You hear it all the time, “Test everything.” The problem is, many marketers who are just starting out really don’t know the proper way to test anything. Yes, there is a proper way. I’m not saying there is one way. I am saying there is a right way…many right ways. What you want to do is stay away from the wrong ways. In this article, I’m going to explain what the right ways are.

Let’s first start off with the wrong way so you can get a clear idea of what not to do. Let’s say you’re testing a sales page for conversion and you want to see if you can increase that conversion. This is a common enough practice. Okay, what are you going to do? Well, there are many things you can do. A sales page is composed of quite a few parts such as your headline, bullet points, guarantee, price and so on. Now, the wrong way to test this would be to change more than one item at a time. You’re probably wondering why this is wrong. Let me explain.

Let’s say you change your headline and your price. These are two very key elements to your sales page because the headline is responsible for 80% of your sales and the price is probably one of the biggest determining factors to making the sale. If the price is too high, people may not buy because they can’t afford it. If the price is too low, they may think that it’s a cheap product. So, if you change both of these elements and the conversion goes up, you have a problem. You don’t know if the conversion went up because of the headline, the price or both. The conversion may have actually gone up more or less if you changed only one of the items.

Now, let’s say you’re testing out the best times to send email ads. Well, you want to conduct this test over a period of time, at least a week. One day isn’t going to mean anything. That particular day may just be a good or bad day anyway. If after a week of emailing at 7 AM and then a week of emailing at 7 PM, you see the 7 AM mailings get a better response, it’s pretty safe to assume that the 7 AM time slot is better for sending email ads.

What you should have gotten out of this article is two things. One, you should only test changing one thing at a time and two, you should test for a long enough period of time to make sure your results are accurate.

If you do this, you’ll find that your testing becomes a lot more meaningful.

To YOUR Success,

Steven Wagenheim

Tired of busting your behind for peanuts online? Go to my web site and find out how I earn a monthly income that exceeds 5 figures and how I can help YOU do the same. Get your free report at http://www.mysecretarticles.com/report.html