Testing your best headlines with Google Analytics Experiments

By  •  Updated: 06/09/16 •  8 min read


In this next lecture, we’re going to go through an incredibly important tool using Google Analytics and their Experiments feature. Now this is used to try out different versions of your landing page, which is a really important thing to do. Very rarely when we set up landing pages do we get the best result possible the first attempt at headline. Absolutely the best way to constantly improve on your marketing is to try different things out and test one against the other. Not pull things down and replace them, but to try them in parallel and see which ones actually work best. If you just keep changing things, there’s too many other variables, like a sunny day might mean that fewer people are at their computers, so you’ll get a lower response. Things like that will go on.
I’m going to show you today how to test two different versions of our landing page. We’ve got the one that we set up during our course here, which you can find at sportivetraining.net. We’ve already set up a Google Analytics account. You can see there’s a little bit of data in there, but I’ve cleared this out and I’ll show you how to do it. What we’re going to do is to take this version of the landing page, which is our current index page, and we’re going to try out with two different versions of this headline. A relatively small change, but a really important one. Headlines are probably the most important thing to test. Let’s go on and have a crack at it.

Here are the different pages that we set up. The index page is our [inaudible 00:01:45] page. Thanks page, just to remind ourselves, is what you get after you’ve put your email address in. Then when you click through on the confirmation page, then you get this video sales letter with a link to the freebie they were offered at the start, and also the link to our [inaudible 00:02:17]. Okay. What we’re going to be doing first of all is we’re going to take a copy of this page. Now we’ll start off by opening these up in Pingendo, which if you remember is our editing tool for these pages. Now this does require a bit of code, but we’ll try and make this as simple as possible. The first thing we’re going to do is to clone this page. I’m going to right-click on that, duplicate, and I’ll change that name to “Index 2”. I’m going to open that up in Pingendo, as well.

Okay. Let’s just change this headline. We’ll go from just getting through sportive to trying something a bit more ambitious. I’ll try and spread that out. Okay, so there’s two different headlines there. We’re just going to save that one. The rest of it is the same. Just that headline changes. Okay. Now we’re also going to open up our thanks page and our content page. What we’re going to do is add our Google Analytics code to each of these. If we go back to Google Analytics, we click on admin, and tracking info, tracking code. This will give us to code to add to each of our pages. We’re going to copy this and we can see it says here, “Add this immediately after the opening body tag.” Now I’ll show you where to find that.

If you go down the page here, you’ll see a section that says HTML. That there is the opening body tag. It’s on about line 10. Just after that, we’re going to paste that code. We’re going to save it and we’re going to do the same on each of our pages. We’ll close all of the pages apart from the first one. Now let’s go back. Let’s go back to the reporting tab. We’re going to go to behavior experiments. We’ll create a new experiment. Now an experiment is just trying out two different things against each other. We’ll call this “Headline test number 1”. Create new objective. Okay, so we’ll use this create an account template. Click on continue.

Okay. We’ll go back to reporting, to our test. Have our email signup as our goal. We’re going to do it on 100 percent of the traffic. We won’t have much traffic on this to start with. If you’ve got a very big site and you don’t want lots and lots of people seeing your tests, sometimes you can show it to just a few of them. I’ll leave email notifications off. Let’s have a quick look in there. Now Google will optimize itself to show you the best version as time goes on. We’ll leave that off, so that will run. The minimum time for the experiment to run. We’ll leave that as 3 days. Our confidence threshold, this is how sure you are that the results are basically going to stick. If you get lots of people signing up to your original version, how confident are we that that’s the right one and that we haven’t just had a freak of statistics? 95 percent is pretty solid for what we need here.

Okay. We then look at our two pages for our experiment. Index HTML and index two HTML. Now what we need to do is to manually insert this code at the top of our page is what it says, after opening head tag. That means further up and we can see there is our head tag. Save that. Now we’re going to load this up, the same way as we’ve done all the way along. Zip these files up. Log into our forward website. Drag the archive onto the page. We can now see that index two page. Okay. There we go. Now we need to leave this for a few days and get some traffic to the site. We’ll come back later to look at that.

To decide which of our two landing page versions [inaudible 00:10:09] we need to put some traffic to it. For this, as our supportive training website doesn’t have much traffic yet, I’m going to show you another page, which is almost exactly the same in terms of how it’s laid out. That’s the freefunnelformula.com, which is the funnel that leads to this very course. You can see here, we’ve had a little bit of traffic, which has been driven from some Facebook posts. Down here, under behavior experiments, we’re running a headline test. Just like the one that we ran on supportive training. You can see here, we’ve had a couple of hundred visitors to each version of the website and a reasonably amount difference between the conversion rates on each one. 21 and 12, that’s almost 9 percent versus about 6 percent.

You can see here that we haven’t quite got enough data yet to have picked a really reliable winner, but we’re not far off. You can see how these have actually moved backwards and forwards over the course of time a little bit. There’s not a vast amount of difference between them at some points with small amounts of data. You’re not going to get a conclusive result, but you can see from this experiment, this one is looking like it’s the better performing one. I’m just going to show you the two different versions of this. You can do that by clicking on this little arrow button here.
The first one is our normal free funnel formula page. Learn how to build a marketing sales funnel with 100 percent free tools. The second one, I wanted to try out whether speed would be more useful to use than the fact it’s free. Learn how to build a marketing and sales funnel in under an hour. If you’ve been through the course, you’ll see that the course itself is less than an hour’s work. I recorded the whole thing start to finish and editing in a couple of hours. The fact it’s quite quick to build these funnels is attractive to some people, but you can see that the fact it’s free, so this one, the original free tools version, is the one that’s winning at the moment.

The nice thing about Google Analytics is when it’s got an eventual winner, when it’s got enough data through it, it’ll just automatically switch to show the one that’s winning. Then you can go back in and check it and just remove the other pages. There we are. That’s an excellent way of testing out headlines, which are really the most important part of grabbing your customers’ attention and showing the proposition you’ve got for them. If you’re driving advertising and you’re not getting much in the way of response, I think that’s really the first place to start looking. Firstly, at your advert text, but also the headlines on your landing pages as well.

I hope that’s useful for you.

Stephen Pratley

I build email lists, that grow into one-man businesses.