
Your Opinion Means Nothing
That's right; your opinion means nothing when it comes to making decisions on design, copy etc... Your opinion is exactly that, your opinion, it is what YOU believe will work best or what looks best. It doesn't mean it is right or even the best option available to you, it is just what you think will work best. You follow your feelings when you first create something and then you turn to statistics when you actually want to improve your stats.
Why do you turn to numbers over opinions? It's simple, it provides you with conclusive data which you can measure which scenario is best. A great example of this would be the use of different opening headlines. You have headline A, B, C, D and E, you really like the looks of A, it appeals to you and you think you have a winner with it. You collect data via split testing and it concludes you made more money from headline E, what one are you going to go with? It's obvious, you will be picking headline E all day long because it is making you more dollar. Now, if you followed your gut then you will not know conclusively that you are doing the best for your business.
One of the critical errors that people make when it comes time to collect some concrete data and that is making to broad of changes which are un-measureable. There is a time where you will use broad changes and test pages but I'll talk about that in a second. When you want to improve your website for conversions etc., you will need to make small changes but test different variations of the thing you want to test. For example, background colour, you want to see if red is better than blue, green or white. You create the pages, put them up for testing and see which one works best. Providing you collect an adequate amount of data (Discussed later) you will conclusively know which one works best for the page AT THAT TIME. There is a reason why you test one thing at a time compared to multiple things at once and that is because it measurable when testing one thing. By testing one thing at a time you can conclusively say that X is the reason why my page is converting more. If you change 4 things at once, put 5 different variations up, collect all the data, you can not conclusively pinpoint which change on the page is responsible for the increase/decrease in conversions.
For you to be able to measure something there is a general rule for statistics, the bigger the sample (Data collected) the more accurate you can make a conclusion. For example, if you send 30 people to each of your pages and draw your conclusions from a sample size that big, it will not nearly give you a conclusive answer compared to 1000 people. The more data you can collect before you make a decision the better. Obviously this is business and we do not have the luxury of collecting tons and tons of data because it would take too long and be cost in-effective. This is why you test pages till you can see that there is a page out performing the rest, if two pages are performing equally or very closely then put your efforts into them two pages to see which one performs the best.
If you really want to you can perform statistical tests on your data but I will not be going into them because some are quite a bit of effort to explain, instead I will mention a few which may help. T test, Chi-squired, Mann Whitney U, one way ANOVAs and two way ANOVAs. These are just a few of tests which could be used, research them and see which applies best to you.
So when is a broad test used? When you start a website of course, you may get 5-10-50 different variations of a sales page made up, you send traffic to them and decide which ones are performing best. From there you narrow down the field until you get your final page which you know needs further testing to be optimised to its best.
So what is the reason of this post? To tell you to test! Testing is never ending and can make a huge difference in the amount of money earned from a page. Just remember this, only because you tested something, eg, background colour, it doesn't mean that background colour shouldn't be re tested after you have made other changes you have tested.
By testing it takes emotions and thinking out of the equation, it gives you raw data which is used to make decisions for you. It tells you EXACTLY what is best for you to do, if you don't like what the data is telling you, tough, it is there to tell you what is best, if you choose to make the change, well, that is up to you.
Easy email marketing automation without moving your lists.
Join Next Live Mastermind Zoominar 100% Real World Secrets to Get Up And Running. Are you Stuck? Donât miss it www.MonthlyMastermind.org
Then you need this: 21 Days To Business Success
Two Signature lines for rent.