Try our conversational search powered by Generative AI!

The problem of A/B testing content when the conversion landing page is the same content


Say I have a page called "Products," and I want to test a title change to see if I can get people to navigate to it more often. The page appears in a couple places around the site, all listing the Name property from the page itself.

Example: the Products page is linked from (1) the overhead menu as an iteration of the top level of pages from the tree, and (2) in the footer, as a Link Collection. In both cases, the link is formed by simply using the name of the page itself. So the name of the page hasn't been entered "somewhere else," the HTML is being rendered by iterating some collection of pages and outputting the Name property.

The Key Point: The only way to change the text of the link is to change the name of the page itself. (Yes, you could have separate properties for "Overhead Nav Title" and "Footer Title," but you'd still have to change a property on the page to make the link text change.)

So, I edit the Products page and change the name to "Solutions," then select "A/B Test Changes."

Since I want to test if more people will click on the page if it's named "Solutions," I pick the same page as the conversion page.

Spoiler: this won't work. You cannot change a page, then select that same page as the conversion landing page.

So, how can I test my title change? The only thing I can think of if that all appearances of my page that I want to be testable will have to be...something else. I would have to set up my overhead nav and footer as separate content elements that I can change (like a Link Collection with manually-added links that can have title changes), then initiate A/B tests on those, selecting my Products page as the conversion landing page.

Is there a solution here that I'm not considering?

If not, then I'm thinking that the desire to perform A/B tests like this needs to be planned for and architected on the original build of the site. If you want to test navigation changes like this, you need to make sure all the navigation is rendered from a logically separate entity than the content to which it links. If you do this, then you can change the "nav content" (whatever that ends up being), launch an A/B test on that content, then select the target content as the conversion landing page.

Sadly, this is something that (1) you wouldn't likely think of doing for an otherwise normal site build, and (2) isn't trivial to retrofit after the fact.

Edited, Jun 19, 2017 21:21

How about creating a new page and use the Shortcut property and set it to a shortcut to the original Products page? Then you can use the new page as the conversion landing page, and the users will still end up in the correct place.

Edit: Haven't tested this though

Edited, Jun 20, 2017 10:50

There are a couple of ways to solve this issue.

First you could use the out of the box Time On Page conversion goal that will trigger after a short duration such as one or five seconds, depending on if you want to account for possible misclicks landing on your page. So someone loads the page, then they stay for the duration and then the page will convert automatically without needing another request for a landing page.

Second, you could develop a custom KPI (custom conversion goal) that triggers when the page itself is loaded. This is somewhat tricky because you have to make sure that any load of the page doesn't trigger a conversion, such as the call to load the content for the title in the navigation.

Hope this helps.

Jun 20, 2017 16:27
* You are NOT allowed to include any hyperlinks in the post because your account hasn't associated to your company. User profile should be updated.