My Affiliate Marketing Experiment – Part 6

Part 1
Part 2
Part 3
Part 4
Part 5
Part 6:


At this point in my little experiment I’ve made some money and learned a few pro’s and con’s about this affiliate marketing industry (at least the low-end type):

Pro’s of simple affiliate marketing:

  • It’s possible to start making money very quickly with relatively little work.
  • It’s possible to scale up successful campaigns to large levels.
  • It’s possible to “sell” a product without knowing anything about it.

It’s not hard to see why so many get-rich-quick books teach this very method of making money…because it SEEMS like it could work for anyone.  Some people might even have some nominal success trying something like this, but very quickly the con’s start showing their head:

Con’s of simple affiliate marketing:

  • It’s possible to LOSE those quickly-gained profits very quickly also.
  • Most successful campaigns you create will start to garner much attention from other affiliate marketers who quickly start copying you.
  • You have to be hush about your work….you can’t openly talk about what you’re promoting or your methods.  This is why many people I talked to about this were so secretive.

By and far the largest con I see with doing this simple type of affiliate marketing is there is no long term value being provided. With my business HouseOfRave for example, everything I do today pays off for months at a time, even years.

With these small affiliate promotions, there is very little value ultimately being imparted to the customer.  Ultimately a beginner trying to simply “make money online” with these almost-scam-like promotions will probably end up disappointed.

But screw that, I was now up to $371 in profit with JUST BluCigs and I’d barely spent $70 to get there!

….and this experiment called for finding out how to make a landing and optimize it with Google Website Optimizer, so that was the next step.

Basically what I’m trying to do in this step is have several versions of the same page and measure the effectiveness of each.  After X-amount of people cycle through each version, a clear winner is usually identified.  This is called an A/B Split Test…and if you do this enough, you can identify which pages convert the most people.

For this I had to do several things:

  • Make a separate landing page to send traffic (I chose ecigaretterecs.com).
  • Make three different versions of the page.
  • Enter all this stuff into Google Website Optimizer.
  • Drive traffic to the site, wait for results.

I didn’t say this experiment was going to be pretty, I just wanted to learn the underlying lessons in it, so here goes:

Step 1.) Buy the domain name.

I chose ecigaretterecs.com because all sorts of other combinations were taken.  Ultimately it didn’t matter what the domain was.  Even I kept forgetting what name I had purchased.  I setup the site on my server and was up and running.

Step 2.) Make three versions of the page.

I wanted a basic page that looked sort of like a reviews site, a site where you go to get reviews of a particular genre of product.  I actually first purchased a WordPress plugin that makes a fully-functional reviews site, but I wanted this experiment to be extremely quick and dirty so I did it the ghetto home-made way.

Since E-cigarettes are quite new, there are multiple brands people want to look through before choosing one.  I would provide this review information for them on this (ghetto and poorly made) page (ecigaretterecs.com).

I decided I would make three very different versions: 1 stylish, 1 simple and plan, 1 supper-ghetto:

(This was the “stylish” one…shut up, don’t make fun of my web design skilzz).

(This was “plain and simple”).


(This was the bare-bones and ugly one).

…the point of this is to simply test out different versions and see if there was any statistical difference in how each of them converted into sales.  Sometimes the data is very surprising.  I find from many friends who do this on a large scale that ugly and simply is often better.

NOTE: An A/B Split Test doesn’t need to have versions that are so dramatically different like this experiment, it can simply be a different picture or different headline text for each version.

Step 3.) Enter into Google Website Optimizer

I went through the Google Optimizer process and outfitted each version of the page with special snippets of code (keep in mind this is something almost ANYONE can do) and set all the links to go to the affiliate sites I earlier created with the custom headers.

Now when someone lands on the webpage ecigaretterecs.com, it randomly selects which version of that page to show the visitor. So if you look at that page from your computer and your friend views it from a different computer, there is a good chance you will see a completely different page.

Google Website Optimizer then collects information over time and tracks the conversion rates to see if there is a statistically “better” version.  When I ended up pulling the plug on this experiment there were 190 page visits, but no statistical variance shown (you usually need around 50 conversions per variant before conclusions can be drawn).

I didn’t get the final results before I pulled the plug on this experiment, but I didn’t care.  I had learned how to A/B Split Test which was the important thing….now I can use it to make already-profitable pages more profitable.

So far I’ve accomplished almost everything I set out to learn with this experiment, and I even made some money.

I was still making money, so why did I finally end up pulling the plug on this experiment?

Go to Part 7.