The Truth about A/B Testing
A/B Testing is widely used in SEO for assessing and improving conversion rates. This testing procedure has raised several questions and concerns, ever since it was first introduced. Experts at a San Diego SEO agency offer a few details about A/B testing which every online marketer needs to know.
Benefits of the Test
There are a number of benefits A/B testing offers. It is widely used by SEO experts in conversion rate optimizations (CRO)- a process that improves the percentage of people visiting a particular webpage and taking the desired decision.
Shift in Investment
If you can improve your conversion rates through A/B testing, you can invest in other areas of your business (such as finance and operations). In an A/B experiment by Pinterest, they found out that investing in text descriptions, through visual analysis, played a vital role in improving CRO.
Avoid Bad Decisions
Instead of blindly rolling out changes, A/B testing allows you to add a limited number of changes, observe the impact on traffic and then roll out all other changes. Whenever you witness a significant decline in web traffic, you can simply run A/B test and determine bottlenecks.
How to do an A/B Test
Traditionally, A/B testing has involved creating two web pages with minor differences and redirecting traffic equally between pages. However, there is only one Googlebot now and it penalizes websites with similar/near-similar pages. So this approach may not be as effective as it once was.
Instead of creating different groups of users by creating a single, similar page, it is better to use A/B testing to create different groups of pages. This method is safer as only one version of page is designed and used and shown to Google’s crawlers. It is also very effective.
The general A/B testing process involves the following steps,
- Identify all the pages you would like to improve
- Select the test to run across the pages selected
- Randomly regroup the pages into variant and control groups
- If the variant group exceeds its forecasted expectations and the control group does not, then your A/B test is a success
Length of Test
Google has, over the course of time, become more rational in its approach to redirect traffic to websites. To determine the length of you’re a/B test, you will have to decide on your approach. If you want to verify that the tests results are positive, then a pragmatic approach is advisable. Once the results come in, you can roll out the changes immediately.
When determining the success or failure of a particular test, it is best to rely on organic search traffic as the best metric for measurement. Organic traffic can help in improving your website rankings and is easy to detect.
Practically, it is difficult to identify and track long keywords. Some changes can help improve the overall click though rate, without improving your ranking position. Similarly, you can set up A/B tests to measure the improvement in your total conversions– between the different groups of pages. However, this test may converge too slowly on many sites.