Simple Split Testing

It took me a long time to get into the habit of split testing things, to me it’s one of the buzz phrases that I use to roll my eyes at. After all, I thought I knew the market better than most people, and I foolishly thought I didn’t need to split test everything. But slowly I started split testing here and there, and when I finally started to analyze the results, I was shocked at the results.

I’m not going to lie, I thought split testing was super boring and I thought my time was better spent on other things. But it seemed like everywhere I looked, every blog I read, and everyone smart person I spoke to talked about the need for split testing. It got harder and harder to ignore, and when I finally decided to give it a try, I wasn’t even sure where to start.

One thing I noticed was that Email Split testing was always being mentioned by people like Neil Patel, Jeremy Shoemaker, and John Chow. I have a ton of respect for all of these guys, and I’ve been lucky enough to have conversations with them all. So I knew if they were saying I should do it, then there’s something to it.

Email Split Testing

My First Split Test – 700% Increase

Lessons #1 – Split Test Using Images - Simple Changes Make all the Difference

Back in 2009 I decided to do my very first split test with the help of my Intern (And Good Friend Greg B). So we setup a split test on what was at that time my highest traffic website, a site called Twilight Gear.  During this time in 2009, The Twilight Book were all the rage, and the movies were just coming out. So it was a HOT market. And luckily we jumped on it early, and our site was one of the largest “Twilight Fan” sites. Since we had a lot of traffic (20k+ Visits Daily) then that made this a perfect place to conduct some basic split testing.

I’ve mentioned this original split test before, and I wrote about it on John Chow’s Website, and it got so much attention the Aweber Marketing Team contacted me and asked me to help with several customer training modules to help other Customers learn how to conduct split test and increase their optin rates. You can read about this split test campaign here: How I Increased My Optin Rate By 700%

Here’s What We Did

When we started collecting email fro the Twi website, I was using a basic optin popup box. It was super simple, and it read:

Please Join Our Newsletter!

Enter Name

Enter Email

After a couple months of not paying attention, this optin form was getting a Less than .05% Optin Rate. Obviously that sucked! We knew we could do better, we had the traffic, Twi Fans are hardcore, and they wanted to be kept in the loop. So what was the problem? Why Weren’t people signing up? The answers were simple.

1. I didn’t give them a specific incentive (All I said is they would get a newsletter)

Optin Rates are always better when you provide a clear incentive to sign up – A Free Book, a Free Ebook, a Poster, etc – this can be anything really, but people want stuff.

2. The optin form wasn’t eye catching, it was just plain and simple. Most people auto close popup’s like this daily without even thinking about it.

What We Split Tested

We knew we could do better capturing emails, so we decided to setup a simple split test campaign, which is super easy using Aweber. Once you create an optin form, you can create another, and another, and another (As many as you want) – And then you just have to decide how often you want each one displayed. So we created 4 New Option forms, and we chose each one to display 25% of the time. So Visitor 1 would see one optin, Visitor 2 would see another, and so on.

Here’s How We Broke it Down

Optin #1-  Showed our Original Plain Optin Form as mentioned above

Optin #2 – One was a slightly modified version of Form 1

Optin #3 – Displayed an Optin with Robert Pattinsons’ Picture (He is the lead Twilight Actor

Optin #4 - )Displayed an optin with a picture of Taylor Lautner (Co-Star of Twilight)

So we added some visual aspects (Pictures), and that’s about it. Nothing drastic, nothing major. Also, keep in mind it takes less than 2 minutes to setup optin forms, not hard at all.

The results were amazing, just adding different pictures increased the optin rates dramatically! As expected the original simple split test form didn’t do as well as the two forms with Picture of the Lead Actors. Overall, the optin form with Robert Pattinson out performed all other forms. Out optin rate went from .05% to over 2%!

What Lesson Can You Learn?

Not only did adding a simple image of one of the main characters do the trick, the simple act of adding an image of a character they LIKED made a big difference. So the lesson here is simple, small changes can make a HUGE difference. See below

Email Split Testing

List Building Tips

So as you can see some simple changes, and a little split testing goes a long way! Also, keep in mind this was originally done in 2009, and we are still constantly changing and testing our optin forms for this site, and a high percentage of people who opt’d-in 4 years ago are STILL active subscribers, and we continue sending out emails and profiting from these lists.

It’s hard to get return visitors to a regular website, but when you have their email, you can stay in touch for years to come! (And Keep Making Money)

Lesson #2 – Phrasing is everything, use action words!

This is probably the easiest thing for webmasters to split test, simply changing one or two words can make a world of difference. But the problem most people have is they grew up listening to commercials and infomercials. Even today you see commercials selling some product with phrases like:

  • Act Now Before it’s too late!
  • Don’t Miss Out On This Opportunity!
  • You Can’t Afford To miss this Sale!

Don’t get me wrong, these are solid Calls to Actions (CTA’s) that work depending on your target market and you sales platform. For commercials and infomercials these types of CTA’s work well. But translating them onto a web page just doesn’t work.

Let me give you a few more common website CTA’s that don’t work (According to my split testing and Experience):

Download Now - Why Doesn’t this work? Well mainly because since the world started on the net, we have been conditioned to be paranoid about what we download, and for good reason. I’m a pretty tech savvy guy and I still end up with some type of malware/spyware a few times a year.

Here’s another one That doesn’t work great:

Visit Now – Although this is much better than Download now, it still doesn’t give the user much incentive to click the button or link.

Here are some better CTA’sHow Do I know? I’ve split tested these of course!

  • Get It Now!
  • Buy Now!

(For My Gambling Webmaster Friends)

  • Claim Bonus Now!
  • Play <RoomName> Now!
  • Get Rakeback Now!
  • Claim This Bonus!

All these CTA’s have one thing in common, they tell the user what to do, and they offer something. People are motivated to “Claim a Bonus” or to “Get Rakeback” – This also removes any doubt as to what you want them to do, and what they get for doing it. At it’s core, these are simple concepts that make all the difference.

Bullet Point are also very helpful when used with a good CTA – Simply laying out the benefits removes any doubts and questions. So for example. here is a call to action I would/have used:

Play PokerStars Now!

You Will Get:

  • 100% Match Deposit Bonus up to $600
  • 1 Free Tournament Ticket (Worth $100)
  • 1 Free Entry Into The Depositors Freeroll
  • Automatic Entry into PokerStars Poker School Club

 >Claim This Bonus Package Now!<

Here are some results of Split Testing Phrases 

I wanted to test my PokerStars Landing Pages – I create a split test using a simple split testing code, which allows me to setup multiple split tests easily in wordpress. So my goal was to split text several Call To Action Phrases and images. So the only variable in these tests are the Pokerstars Logo, and the CTA phrase.

Things to Know

1 - For each month, the number of page views is approx 900 Per Month to my PS Landing Pages

2 - When testing began the Visit to signup percentage was 11 signup per 100 visits. So a 11% Conversion Rate

- Previously before I started testing, I was using a standard Pokerstars Dark Logo, and a Blue Button w/ the phrase “Visit Now”

Testing Begin

  • CTA #1 – Standard PokerStars Logo – Phrase = Play Now! 
  • CTA #2 – Image of PokerStars Pro Daniel N – CTA Phrase = Claim Bonus Package!
  • CTA #3 – Enhanced PokerStars Logo – CTA Phrase = Claim Your Bonus!

Split Test Results

  • CTA #1 – Conversions Increased slightly to 13% Per 100 Visits
  • CTA #2 – Conversion Dropped Slightly to 9.5%
  • CTA #3 –  Conversions Increased to 14%
Each Added Conversion is worth Aprox. $200 – So if I end up getting 3-5 additional conversion, that adds up fast!

So CTA #3 is a Winner!! Now, as you know a lot of other things can have an impact on these conversion rates. But this was a simple test to do for any website. Obviously, this is a very specific example on a site with existing traffic, and traffic coming via a specific set of keywords. So mileage may vary.

I am constantly testing various phrases, Images, Button Colors, and pretty much everything possible. The key with testing is tracking, if you’re not tracking then you are wasting your time.

Over the What Lesson Can Be Learned

So you can see how effective simple split tests can be. It doesn’t have to be super complicated, and the more times you try little test, the more information you have to conduct more tests. Simple and Effective = $$$$$

Obviously what CTA’s you need to use varies depending on the type of traffic coming to your site, and what you have to offer. But these are all easy things to analyze and figure out. Also be sure that once you have split testing your CTA’s that you don’t get lazy, and stop testing new things. As we went over in the first example, adding images can make a big difference, and as in this example, phrasing makes a big difference.

With a few simple change you could be adding hundreds or even thousands of dollars per month to your bottom line, and that can make the difference in paying the bills, or going bust.

 I’d love to hear your split testing advice and results, please lets us know what you have tested and how it turned out in the comments below!

About Randy

Hi, I'm Randy and I've been working in the online marketing industry for close to 10 years now. I'm not the smartest guy around, but I push myself to learn something new everyday, and I try and surround myself with great people.

I love playing poker, spending time with my family, and I'm a serious SEO/Marketing geek. Please leave a comment because I love to hear people's opinions.

  • Pingback: NFL Betting Picks - Marketing Case Study

  • Poker SEO

    Nice post. Have you tried split testing with a period versus with an exclamation point at the end of your CTA?

  • randylayman

    Ya know I haven’t, but I know I’ve used both those before. While I was writing this that exact idea occurred to me.

    One thing I’m trying to be careful of is split testing too many things at once. So right now I’m focusing on testing Button Color, and Phrasing. But I love the idea of testing various punctuation :)