You can split test an opt-in page on your WordPress website for free using VWO. Yes, VWO has expensive paid features. But split testing landing page conversions isn’t one of them. As long as you’re not processing tens of thousands of visitors a month, the free tier covers everything you need to run a legitimate split test.
This tutorial will walk you through how to create a split test using a WordPress website in order to determine which page has a higher conversion rate.
How You Will Track Conversions
There are a lot of different ways to track conversions… in fact, there are a lot of different conversions you could track! Our case here is determining if signup page A or B has a higher percentage of visitors who convert to webinar registrants. The same approach works for any lead conversion on WordPress: opt-in forms, contact pages, sales pages, or any page where a form submission redirects to a thank-you or success page.
The way we will determine this is by telling VWO to count how many people land on our ‘success page’ from Page A and how many people land on that page from Page B.
Our control is the page which has perviously beat other designs. This is our Waterfall template which we mirrored on the standard Demio page. We still use Demio for webinars, but we wanted more control over design, tracking, etc.
Our variation is a two-column registration page with some color variations. Each of these pages has it’s own URL on the same domain.
Both pages will point to our ‘success’ page which offers the user the invitation to purchase the training lifetime access. When a lead signs up, they are redirected to this page and when they hit this page, it will track that as a conversion.
Install VWO On Your Website
If you don’t have a VWO account yet, you’ll need to sign up for one. It’s free. During your free trial you’ll have access to all their products. When your free trial is over, you’ll need to ‘subscribe’ to the Testing Product, but just select the free tier and you’ll be good.
There are two primary ways to run VWO on a WordPress website. The first is by using the VWO plugin.
The second is by installing the HTML code which goes in the header of your website. I have opted to install the code via HTML in the header code in effort to avoid running another plugin on my website. Either method works. I prefer the HTML route to keep my plugin count down and avoid any potential conflicts with caching or performance plugins.
VWO specifically mentions to avoid using tag managers like Google Tag Manager for handling the code. Because of this, my code is installed in my website theme files.
Instructions for WordPress and Plugin
VWO Will Route Your Traffic
You may be wondering, “How will people see both pages if I only supply them with one URL?”
VWO handles this.
When your campaign begins, no matter which of the two URLs you direct people to, VWO will automatically route visitors to the page. It will do its best to distribute your visitors according to the traffic parameters you set (below).
In other words… if you have 500 visitors and tell VWO to send 50% of traffic to your control page and 50% of traffic to your variation page, it will automatically route approximately 250 visitors to the control page and 250 visitors to the variation page… even if you only share one URL with all 500 visitors.
Setting Up Your Split URL Test
There are a lot of conversion tracking functions in VWO. The one we’re looking for is Split URL. Why not A/B? A/B testing tweaks elements on the same page (a headline, a button color, an image). Split URL testing is for when you have two completely different pages at two different URLs and want to know which one converts better. That’s what we’re doing here.
Split URL is when you have two different URLs you want to test against each other.
Add Your Control URL
Copy the URL for your control page. This is usually going to be the page you have been using. You can change the variation name if you want… in this case I did to the template style.
There is also a dropdown for Page URL where you can select one of two options:
URL matches
URL matches pattern
I always choose ‘matches pattern’ because it tracks the URL whether or not there’s a trailing slash. Different tracking tools handle trailing slashes inconsistently, and this setting protects you from losing data over something that trivial. In addition, tracking applications generally avoid counting query parameters like UTM parameters as a part of the actual URL. These are the characters which come after the ? in the URL.
My choice of URL matches pattern is more of a safeguard against one of my teammates choosing to include or exclude a / when promoting the webinar.
VWO allows you to have more than two variations, however, unless you’re processing thousands of visits a day, you likely won’t accumulate enough conversion metrics to really get solid information. Two pages will give you a better picture.
You can also choose how to split the traffic. I will typically do 50/50 across my two pages. You can alter this if needed or desired.
Lastly, you’ll see an option for creating a URL of the page in previews. This is purely subjective. It doesn’t matter what you put here.
Defining Your Conversion Metric
The next step is to tell VWO what you count as a conversion. As we already discussed above, we’re counting a visit to the success page as the conversion. If a user completes a form element and hits submit but does not stick around long enough for the success page to load, the conversion won’t track.
This is largely not an issue because users expect to be redirected to some sort of confirmation message, so I’m not worried.
VWO allows you to measure a Primary metric and a Secondary metric. Primary is going to be the most important metric to track. In this case, how many people visit the success page.
You can track a second metric link How many people scrolled down on the page, or how many people clicked a button on the success page, or a number of other metrics if you want more information.
For this test we’re going to monitor just the Primary metric.
If you’ve run tests before, you may see saved metrics already available here. VWO stores past metrics when you check ‘save to Data360 for future use,’ which saves setup time on future tests.
Here are the steps & settings:
Click +Add
Click +Create
Visitors event you want to track: Page Visit
Where > Page URL
Include pages where > URL matches pattern
Metric will calculate > Unique Visitors
When creating a metric there is the option to “save to Data360 for future use” which essentially allows you to use the metric again in another test. If you opened up the Metrics page and you saw a bunch of options already available, these are past metrics that were saved.
After you add all your settings, click Create.
Targeting
With a free trial you can tap into measuring all sorts of types of traffic. Maybe you want to know the conversion rate just from people who register from an email versus those who come from an ad. Again, unless you’re dealing with a large amount of daily traffic, you’ll want to just focus on all traffic for the pure split test.
Also worth noting, while you have access to more traffic segments during the free trial, when your trial is up, segment configuration is only available on paid plans.
All Traffic is selected by default.
Integrations
Nothing to setup here. There are integrations you can use to forward your data into other applications. You don’t need to worry about that for this test.
Review Your Campaign
The Reviews tab will allow you to see all your campaign settings in one view. You can see everything mirrors what we created above.
Start Your Test
To this point everything is configured but your test hasn’t started… which means that traffic is only going to see the page you send them to. Remember, we had two URLs we’re testing. Until I start the test, there will be no traffic routing.
To start your Split URL conversion test, you’ll need to click the giant button “Start test”.
At this point, traffic will begin routing to both pages.
If you paste your URL into your browser 10 times and keep seeing the same page, don’t worry… it’s not broken. VWO knows that you are the same viewer and will keep showing you the page you first saw.
Once you start the test, you’ll be directed to the Reports page where results will begin to be showing in the coming days.
Measuring Your Data
Over the coming days you’ll begin to see data populate in the Campaign Reports tab. You’ll also notice there is a message saying “Collecting minimum data for statistical calculations.”
What this means is that you don’t have enough data yet for VWO to actually predict which will perform better.
This goes back to how much traffic you’re sending to your page over the reporting period selected.
A general rule of thumb: aim for at least 100 conversions per variation before drawing conclusions. If your traffic is low, run the test longer rather than calling a winner early.
Don’t let this bother you — you’ll still see enough data to know whether or not your page is converting.
If you look at the example above you’ll see one page is at 11.30% and another is 12.59%.
This difference in these numbers is so small that, in effect, it doesn’t matter which page you run. They’ll both produce the same amount of conversion.
While we’re not getting into conversion optimization in this article, the core insight you can take away from this data is that your two pages are too similar. You need to have more variance between the two pages. They need to be more drastic in their differences in order to reveal which direction is worth pursuing.
You Can Measure Anything with VWO
I have looked at a lot of different conversion tools, apps, services, and plugins. VWO has provided the best experience and has been the easiest to work with. I wish their paid plans were more accessible. But for what we’re doing here, the free tier is genuinely capable. Set up your test, let it run, and let the data tell you what your instincts can’t.
Give VWO a try and start measuring just how effective your landing pages are at converting traffic.
Yes. VWO’s split URL testing is available on their free tier as long as your site is below their monthly visitor threshold. You don’t need a paid plan to run the test described in this tutorial.
A/B testing compares variations on the same page. Split URL testing compares two completely different pages at two different URLs. Use split URL when you want to test entirely different designs or layouts.
Long enough to collect at least 100 conversions per variation. For lower-traffic sites, that may mean running the test for several weeks rather than a few days.
