The Small Business Introduction to A/B Split Testing

I don’t know how many gyms have used this phrase in their advertising, but just because it isn’t original doesn’t make it irrelevant: Summer bodies are built in winter. Before COVID, this was a much less complicated concept. Lots of people join gyms during winter than at any other time, particularly right after the holidays.

Make your SEO content writing count

In our previous post we discussed content management for small- and medium-sized websites. With a spreadsheet and Google Analytics you can keep a bird’s eye view on your content, and make more accurate decisions in terms of the content your visitors will love. At some point, however, your SEO content writing service, or person, may produce two versions of the same page – both of which look like they’re capable of doing a good job. How do you choose between them? Instead of spending hours pouring over both versions, rather use Google Analytics to A/B split test each one.

A/B what?

This post will provide you with a brief introduction to A/B split testing, and links to Google Analytics pages where you’ll be able to learn more, and start your own content experiments.

In most cases you’ll read about A/B split tests that have improved conversion rates from 0.45% to 1.17% – small figures ostensibly meaningless to anyone not familiar with the context or industry they’re used in.

But when your revenue increases by 114%, it’s another story. The entire case study is rather technical, but, essentially, hosting and monitoring company Server Density, wanted to see what type of pricing structure would prove more effective: one where audiences are able to choose what they pay (e.g. $1 for 1 unit, $2 for 2 units, etc.), or a packaged deal (e.g. $5 for 7 units, $10 for 15 units, etc.). Although the packaged deals were lower in price, they were higher in Average Order Value. After running the test, Server Density found that their revenue increased from $394 to $833 with the packaged deals.

The point in the example above is not increased revenue, but the fact that Server Density never would have known, had they not run the test.

So what is A/B split testing? Let’s say you have a page with a heading, "Lose 5 pounds in 3 days from your couch", and your SEO content writing service suggests "Lose five pounds in 72 hrs" as an alternative. Tough choice – you’ve had good results with the current version, but there’s a good chance audiences will respond equally well, or even better, to the suggested title. A/B split testing lets you compare two versions of the same page against each other with actual visitor data. In other words, both versions are live on your website, and are alternately served to visitors as they arrive.

Preliminaries

Sound like something you could use? Of course it is – every website owner can benefit from A/B split testing because, even though we know our audiences, they still manage to surprise us again and again.

But jumping into conversion testing with your guns blazing and no prior planning isn’t a good idea. Many companies make the mistake of testing for the sake of testing, instead of considering what their audiences will respond to. Start by analyzing your Google Analytics data – that’s our recommendation. If you see you get a fair number of visitors to a certain product page, or a page that leads to a product page, but the click thru rate is almost non-existent, there must be something on that page that doesn’t work well – perhaps just some copy that your content writer has to fix?

Perhaps. But don’t guess. Experts suggest combining quantitative and qualitative data to help make that important decision. In other words, you’ll have to get in touch with members of your audience and ask their opinion, too. Perhaps it’s the copy, or maybe they just don’t like the colour of a button. Polls, email questionnaires, campaigns that incentivize response, or even phone calls, are all valid methods of obtaining customer opinion (qualitative data).

Start testing

As said previously, we won’t go into the details of actual testing here. There are a lot of software packages designed specifically for that aim, and they all work differently. But if you’re looking for a free tool, we recommend that you start on Google Analytics’s Move from Website Optimizer to Content Experiments page in case you’re using Website Optimizer, and then move on to the Overview of Content Experiments page, where you’ll be able to learn more about the tests, and set them up. The end result is that you’ll be in a position to make an informed choice between two versions of the same page produced by your SEO content writer, choose the best heading, or change the colour of a button – all with the aim of increasing your conversions.

Google