Join Steve Harris for an in-depth discussion in this video Using Optimizely for A/B testing, part of SEO: Ecommerce.
- Now that we've successfully optimized our site it's time to look at some great tools around the web that you can implement to continue making improvements. In the world of e-commerce, making minor tweaks to the design and layout of your site can have a huge impact on the number of sales that you make. While this type of testing isn't directly related to Search Engine Optimization, learning how to A-B-Test design tweaks on your site can clarify exactly how to present your information to visitors that you've worked so hard to acquire.
A-B-Testing is also called Split Test and it means that you're going to show slightly different versions of your site to a segment of your traffic. My preferred tool for split testing uses a service called Optimizely. You can see it on screen here at optimizely.com. There are several great tools on the market that do a very similar thing. In order to use Optimizely you simply need to make a few design tweaks to your site which are actually made directly in the Optimizely interface and determine your goal whether it be a purchase, newsletter signup, or simply clicking through to a specific page.
Then we can compare which version of your site performs the best. Let me show you how it works. The first thing I'll do is log into Optimizely. One I've logged in, this is what our backend dashboard looks like. So it says on the top My Web Project and below that we have a number of experiments that we can set up. So let's set up an experiment using our Mix Athletic site. The first thing that we're going to do is click this New Experiment button. When I do that we need to give our experiment a name and a URL.
I'm going to test a design tweak to a product page on Mix Athletics. I'd like to see if changing the Add to Cart button to red encourages more clicks. This is the perfect type of experiment to run on Optimizely. So I'm going to copy the URL of this page let's go back to Optimizely, let's give this experiment a name; Add to Cart. Now we'll paste the URL in this box. And we go Create Experiment.
Once we've created the experiment, Optimizely loads a preview of your website. And in this preview you can actually click on elements to edit them. So you'll see in the top left here, we have variations. We have the original, and we have variation one, and right now the two of them are identical. Let's make some design changes to variation one. Insuring that this is clicked, let's scroll down to that Add to Cart button, and you can see our mouse highlights it within Optimizely's interface.
If we click on that we get a fly-out menu here that gives us some editing controls. So let's edit the Element and let's edit the Style. Once we've clicked Edit Style we have some other tabs on elements that we can change, but I'm going to change the color and background and let's change the background color here using this color wheel and let's select red. Just like that, and click Done. There, so you can see now our button has changed, even though we haven't made any edit to our source website, Mix Athletics, we can do all of the edits within Optimizely really easily.
And if we compare our variation, so the original and the variation you can see the design changes you've made. We could add multiple variations to this as well. The next thing that we need to do here is, I'm going to just save the experiment, and we have some options to set up targeting and goals, and those are these two buttons on the top right. FIrst let's look at Audience Target, if I click on this, you can see that we can set up a specific audience that we want to see this experiment.
So right now it says to everyone, all visitors will see this experiment, if we wanted to we could create a new audience and we could add some dimensions to this. For example, we could call it: Users That Come Directly from Search. And then all we would need to do is set up a few conditions. If I scroll down in the Conditions list you can see that we have things like, Traffic Source. So I could drag that Condition out here and we could set some options like Coming Directly, Coming in Through Search etc.
For now I'm not going to set up any custom audiences so let's just click back to go back to the previous screen and close this, I'm going to show this experiment to everyone. The next thing we can do is setup a goal. If we click on the Goals button, what we want to do here is define something we can measure as a success so we know which experiment worked better, or which variation of our site. So we can click Create a New Goal we could add a goal name, let's call this Add to Cart and what we want to do here is define what exactly we're tracking, we could track Clicks, for example on this button.
So it shows us a preview of the URL we have and we can just simply scroll down, select our Add to Cart button and now it's tracking clicks on that button. And then we just need to hit Save. There, so now we have a goal and we can measure which one of these design changes works best. Let's close this window and click Start Experiment. Once we hit start experiment we just need to make sure that we've installed this code on our Mix Athletic site so that Optimizely has the ability to control the page.
This small block of code can be easily copied and paste directly into your site using shopify's backend interface. For now I'm going to hit close and let's drop back to our dashboard and I'll show you what the results of an experiment like this look like. Here's a previous experiment I ran testing different colors of Signup buttons. If I click on this experiment and click Results, you can see that we get a nice overview page showing the results of the experiment. So out of 1,113 visitors in three days running, my variation one had 542 visits and variation two had 571.
Of those specific visitors, our goal, which we setup, which was membership signup, we saw a 40% increase in signup's by using variation one. So this was a very successful test showing us that the design tweak we made on variation one is going to be more successful. A-B-Testing seems initially, complex but it's a common misconception that you need to actually create alternate pages directly on your site. The optimizely system takes care of presenting the tweaks you've made, so setting up and analyzing experiments can be done in just a few minutes.
- How SEO affects the buying process
- Researching customers, keywords, and competitors
- Using simple and long tail keywords
- Eliminating duplicate site content
- Using rich media
- Designing landing pages to drive engagement
- Blogging for SEO
- A/B testing with Optimizely
- Measuring traffic with Google Analytics