How to improve your website conversion rate
Did you know that even the smallest website can use testing to find out which page layouts convert best? If you have an ecommerce site and want to find out if a change to a page improves or decreases sales then Google have made it really easy.
You’ve probably from time to time noticed changes on large sites like eBay which disappear if you refresh your cookies. Doubtless you’ll have seen announcements saying that they’re running tests for a proportion of users. They then measure the results of both pages to determine which works best before migrating everyone to the best performing page.
Google Analytics now gives you the ability to do this on your website with “Content Experiments”. You create two or more versions of the same page, tell Google what percentage of visitors will see each version and specify what a successful action will be (e.g. add to shopping cart).
Google will then display the different versions of the page to different visitors and will track the results. Over time they’ll build up a report to let you know which is the best converting page from which time you can then make this the default view.
Content experiments can quickly make a big difference to your conversion rates. Do you know whether “50% off” or “Half Price” would result in more sales? Google Content Experiments can tell you. Sometimes something as simple as the colour of the “Add to shopping cart” button can make a big difference and again this is something you can test to get the empirical answer.
Stop making changes to your website that you “think” will result in more sales. Start testing and then implementing changes that you “know” will perform.
Mark says
12:25 pm on 08/07/2012
Deleted comment: I watched the video and now it makes sense
Chris Dawson says
12:29 pm on July 8th, 2012
You create duplicate pages with the difference you want to test. You tell Google the URLs of the pages and on the original page you add a few lines of code that Google Content Experiments generate.
Google’s code then directs a proportion of site visitors to the alternative pages and measures the results of the action you specify.
It’s an ideal experiment if you for instance create a landing page for a specific promotion and want to see which variation is best.
whirly says
7:19 pm on July 8th, 2012
Good heads up Chris, thanks.
st georges dragon says
7:18 pm on 08/07/2012
I watched the video and felt gormless,
st georges dragon says
8:29 am on July 9th, 2012
the dementia must be taking hold as we always thought sales and money in the bank was the best indication of a sucessful stategy
Digby Geen says
12:07 am on 09/07/2012
Surely by now webmasters should know what works the best.
Eg the best colour scheme
50% off ?
Free Shipping.
ebuyerfb says
9:00 am on July 9th, 2012
Not at all. A/B testing like this is extremely valuable and very relevant especially today. I could have sworn though that Google Analytics has had this feature for years. Probably just came out of beta.
Chris Dawson says
9:05 am on July 9th, 2012
You’re thinking of Google Website Optimizer which is about to be retired.
Google Content Experiments is new and built into Analytics.
ebuyerfb says
9:24 am on July 9th, 2012
That’s what I was thinking of.
st georges dragon says
10:06 am on July 9th, 2012
exactly .actual selling seems to have become secondary to data and graphs etc pages hits and ranking appears to be more important than sales
Chris Dawson says
11:35 am on July 9th, 2012
I’m with you on this – sales are what counts.
However if I looked at your website and say if you changed the colour of your add to shopping cart button from blue to red it would double sales would you be interested?
That’s what Google Content Experiments can do for you.
Not using it won’t mean you won’t be profitable. However for those with time to tinker it can make a real difference.
st georges dragon says
2:47 pm on July 9th, 2012
we dont get time to breath, these days,
wish it were the old days when even working out how to load up a picture gave you a buzz