Content Experiments in Google Analytics
Web Optimiser has been renamed to Content Experiments and has now been integrated with Google Analytics as of the 1st August 2012.
This can now be found in Google Analytics under Content > Experiments. If you are still running tests in Web Optimiser it’s recommended to switch the Web Optimiser tags to the new Content Experiments tags.
What Are Content Experiments
For people who are unfamiliar with Content Experiments and what Content Experiments can do, now’s the time to take a look. In short, they are a great way to test landing pages on your website. You can run up to 6 different variations of a page and can set up tests on all type of pages, landing pages and goal pages.
There are plenty of things to test to increase conversions. For example, you might want to test one USP against another or include the VAT at the beginning of the checkout instead of the end.
How To Run Tests
When running any test you need to establish your goals. In this case, the goal was to increase sales but not increase the marketing budget. The best way to do so was to take a look at the Analytics data. If one product stands out, for example, having had 200 visits but zero sales, you might want to start by looking into this.
Do some research on the product first before running a test to see what to assess. If you are competitive on that product, try test pages using a CTA such as ‘free delivery’ or ‘was £29.99 now £19.99’ or any other special offers.
An increase in conversions is an increase in profits. Why would you NOT be running tests?
How To Use Content Experiments
Content Experiments works by having two variations of a page, the original page and the test page, and Google then uses third party cookies to run tests on these two pages.
The process is that one person will see one variation of the page and another will see the other variation of the page and if you return to the site you will be shown the same page again. All the data is stored in Google Analytics so you can see which of your internet marketing experiments are working or not as the case may be.
