounded personal accusation. Never mind the fact that I don’t even have a subscription product to sell. And, I really don’t need the $2.50 from an extra book sale. His questions were totally fine up until that point when he revealed himself as a troll.

  14. tamar says:
    January 15, 2014 at 11:53 am

    Great article, we haven’t redisigned the website but we made some changes with our company logo.

Show Me My Heatmap

@CrazyEgg Thanks for no longer truncating naming conventions of exported snapshots. Saves so much time!

Nicole Mintiens


What makes people leave your website?