A/B Split Testing Versus ABAB Split Testing

a-b-split-testIt’s amazing to me that after writing about Split Testing so many different times on this blog, that I still receive email’s daily asking me the best methods for split testing, different ways of split testing, which options work best, etc.

If you’ve ever spoken to me via social media sites or via Instant Messaging for any length of time, you’ll know that I’m not the type of person to say one method is better than the other.  Unfortunately in this business, certain methods work for some people, while completely different methods work for others.  There is no clear cut solution or any clear cut path to follow in order to succeed in affiliate marketing.

However, I received an email about two weeks ago with a very specific question about split testing.  I responded to the person with the same advice I’m about to talk about in this blog post.  The email asked…

I know split testing is important, in fact, from what I understand it may be the single most important aspect I can do in affiliate marketing to influence the scaling process of my campaigns.  However, I’ve seen several methods for split testing and I was wondering which method worked best.  Some sites I’ve read say that you should split test two versions of whatever your testing at the same time, while others suggest split testing one type one day and then another the next day.  Which method do you feel would have the greatest impact on my split testing efforts?

Now, when I started split testing, I always used the A/B method of split testing which means that you split test two variations at the exact same time giving both equal exposure.  In fact, it’s not something I have even really thought about thus my reason for never writing a blog post about it before now.  However, after thinking about it, I felt that it was worthy of a blog post because I do believe there is a very clear cut way of doing it that will yield the best results!

The other, most common, method you see affiliates use is the AB split testing method.  The only difference between the two is that with the A/B method you are split testing both things at the same exact time giving them both equal exposure.  With the AB method you are split testing them at different times, or days.  For example, you may split test a landing page on Monday, document the results, and then put up a new landing page on Tuesday, compare the results and pick the best one.

The downside with the AB split testing model is that unfortunately data can be very different on different days and even on the same day but morning vs. evening.  Some campaigns do really well during the evenings, but horrible during the mornings, if you are running one page in the morning and one in the evening, your better landing page may actually be the one that you are running in the mornings but because conversions suck for your niche in the mornings, it looks like your landing page doesn’t do well at all.

However, if you are using the A/B method you know that both items you are split testing are getting the same exact data.  They are sharing visitors equally and both are being displayed both in the morning, afternoon, and evening.  This insures that you aren’t receiving data that is skewed because of changes in the visitors you receive throughout the day.

Now, as I said in the opening paragraphs of this blog post, not every method works best for everyone.  There is one other method that I’ve found works decent and that is the ABAB method.

This works just like the AB method, but there is one major difference.  You split test landing page one on Monday, then landing page two on Tuesday, however, you then continue on giving each one more time.  So you’d split test landing page one on Wednesday for the second time, and then split test landing page two on Thursday for a second time.

Personally I still don’t believe this is the best method and when compared with the A/B split testing method, I’m not sure why you’d choose it, however, there are some that believe that the ABAB method is the best choice.  In fact, it is a common way of split testing in other fields like Pyschology.  These fields use the ABAB method for most experiments.

Ultimately the choice is yours, however, I’d highly recommend going with the A/B method or at the very least the ABAB method.  Using the AB method is in my opinion simply being lazy and it will impact your business in a very negative way!

Which Offer Should I Use?

testingSplit testing an offer doesn’t have to be as hard as people make it!  In fact it’s actually quite simple if you just understand what exactly you’re doing when testing different offers and deciding which offers to use.

First let me start off with the biggest pitfall that most affiliates fall prey to when it comes to offers that affiliates are promoting…  They quit testing new offers!

Affiliates find an offer that’s working, working well in fact, and they immediately stop testing new offers within their niche.  The reason for this is actually quite simple, they are afraid that they’ll find an offer that converts worse than the one they are using… and a worse converting offer means they lose money!!!

First, let me explain why it’s worth losing a little bit of money now or taking a chunk of your traffic and split testing an offer that may actually convert worse, sometimes…. a lot worse!

The bottom line is that things in this industry change, and they change fast!  Offers come and go and great converting offers sometimes become horrible converting offers.  This can be caused by several different factors, including but not limited to, too many affiliates pushing the offer (as an offer converts really well, more and more affiliates find out about that offer!), the advertiser changes their landing page too drastically, or other advertisers start promoting something better than what offer you’re running is!

So what do you do?  How can you test other offers without losing too much money?  I have to be honest, the main way is going to sound soo simple, in fact, many of you may simply be like “Duh!”, why didn’t I think of that, but seriously it’s not your fault!

Don’t split test 50/50 on your offers!  I know this is something that is drilled into affiliates heads from just about every blogger out there including myself, as it works for most stuff, however, there really isn’t much of a need to split test your traffic 50/50 to test out a new offer.  Instead, maybe try an 80/20 split, sending 80 percent of your traffic to the offer that you know converts well while only sending 20 percent to the new offer.

Secondly, don’t split test more than two offers at a time!  The more offers you’re trying to test at a time, the more you’ll have to split up your traffic in order to get a good read on each offer!

Thirdly, keep a close eye on the new offers, as soon as you feel you have enough data (clicks/leads) that have come in to get a good reading at how the offer converts, then move on!

Lastly, Talk to your affiliate managers!  Most networks have the average conversion rate of what an offer gets across the entire network, while this will change person to person, it can at least give you a starting point as to what offers to split test first!

I know these sound like some easy tips, and for my more advance readers, sorry, but I get tons of emails a day asking me how to properly split test offers, so I thought I should write up a little article on how!

You’re an Idiot if Your Not Split Testing

idiotI talk to affiliates after affiliates that are starting to make money in this business, and one huge common problem that I hear and see from these affiliates all the time is that they are not split testing anything and if they are, they are doing it so half-butted that its not doing them any good…then they have the audacity to complain that split testing doesn’t seem to work for them…

At first when I started mentoring and consulting individuals, I thought to myself, well maybe they just don’t realize how important split testing is, but than I realized that just about every single affiliate blog out there talks about split testing ALL THE TIME!  In fact, I’ve talked about it in depth several times on this blog and touching on each aspect of split testing, whether it’s split testing landing pages, or split testing ads, split testing offers and networks, or whatever!

I’ve explained in these posts, why you need to do it, how extremely important it is, and what types of benefits it will have on your campaigns, in fact so have a ton of other bloggers like Chad from CDF Networks in which he wrote a post entitled Split Test Everything!  Jonathan Volk wrote a post on Increasing Landing Page Conversions where he talks about split testing and it’s benefits and why you need to be doing it!

I could go on and on with posts all over the internet that talk about split testing, yet there isn’t a need for me to, simply  go to any affiliate blog and type in split testing, you’ll see that it’s a subject that is talked about all the time, yet very few affiliates seem to be doing it!

Instead, they launch a campaign in a niche like payday loans, it isn’t profitable and they move on to another niche and so forth over and over wondering why they aren’t making money!  Instead, find out why you campaign is failing, whether it’s your ads, landing page, keywords, etc and split test them!  What???  You aren’t tracking each of those things to find out where your problems are?  Then you’re an even bigger idiot!!!

Quit being lazy, quit making excuses and start split testing…  I promise you if you do, you’ll drastically see increases in your profits as an affiliate marketer!

Sorry guys, I love you all, but really felt like I needed to be blunt with this one!  Hugs and Kisses!

Sometimes It’s The Small Things!

I think we sometimes forget as affiliate marketers just how much the small actions we do can increase conversions or click through rates when it comes to our landing pages.  We always worry about split testing the main headlines, the big picture at the top, the call to action phrase or button, etc.  We focus on these huge things because just by increasing one of them we can possibly raise our conversion rate or CTR by 5% or even 10%.

However, there are only so many big things we can change.  When it comes to small things though, we have tons of little settings we can fool around with like colors we can change, or certain words we bold, italicize, or highlight, etc.  Each of these small things may only raise our conversion rate or CTR by 1% or even .5%, but if we do 30 of these small things we could raise our CTR or conversion rate by 30%!!!

So what exactly are some of the small things that we can test?  Did you know that certain fonts convert better than other fonts?  Did you know that certain background colors convert better?  How about the fact that by bolding or italicizing a couple words throughout your landing page could increase conversions?

Have I sparked your curiosity yet?  If so, then hopefully the list below can kinda be a checklist for you to help you remember to check the little things when you are split testing your landing pages…

Things to Split Test on your Landing Page:

  • Split test what font you’re using!
  • Have you split tested your background color?
  • Have you tried to bold or italicize certain words or phrases that could help sell users?
  • What about underlining certain things?
  • Are you linking to the offer page throughout your article?
  • Are you including a Go Back To Top Link at the bottom of your landing page or another Button or Link at the bottom to take your user to the offer page?
  • Don’t forget to add testimonials to help build trust with your visitors if your niche allows for them.
  • Can you include any endorsed by, affiliated with, or as seen (in, on, at) buttons?  For example, As seen on Oprah or Endorsed with the Good Housekeeping Seal!

The list above is just a couple of the “small things” you could be doing and adding to your landing pages in order to increase conversions and CTR’s.  If you aren’t doing them, you need to start!  If you are using them then good for you!

If you have any other “small things” that you know of that work well in increasing CTR’s and Conversions I’d love for you to share them with me in the comments below.  It’s the smalls things that sometimes make huge differences!

Split Testing Offers and Networks

Axe Splitting Wood

I’ve decided to write a couple more posts on split testing.  I’ve currently written one for both split testing ads and split testing landing pages.  In this post, I’m going to be talking about split testing offers and using different networks to do this.  In future posts, I’ll be talking about other things you need to be split testing.

Most blogs you read talk about split testing ads and then they leave it at that, however, as you’ll see from this blog there are many more things to split test!  I wasn’t going to do anymore articles on split testing because I wasn’t sure if it was necessary to touch on each one, however, after reading a post by Chad from CDFnetworks entitled Split Test Everything, I decided to go ahead and continue doing articles on split testing.

Split Testing Offers

Most affiliates leave this step out and in my opinion it’s a very important step.  In fact, some do start split testing their offers, however, after they find a decent one they all of a sudden stop!  Instead of pushing forward to try and find an even better converting offer, they become complacent.  So what exactly do I mean when I say split test different offers?

Split testing your offers is the same process as split testing your ads or landing pages.  In essence you are simply swapping out your affiliate link from one offer or network and replacing it with a different offer or network to see if that other offer is able to give you a higher ecpc.

So I simply try to find an offer with a higher conversion rate?

No, most affiliates fall into this pattern of looking at the conversion rates of offers!  So before I continue talking about how to split test offers and common practices, I want to clear this up!  You are looking for an offer that has a higher ecpc (estimated cost per click).  Let me explain why real quick…

If you have an offer that pays out one dollar for every lead and it converts at 25% that means for every 4 people you send you’ll earn .25 cents.  .25 cents is your ecpc.   That means that the highest amount you could bid in order to break even.

Now let’s say you have an offer that pays out ten dollars for every lead and it converts at 20% that means for every 5 people you send you’ll earn 2.00 dollars.  2.00 dollars is your ecpc.  That means you could bid up to 2.00 dollars on your keywords and still break even.

Of course that isn’t always true, sometimes you have an ecpc of .50 cents but if you raise your bids to .50 cents you’d start to lose money, however it gives you a general idea on what you could spend up to.  That’s why it’s important to ask for the ecpc for a particular offer when looking.  Simply ask your affiliate manager and they should be able to supply this information to you immediately.

Moving On

So, you’re not looking at the conversion rates, but the ecpc.  So after you have selected your first offer to run, you will want to allow time for data to come in.  Normally a week is about what I like to give each offer personally.  This allows me to make sure it’s as accurate as possible since some days perform better than other days during the week depending on the offer.

After you have run your offer, you simply replace it with another offer to see if it will perform better or not.  Okay, I know your next question…  How do I know which offers to try and which ones to stay clear of?  Well, unfortunately there isn’t any surefire way to tell if an offer sucks or not without testing it yourself in my opinion, however, there is a good way to get a starting list.

Simply email your affiliate manager and ask them if they could send you a list of all the offers they have for the niche you’re working in and ask them to include the ecpc for each offer.

NOTE: Make sure if you’re only running PPC traffic to it, that you tell them to give you the ecpc for search traffic only.  This also goes for organic traffic or email traffic.  That way you ensure that you’re receiving the most accurate data you can.

After you receive this list, you should have an order in which to test the offers out.  Starting with the offer that has the highest ecpc and moving down.  Easy enough right?

Testing the Same Offer on Multiple Networks

Did you know that multiple networks normally have the same exact offers?  Did you know that most even have the same exact landing page?  Okay smarty pants…Did you know that the same exact offer with the same exact landing page can possibly convert better and have a higher ecpc on a different network?

That’s right, your ecpc can vary depending on the network you’re using!

Not only do you need to test different offers, but you also need to test different networks.  That makes it a little more difficult, well not really difficult, just more time consuming!  I won’t beat around the bush, some networks scrape leads, although sometimes it’s not really the networks but the software they use…*cough* Direct Track *cough*

So the bottom line is you need to track both different offers and the same offer on different networks!

Finding a Great Offer with a Great ecpc

As I mentioned at the beginning of this article, I see affiliates all of a sudden stop split testing offers.  Why?  Well, my guess is because they find an offer that does really well and have a great ecpc so they don’t want to keep testing and drop to an offer that does worse, however, you must also remember that you could again find one that does even better!

What I Recommend

If you find an offer that is doing exceptionally well, instead of split testing new offers across your entire campaign, only send maybe 10% of your traffic to a new offer in order to keep testing!  This will still give you an idea on offers and if you find one that is doing better in that 10% then try it across the whole board!


Do you currently split test your offers?  Do you split test the same offer among different networks?  Do you feel that all affiliates should be doing this simple step?  If you could say one thing that could encourage some affiliates to start split testing their offers if their not currently doing so, what would you say?