Test Your Marketing with Statistical Significance
One of the best things about internet marketing that everybody knows is how easy it is to test everything, but what I find is that most people give up on their tests way too prematurely. You want to make sure you accumulate enough data to actually be able to accurately judge any landing page, banner ad or website that you launch.
Let me give you an example of how I see a lot of people do this. Let’s say that you’ve designed a new script that you’ve given to two sales people. Both sales people are going to call 20 people in a day and at the end of the day you’re going to decide which is the better script to give to the rest of the sales team.
In the morning both sales people get about six phone calls in before lunch. When they sit down to confer with each other they find out that sales person A had converted on every other phone call or three out of six sales. Sales person B had only converted on one call. The two who talked to me almost gave up on not seeing the afternoon test through because clearly script A was superior.
Now what happened in the afternoon is that sales person B actually found much greater success. They found out that by the end of the day he had converted on almost all of his phone calls except for the ones in the morning. Now if they would have made that decision at lunch time they might have missed out on picking the right script.
This is what I see a lot of people do when they launch a new ad or a landing page or do anything on social media. The magic number I want to recommend to you is to get 1,000 people on any piece of marketing before you make a decision about it. Send people to a landing page 1,000 times or get 1,000 people to click on two different banner ads before you make a decision about which is the better performing ad. That’s called statistical significance and that’s what you need to make a smart decision about your marketing. Test more than you think you need to and test often.