Split Testing Secrets

What You Need To Know About split testing.

thThe majority of very smart internet marketers make their money because they test and tweak every sales message until they get a huge and positive response. They do not under-estimate the power of A/B split testing, in order to get your split testing strategies right there are three very important things to look out for and implement.

Firstly, be aware that it is harder to set up a split testing campaign than you might think.

For example:
You must upload a new sales page with just a single change on it. You may also have to create two completely different sales pages, depending on how your tracking is set up. It is imperative that you have a script that splits traffic between the two mentioned pages. This traffic splitter should split the traffic randomly and just on every other occasion. The traffic splitter must place a cookie on the visitors machine, thus sending them to the same page the next time they visit.

A tracking system must be put in place so you can keep a track of visitors and sales for both pages separately. You may also have to set up a data base.
This is just the beginning because while your test is running bad things can happen. For instance you may notice a slight typing error on advert B. It may not affect conversions but you don’t like it so you feel it must be rectified. It is ok to do this but it is probably not a good idea to change a page during a test. A reason to be careful changing anything whilst testing is because it has been known for editing software to change the names of the fields in an opt-in form so it doesn’t work anymore. If this does happen you may not discover the fault for a week or so, after you marvel at how well test A has done.

thOther things can happen, ad campaigns can stop running, you can suddenly get listed on the front page of Yahoo and be getting traffic from different sources to what you were targeting in the first place. All of these things will affect how your ads perform. When something seems strange with your test results you will automatically want to check it out. Be careful though not to jump to conclusions because even perfectly legitimate results can look odd at times. Therefore you will need to know how long to run the test before the results are significant. Split testing can be a bit of a head ache but once a couple of tests have been done and your tools are in the right place it gets easier.

 

Secondly, be prepared for unsatisfactory testing results.

It is a fact that most of the changes you make to your landing pages will not help improve your end result. As a rule of thumb if you run 25 split tests, 13 will not show any significant difference, and 6 of the proposed changes will have a negative effect. Which leaves just 6 more to have a positive effect. Leaving you with 6 improvements out of the 25 tests you started with. Another thing to consider is that from 100 ideas you have you will only be able to pick out 25 elements to split test. This is because 75% of the ideas you generate for improving your ad will be things you don’t currently have, or are too expensive or time-consuming or they cannot be combined with other things you want to do with your page.

In short this means that you must generate 100 ideas to get 6 that will actually make a positive difference on your page. Unfortunately this process is normal procedure, but no one will admit it to you. split testing is a great way to increase your profits but it certainly is not as easy as you may have thought.

Thirdly, you will not be able to find all the nuggets by yourself.
There are some changes you can make to landing pages that will make responses explode and finding them is of up most importance. You alone will not find all the answers as to how to improve your landing page, so ask for help. Approach other people in the know, internet marketers, product creators, potential customers and existing customers too. Allow them to throw ideas at you about your page, gain inspiration from the ideas you take on board.

Brainstorming sessions are not necessarily practical for all but the opinions of others will certainly get you thinking, you may even approach the whole split testing application differently as a result. Consider checking out exactly what your competitors are doing, anything that will increase your conversions is relevant here. These few key ideas will improve your split testing results, so take some action.

In conclusion, do not get fooled into thinking that split testing is easy and just by changing a single word you will see a 300% improvement. Unless  you are really lucky this won’t happen. It is difficult to get results from A/B testing but not impossible otherwise no one would be doing it, the difference here is, doing it the right way will bring you success.

For split testing tools check out this blog post.

Leave a Comment