A/B testing can help you get a leg up on the competition and help you in avoiding mistakes. Watch this week’s Whiteboard Friday with Jandira to review four surprising SEO test results.

Click on the whiteboard image above to open a high-resolution version!

Hi, Moz. My name is Jandira Neto. I'm an SEO testing consultant at SearchPilot, and today, I am here to go through four surprising SEO test results that we have recently run.

If you're not familiar with SearchPilot, we are an A/B testing platform that tests on enterprise sites, such as e-commerce sites, retail sites, and job listing pages. Our mission is quite simple: to prove the value of SEO to the world's biggest sites while testing agile changes and their impact. Some might ask why we test.

So why can I not just look at my before traffic and my after traffic after implementing the change?

• A/B testing is important because, firstly, you can get a one-up on your competition. While your competition is testing the same old strategies, you are continuously evolving, exploring, and trying new things.

• Also, you can avoid mistakes. Some mistakes might cost you and might cost other departments. Therefore, if we test, we can see if something could have had a negative impact on our traffic and roll that back before it does any sort of damage.

• Also, a great thing you can do is get your development pipeline sorted. We've heard from many customers that the development pipeline is extremely long, and the development team does not want to go through all of that. But if you have some robust and some concise data to show, then you can most likely convince them.

• Also, another thing to mention is that testing might work differently for one site and might work differently for another. That's why it's important to test to see what works for you.

So, firstly, let's go through taking content out of an accordion. Google previously has said that hidden content can be crawled and can be indexed. So we're talking about things in tabs. We're talking about accordions. We're talking about anything that requires JavaScript or requires the user to click to find it.

However, we wanted to test the limitations of this. So, on one of our e-commerce sites, we took out the content from a hidden tab, and we saw a 12% uplift on taking it out of the tab, which was quite surprising because if Google says one thing and we're testing another, then that contradicts their advice.

So that's why it's important to give a test and to see what works for your site. It might not work for another.

The next idea I wanted to go through was capitalizing title tags. You might think, unlike our previous idea, that this cannot work. This is too spammy. This is something that Google has set against. Why would I want to see capital letters in the SERPs?

However, you can't deny that it is an effective way to make your text stand out in the SERPs. Or is it too spammy or too small of a change? How can just capitalizing a couple of words make such a dramatic change? But that's why we test.

So we did a test, and we saw an 8.5% uplift in organic traffic when we just capitalized the title tags, which is absolutely amazing and crazy that with such a small change, we saw such a large increase in organic traffic.

The next idea I'd like to explore is removing brand names from title tags. This is another one you might turn your nose up at because why would I want to remove my brand? Google has said before that in the SERPs, they want brand presence and recognition. So, adding your brand name surely should be a way to improve your brand presence and recognition in the SERPs.

But we wanted to test if that is the case, similarly to how we tested taking the hidden content out. If we were to remove the brand name from a title tag like this example, when we tested it on one of our aviation customers, if we were to remove, we should see a negative impact because now there isn't a brand name in there anymore for us to have a look at, for users to see and to recognize.

So, your organic traffic should decline. However, we did see a 16% uplift when we removed the brand name. So that was completely unbelievable. We thought that was really surprising.

But we do want to bear in mind that, like I said previously, this was for our aviation customer. This was for a specific industry. Not all industries might see an uplift like this. So, that's why it's important to test. That's why it's important to see if that could work for you.

The final idea is article Schema. Google has said before that they love structured data. It helps them understand the site better. Article Schema is a form of structured data that includes the headline of an article, the author, the publish date, and many, many more, and it can also win you rich snippets.

However, we wanted to test again if that is the case. Google's advice might not work for all industries, so we decided to add article Schema to a customer's site. With that happening, we should see an increase in organic traffic, as Google has said that it will understand the site a little bit more.

However, we were surprised to see no impact at all on the site. We saw no increase. We saw no decrease. We saw just a flat line, which was quite interesting because, according to Google's advice, it should understand the site better. Therefore, we should have more organic traffic because of improved clicks, improved rankings, and maybe by winning the rich snippets. But we saw nothing. So that's why it's very important for us to test all the time to make sure that things work.

However, this customer did decide to roll out the change anyway because it can be seen as a positive as opposed to nothing because it didn't damage your site, and it didn't do well for your site. You might as well just implement it anyways because Google has said that this is best practice.

So that's why all of these ideas are sort of best practices that Google has given. However, it's important to test it on your site's industry and on your site's organic traffic to see if it has an impact on you or if it can go a completely different way. This is a few things that we've gone through.

We also have case studies on our site with many other SEO test results that we're quite surprised with, which you're more than welcome to see on searchpilot.com. But that's all I have for today. So thank you very much, Moz.

Transcription by Speechpad