Here's a handy way to know you're on the right track.
Anyone who has listened to my seminars or followed my work will know about the importance of split testing.
To recap very briefly, it is a way to see which adverts are more attractive to your potential customers.
It's a practice of improving your sales results - a legacy from traditional direct marketing if you will.
You post a mailshot to ten thousands prospects with one headline and another ten thousand to another list of prospect - with a different headline.
You then pick the winner that produces the best results. This way, over time and with repeated testing, you are constantly improving your advertising.
But of course it can be applied to web pages, emails, PPC adverts, direct mail in fact almost all forms of advertising. It's a very simple tool. The good old fashioned A and B Split test.
However, one question I get asked all the time - especially when people are trying to optimise their Pay per Click advertising, is how many samples do I need to take to get a positive answer?
For example, if I have 10,000 impressions of advert A and 10,000 impressions of advert B and advert A converts at 2% whilst advert B converts at 5% then I am almost 100% sure that the statistics are meaningful and an extremely good indicator of future performance.
But of course if I'm running pay per click or mailshots or whatever - this could be expensive, so we need the minimum sample to be able to make valid predictions. The more often we get valid samples - the faster we optimise our advertising and the more money we make.
However, if our sample is too small - we can not be sure our data is valid and can make expensive, inaccurate predictions.
The solution? You need a tool that does the maths for you and gives you stats that you can be confident to work with.