*** Please HELP *** A/B tests and optimisation implications

[CommT]

Hi Mozzers,

We've been A-B testing landing pages, and have had some success.

The changes we've been making have been quite radical in some instances - for example we tested this page: https://www.turnkeymortgages.co.uk/todays-mortgage-deals/ against this one: https://www.turnkeymortgages.co.uk/mortgage-quote/ (Today's best deals won, but we've decided to keep the quote page as it does work for some channels). The decision was made to try and optimise Today's best deals for 'best mortgage deals today' rather than 'mortgage quote' because it offers so much more than simply a quote. The quote page is optimised for 'mortgage quote', though it doesn't rank particularly well (I'm not overly concerned by this as even though you'd think that when people are looking for a quote that they would fill in the form, they don't - people are strange!)

As a result of the change above we changed all links that originally went to the quote page to go to Today's best deals instead. As we go through the process of optimising for best conversion will it be damaging if we don't change the url as well. As I can see lots of iterations and lots of work whenever we make changes to the pages (going through the entire site to change the links).

I am worried though that we'll end up with hundreds of landing pages and changing links all over the site - do you think we should keep the URLs the same from now on, unless the content changes as radically as it did in the instance I've highlighted above?

Thanks,

Amelia

[LynnPatchett]

Hi Amelia,

Conversion optimization does necessarily need different urls, so I would say that no you do not need to change url unless as you say there is a major change in the topic or SEO optimization of the page that might dictate a change in url also. It depends a bit on how you are doing your A/B testing - are you using software like optimizely or just measuring different pages in analytics? If the latter then maybe your situation dictates separate urls to test and you are bringing in visitors from CPC campaigns (for example) to test - once a decision is made on which version works best you can update the main page that is linked to from your site - keeping the same url of course!

[CommT]

Hi,

Thank you for responding. I am using Google analytics experiments. You put some code on the original page and 'magically' the two pages get rotated to show approximately half the visitors one version, and the other half the other version. The new version is often called 'page-name-b' (for example) and if it wins then the content moved to the original version. The only exception to this rule was when we used a page called 'todays-mortgage-deals' - because I wanted to try and get it to list for more than just 'mortgage quote'.

Best wishes, and thank you again for responding

Amelia

[LynnPatchett]

Hi Amelia,

So in this case you are fine doing it as you are doing and not changing the 'final' url but just shifting the content as you describe after determining a best case. Have fun testing!

[CommT]

Thank you, much appreciated