Search Engine Optimization split testing follows the same principle as AB testing that you may have read about in one of our other blogs – How To Conduct A/B Testing. The key here is to perform tests on similar pages on your website, specifically pages with similar content, and create slight SEO differences between the pages to see which performs better.
This is incredibly helpful in creating a robust and well-optimized website for search engines to rank accordingly.
We have preached many times before that digital marketing is not an exact science, and that what might work for one company may not work exactly the same for your own. The reason for this is that even though two companies that operate in the same industry may seem similar enough, there are many differences when you consider their target audiences, setup of their marketing efforts, branding, and many more variables.
How To Split Test SEO Efforts
Well as hinted at above, you would need to start by grouping separate pages on your website so that you have comparable pages in order to test against each other. You would then create subtle SEO changes on each group and monitor the performance of these pages over time to see which changes help rank your website better in SERPs.
It is important to take note that it is wise to make a single change to one group at a time. This will make it much easier to monitor performance of a single SEO metric such as the performance of a keyword or Click Through Rate. If you make too many changes at once it will be difficult to determine which change is responsible for the great improvements in organic traffic you may be noticing! Similarly, if you are constantly making changes to both groups that you are testing against, it will be difficult to use either group as a “control” where the performance of your SEO changes can be measured against.
So with that out of the way, let’s look at some of the things that you can split test in your SEO efforts!
This is a natural starting point as selecting the right keywords is often a trial and error experience which may require a couple of attempts before you get this completely dialed in.
But it isn’t always as simple as changing your keywords from bike to bicycle. There are multiple different types of keywords that you will need to consider and play around with in order to find out what attracts your audience best. Here are two keyword types we think are important for split testing:
- Branded Keywords – These keywords are used specifically to highlight your brand, meaning that if you are reporting on Amazon’s website you will use Amazon or Amazon Affiliate as one of your branded keywords. This form of keyword usually works well on larger landing pages of a website, such as your homepage. Test out different combinations for these keywords and see which works best!
- Generic Keywords – Sticking with our Amazon example, a generic keyword could take the form of camera, or 4K TV. These are keywords that someone may help a Google searcher discover your business and more specifically your product/service/content. Playing around with different wording and phrases can help you optimize these keywords on your website. Testing this type of keyword can also tell you a lot about what viewers are interested, and not interested, in seeing from your website!
Meta Title & Meta Description
Similar to what we mentioned with keywords, the Meta Title and Meta Description can play a huge role in whether your website ranks highly in searches and whether your landing pages even get clicked on.
The Meta Title in particular is ultra important as often this is all a viewer will look at in deciding which link to click on from their search. Testing various combinations of titles for your landing pages and including different keywords in the title can help you find out which works best for your website.
Similarly, with the Meta Descriptions for these pages, you will need to test to see which single keywords and long-tail keywords work best for your landing pages. This becomes increasingly important if you would like pages from your website to appear as featured snippets, which of course you want them to!
Test out different Titles and Descriptions and see what works best.
Where the other changes we have mentioned can be quite simple, easy to change, and can be changed fairly quickly, adjusting page content can make it feel like you are starting from scratch and rethinking your whole website strategy!
What does this change look like then? You could start by testing out whether creating a shorter, more-to-the-point landing page is more beneficial to your website session duration, and likewise with using various types of content such as including images, videos and inter-page links.
All of these page content variables can have an effect on the experience of your viewer and ultimately how high your landing pages rank.
I hope SEO Split testing is starting to seem much more manageable, and that you are starting to plan how you can implement this in your own SEO strategy, but before you do, there are a couple more things you should take note of.
Firstly, it is incredibly NB that you give each change time to index and perform. These changes should be monitored over a matter of weeks, if not months before you decide to update your website with the best result and then test again.
Next up, before you even make any changes to your website, you should decide which metric you will be using to evaluate the success of these changes. I mentioned ranking higher in search results as the metric that you could be looking to focus on just because it is what digital marketers will often be using as their benchmark to check progress on. However there are many more metrics such as CTR, website sessions duration, new users etc that can be monitored in order to track progress on a more granular level.
An added bonus of monitoring more granular metrics is that often these will result in a higher chance that your pages will obtain better SERP rankings. Another bonus, depending on how you look at it, is that some of your split testing changes may improve one metric, but drastically worsen another and therefore equalize any improvement in ranking you may have expected. Being able to monitor multiple specific metrics will help you make better decisions about what to test and what to implement on your website site-wide!
Nothing Standing In your Way
SEO can be an incredibly scary term in digital marketing but by breaking it up into small bite-size chunks and testing to see what works for you and what doesn’t, means that you are giving yourself the best chance of success!