How to reduce your bounce rate by over 40% (or at least how we did it!)

So, in case you didn’t notice, we’ve just released pretty radical looking new version of WPMU DEV.

In fact, it was so radical that even the prospect of AB testing it had to be thrown out of the window, it’d be like apples and oranges, and so we took a deep breath, popped her out of the door and waited to see what happened… gulp.

And here’s your first taste of the net result, an utterly staggering decrease in the bounce rate on our ‘projects’ pages – i.e. the ones which showcase each individual plugin or theme . Check it:

That's an over 40% decrease in our bounce rate, want to know how we did it? Thought so :)

So, here’s what we did, first up check out the old project page:

And now, here’s the new one:

You can also check out the real thing here, but I include the images for posterity (and because we’ll doubtless be tweaking away very, very soon).

So, here’s what we changed:

  • Rather than competing right and left stuff, it’s straight down the middle, taking the visitor down to the main content far more easily
  • It’s much more harmonious, while there are still elemnets of color they are much more restrained
  • The sidebar is a lot more clear & simple in the new design, much less confusion or users likely to throw up their hands and give up
  • The blue is, lets face it, far nicer than the red
  • The javascript popup for screenshots is also much more clickable, and enticing
  • And last, but not least, the use of tabs is something that I think folk (especially WP plugin folk) really like
  • And finally, right down at the bottom of the page, we fixed up ‘related plugins/themes’ at the bottom of each page
Genuinely related WPMU DEV offerings

 

How do you think we did this, well, it’s a little embarrassing…

There was no algorithm, no smart tech, nothing fancy at all… in fact Sue and myself simply spent a day going through everything we offer and manually adding in related work.

It was boring and hard work, but it saved us a good week or two on development, and made sure that we had better results than pretty much any tech we had developer would have managed – classic startup, love it :)

So, there ya go, if you can spot any other significant reasons why the bounce rate might have dropped so suddenly we’d love to hear your thoughts, otherwise, happy tweaking.

 

Comments (18)

  1. Thats cool. Yup I think your points are valid but also the new site looks way more professional in my humble opinion and so is inspiring more confidence in your products. Its like I always say “Your website is your online shop front. You wouldnt go into or buy from a physical shop that looks dodgy and its the same online”. Do people listen?…..Sometimes.

    I have to say your new site has inspired me somewhat with one of my own. Stay tuned, I will annouce it on the forums on WPMU Dev for WPMU Devers critique.

  2. I hate to throw cold water on your argument but I think it’s a little early to draw conclusions about the bounce rate. When you make a radical change to your UI, everyone’s first response is going to be to have a look around. It will be interesting to see where your bounce rate settles in after the curiosity wanes.

    I’m sure it will go down based on the good changes you’ve made but probably not as much as you’re currently seeing. At the moment, I think you’re seeing a curiosity bounce.

  3. Nice self – back slap James. I don’t blame you. But Jon is probably correct. All the current members are looking around to find out where things are and how to navigate the new look. Don’t get too pumped up too quick. Once people know where things are, you’re likely to find the trend back the other direction.

    The ratio I would be interested in is “Conversion Rate” off these styled pages. How well are you doing getting new members compared to the old style? Are you doing better? Page layouts matter a lot when getting subscribers and new members. I’d love to hear how it’s working. There is a wealth of study out there indicating that focused, non-divided, attention produces better conversions. If your single-column above the fold look fares better than your two block, text & graphic, that is a very valuable bit of information.

    • Ahhh, the old conversion rate :) Yes, that’s the metric I’m most interested in too, and definitely one I would like to share, maybe, down the track…

      Right now it’s a combination of too confidential (although I’m not sure why), too complex (to many factors) and too much focused on effective marketing and improving said rate to stop to write it all down.

      Don’t get me wrong, I *am* tempted and I think that it’d be amazing to really write about and test some assumptions on, as well as get the feedback from the community, but maybe not just right now :)

  4. Hello All,

    Nice new design and I like the blue!

    With regard to bounce rate…

    James, does your analytics program distinguish robot “bounce” visits from actual people visits?

    I ask because:

    1) if you do, actual people certainly may be engaged longer by visual elements and that supports your contentions regarding the effects design.

    2) if humans are not distinguished from bots, then there might be even more to be learned from your results. For example, what about the changes might have required bots to stay a bit longer on the site? More new stuff for them to index before they leave? etc.

    I am curious because some of our sites have very high bounce rates, but they also have tons of bot visits compared to real people and it seems that all those short bot visits tend to run up the calculated bounce rate.

    Great job on the site redesign… now if I can just ever get any traction on the support forums for my 2 needed plugin tweaks and 1 proposed new plugin then MY life will be GREAT! :)

    Phil D

    • Hey Phil,

      Yes, no robots included, people only (Google analytics) – moved on a long way since awstats :D

      And we’re really looking forward to turning our attention to a lot of other non-internal development work now too, expect fireworks for the second half of the year!

  5. I really the look of certain parts of your old site. The problem was I never was quite sure where I was because it seemed like there were two websites in one.
    I do not care for the one long layout and the all dark blue. I feel you need to add back a little color and white. The other site was too busy but this one is the opposite to me. I think there is a compromise in the middle.
    I do not think I would of bookmarked this site like I did the old one.
    (this was when I first stumbled upon you, before I knew how much you offer and all the great info). So that would of been a loss for me!

    • For the record I also preferred the old one. The long bounce rate could be that, like me, people are just trying to find the stuff we used to visit the site for! Don’t like the dark blue, or the amount of scrolling required to navigate. Yes the old one needed a face lift, but this doesn’t work any better than the old one for me.

      wpmu.org on the other hand is much better in terms of design and usability.

  6. Love the new design as far as visual experience. The UIX is taking me some time to adjust to. Better color palette.. Better design.. UI will probably be tweaked doen the road. It’s not bad.. just different.

  7. Last month I made a couple of significant changes to one of my websites. It just took a change of my overall perspective. From one day to the next I noticed in Google analytics that my bounce rate had dropped from 66% on average to 0,63%!

    In my opinion there is only one golden rule which applies to everything and anything you do on your website.
    Look at it through the eyes of a site visitor, create eyeballs, don’t clutter with too much text on landing pages, get them in and let them move on, show them how to immediately (righthand corner above the fold), focus on creating visuals that lead people to explore. Nobody leaves my website after visiting at least 4 pages.

    If anyone would like me to post the website I’d be happy to.

Participate