Category Archives: Split Testing

Split Testing: The Golden Rule Of Internet Marketing

Note: this an article I wrote a few months back as a guest post at BradleySpencer.com.

Lately I’ve been doing a lot of testing with landing pages, and some interesting things have come out of that. Actually, to be more precise, I’ve been testing ad copy / landing page combinations, but today we’re mostly just talking about landing pages.

Remember – as a marketer you need to test everything on your site, so even if the following has worked for me, you can’t just go and implement the same on your own site without testing – you might find your market is quite different from mine.

So… what is the prevailing wisdom on landing pages? Well, one school of thought (a very popular one, I might add) says it needs to be as short as possible, ideally with everything above the fold. Sounds pretty good, right?

I mean, it makes sense – you want people to see your whole message immediately, without having to do extra work to get to the opt in. If they have to scroll, a certain percentage of the traffic will simply not bother.

HOWEVER – the catch 22 with landing pages is that you need to give them enough information to make the decision to opt in. So there’s a balancing act going on here – you want as little as possible, to make things as easy as possible, and yet at the same time, you need to balance that with enough information that they become engaged.

Recently I tested this idea on one of my pages for a few days. You can see the results from the split test right here:

The variation in this case was the same as the control page, except that I hacked off a few paragraphs of copy that I thought may have been ‘overkill’ for what was needed. It brought the page up above the fold, and basically brought my page into line with the short page school of thought.

Well, as you can see, the short copy bombed. I stress – the length of the copy was the ONLY difference between these two. According to the mighty Goog – the original, long copy version of my landing page out pulled the shorter. Why is that?

Well, it turns out that the bit of copy I chose to hack was apparently speaking to a lot of people. It basically said there are two types of ____. One does XYZ, and the other does everything better than XYZ. (I’m simplifying here :)… My free course can help you learn to be the second type…

I’ve kept it generic so your brain doesn’t get stuck into a box about my particular market – you can use this idea anywhere.

On a side note, you may be wondering why the conversion rate is only 10% or so for an opt in… that’s because of the ad copy I’m using. The ad copy (banner ads on Google in this case) pulls incredibly high CTRs by grabbing people’s interest. The nature of the ad being what it is, many people click off after landing on the page… yet the tradeoff of far more traffic works out favorably for me in terms of cheaper conversions.

A different example – a friend of mine recently created a landing page that broke pretty much every ‘rule’ I know of  on landing pages. It was hugely long, only had a single opt in box at the very bottom of the page. It had hardly any graphics… Basically 2-3 pages of solid text. The only traffic he sent there came off his signature on a forum… and yet in a couple of days he got 180 or so views of the page, and something like 140 opt ins.

How can that be? Because his traffic was pre-qualified… they’d already seen his forum posts and knew something about him, therefore they were willing to read and engage with his long copy page, and many opted in as a result.

So the moral of the story is this: use a landing page that is appropriate for the traffic you’re sending to it. Yes, this means you’re likely going to need multiple landing pages… probably at least one for every campaign you’ve got running if you want optimal results. If you truly want the best results possible, you need to segment each traffic stream and then TEST your landing page for that stream until it performs the best it can.

At the end of the day – no matter what all the ‘gurus’ tell you about internet marketing… the fact of the matter is they’re probably just telling you what has worked for them, as I’ve just done. It doesn’t mean it will necessarily work for you. Each market and offer is quite different.

The only way to know is to test.

Using Google Analytics or Optimizer on Outbound Links with WordPress

GUEST POST by Bradley Spencer

Hey readers of Jonathan’s Blog. Today I’m going to share with you the exact code I use to make my affiliate links a lot prettier without having to buy a script or software or anything. And it only takes about two minutes and works with WordPress, Drupal, Blogger, or any CMS for that matter. And best of all you can track everyone who uses the hoplink in Google Analytics or Google Website Optimizer. This is the same system I created as ‘the missing link’ for using Google Website Optimizer to split-test my landing page at WordPressLandingPage.com and the system I use every day at my affiliate sites.

The problem I had before (that this solves) is that my affiliate links were ugly and people didn’t want to click on them, and also that I had no way to keep track of how many people I’d sent to a merchant from within Google Analytics. And, maybe most of all, I had no place to put the code that Google Website Optimizer gave me to put on a ‘conversion page’ But this little file solves all of that. Sounds great, huh? Well let’s dive right in.

Step 1:

You are going to need a few very common tools to make this work. So the first thing we want to do is gather those tools. Here is what you are going to need:

  1. FTP Access to your server
  2. Text Editor (any will do, but not a big word processor like Word that adds markup. You need a lo-fi text editor)
  3. Download this file which gives you example code you can use to get up and running in minutes.

Step 2:

Once you have your FTP connected, your text editor warmed up, and you’ve downloaded the file from Step 1 above, just watch this video to see how to get set-up.

Making the Most of Your Hoplinks:

So far the best reasons I’ve found for having these hoplinks are that they are better looking links than most affiliate links and that you can use goal tracking to see how many people you send to your merchant. If you can think of other uses for these hoplinks, then by all means leave a comment. I’m always looking for new uses.

About the Author:

Bradley Spencer is a WordPress and SEO consultant who can help you make the most of marketing with wordpress.

Website Optimzer: Fooled By Randomness

Lately I’ve been playing around with Google’s Website Optimizer a bit more. One of my landing pages is getting a good thousand plus views per day, so that gives me a nice platform to split test ideas on. I can run a split test inside of a week and get some relevant results.

Lately however, I started to suspect that perhaps the auto responder follow up series for one of my products was a bit weak, or contributing to poor conversions. So I decided to split my traffic, using the website optimizer, into two separate auto responder streams to follow up.

So the way I did this was very simple: I duplicated the landing page, and simply swapped the opt in code.

You wanna see the results? This is funny:


As you can see, Google has declared one of my pages a surefire loser. That’s like me losing against myself in a foot race!

This is the tricky part with split testing… you gotta split test, because if you don’t, you just have no idea. But even when you do, often times we take results as being a pretty sure thing, when in fact they’re not.

If two identical pages can get these sorts of results, then surely two similar pages can show up as winner/loser when in fact the truth is the opposite.

How do you get around this? As you can see, I ran about 3700 clicks through this particular test and came up with this result. I since realized (duh!) I need to be measuring sales and not opt in conversions so I’m not going to continue this test any longer, but probably the best thing to do is to run the test for a LOT longer than Google tells you you should.

In this particular example, truthfully, I would expect that if I ran 100,000 clicks through it, the results should stabilize. However, you never know…

Any weird things you’ve run across with Optimizer? Leave a comment below and tell us about it..