Before FREE tools, like the one we will look at became available, better marketers used to pay a fortune for commercial products or they used to engage in simple, manual split testing.
Manual split tests are typically what are known as A\B Splits.
An A\B Split test consists of an original webpage which runs for several days, weeks or months, depending on the volume of traffic going to that website. The higher the traffic the shorter the run period needs to be and vice versa. During the run, statistics are manually compiled to determine the “conversion of sales” proposition for that website.
At the end of the original run period, a new alternative page is loaded and run for an equivalent period.
The difference between the original and the alternative pages is held to one single item such as headline, offer, introductory paragraph or image.
The reason the change is limited to only one item is because, if you made more than one change and the alternative “converted more sales,” you would not know what change to attribute the increase to.
At the end of the alternative page’s run period, a manual comparison of the conversion rate statistics takes place. If the new page achieved better conversion results, it becomes the new control and another alternative is created with a different modification in one item. If it was worse then it is discarded and the cycle starts all over again.
The main problems with manual A\B split testing are: 1. It is slow, testing only one change at a time over a period that can be up to several months long.
2. It is inefficient depending as it does on manual recording of statistics.
It sounds like the definition of tedious, even if profitable work.
Eliminate Guesswork From Your Marketing Strategy And Convert More Sales With FREE And Simple Split Testing.
11
Charles Kangethe - ©2012 www.wealthcreatingprogram.com Can you see why Splitter was not very keen on doing this type of testing? Unfortunately, because he did no testing at all he had a miserable trading year!
Automated A\B Split Testing.
It was not long after split testing became a tightly guarded secret amongst the Internet marketing superstars, that a clever technology type decided, “We can automate A\B splits.”
And so the first automated tools started to appear. Initially they were crude and quite technical but without exception, they were all expensive.
They did exactly the same thing as manual A\B splits, but with a twist.
Using an automated tool allowed the ontrepreneur to create many test pages and load them all to the server. They then embedded some code in the HTML of the pages and started the test.
The split test tool would then a different page on a rotational basis to website visitors. It would also compile statistics on the fly about which page was “converting more sales.” Web masters and ontrepreneurs would then review these statistics and make changes to show the best performing page to visitors.
The limitations of the manual A\B split test had been overcome. Now the tests were quicker and the statistics more accurate. However, a few new problems cropped up.
Firstly, the new tools were expensive. A recent search revealed that even today in 2011 a simple A\B split testing tool will set you back just under $150.00. Splitter made no money in his first trading year, $150.00 in that context is a lot of money!
Secondly, even though the tools make testing quicker you are still only ever testing one condition against the control at a time. Many different conditions on a website influence visitor behaviour. Changing one condition per page and comparing it against the original or control is still a long process even with an automated tool.