Follow these seven stages to get started with an SEO test:
Form your hypothesis
A hypothesis is a guess about what will happen. It’s when you decide on what you’ll change and how you believe it will improve your website’s SEO ranking.
Here’s a basic hypothesis formula:
[change] leads to [effect] on [types of pages]
When performing this type of test, make sure to pick a set of pages with a lot in common—like blog entries or e-commerce category pages—to reduce confounding variables. Your hypothesis should be based on knowledge rather than chance.
Choose your pages carefully
It’s recommended that you only run tests on sites with a lot of organic traffic. It’s because a page that receives little or no organic traffic will provide you with very little information. It will simply distort your test findings.
In Google Analytics, you can discover websites with decent organic traffic:
- On analytics, go to Behaviour > Site Content > Landing pages
- Choose to only filter by organic traffic
- For the pages, you want to test, use a filter. (e.g., pages with /blog/ in the URL)
- It will show only pages with at least [x] sessions.
Pull a random sample
You’ll be doing an AB split test, so all you need to do is modify a sample of pages. The remaining pages will serve as your control group.
If you’re employing a specialized SEO testing tool, it may handle random sampling for you. If you’re doing things the old-fashioned way, though, Google Sheets’ “Randomise range” function is the easiest method to choose a random sample.
Here’s how to go about it:
- Import your data (URLs, etc.) into Google Sheets.
- Highlight everything
- Randomise the range by right-clicking
It replaces the percentage percent sign with a random number, making it possible to use a portion of the top URLs as your variant group.
Decide on the test duration
It isn’t easy to know how long to run a test, but it should take as long as it takes to gather enough data. This might last anything from a few days to a few months. It depends on how much traffic you have and how many pages you have.
Some extra tips:
- Run the test for long enough for Google to re-crawl the pages in your variant group. Unless the sites have been re-crawled, you won’t be able to attribute any fluctuations in traffic or ranks to your modifications.
- Run the test for a long enough period to obtain decent traffic to your pages. Don’t make any judgments until your pages have received significant traffic, as they’ll most likely be statistically insignificant.
If you’re unsure how long to keep the test running, I propose beginning with a month. If you need more data, you may extend the test indefinitely.
Decide how to track test results
You need a method for tracking results before making adjustments to your website. The most straightforward approach to accomplish this is to utilise SEO testing software. It connects to your GSC account and keeps track of the outcomes for you.
However, you may log results if you wish.
Use Google Search Console or Google Analytics to check changes in organic traffic. Compare the average variations in traffic for both groups for the period before and after the adjustments were implemented.
Use GSC to check the average CTR changes.
Google Search Console and rank tracking tools are options for monitoring changes in organic ranks. The former is unquestionably the superior alternative since it gives you actual ranking positions, not just averages.
Make the changes
It’s now time to put your AB test into action. Here are some helpful hints for completing this:
- Don’t make any modifications to the control group’s pages. The term “control group” implies it; you’ll compare the effects of your modification to its baseline.
- Change one variable at a time. If you make numerous adjustments, determine whether the result was due to a single or a combination of factors.
- Keep a note of any modifications you made, as well as the date on which you completed them.
- Ensure you have a method to return if things don’t turn out as expected. You’ll need to do this if your SEO test pulls up negative or inconclusive results.
- Don’t waste your time with pointless activities. If you can and quickly apply an SEO best practice, don’t bother testing it—do it.