the control, then you make it the new control and continue with testing another variation of the headline or a different variable. Perhaps you might next test whether the color of the headline makes a difference. Suppose that changing the color of the headline resulted in another 1% increase in sales to a 3% response rate. This would be another 50% increase in sales and would be triple the number of sales averaged from the original sales letter!

It is very easy to be complacent with a working sales letter and not test further, but this can be a costly mistake. Any change in a variable could result in higher profits. For example, suppose that you determined that you consistently had a 1% response rate when selling your product for $97 but a 2% response rate when selling the product fo

What if you could double your sales overnight by applying just one Internet marketing secret?

One of the greatest secrets of direct marketing is that it is scientific. With many forms of advertising, you cannot tell which ad or marketing campaign produced the results. But smart direct marketers know exactly which ad or sales letter produced what results because the results are tracked and measured.

Direct marketers have a secret they have been using to double their income - or more - for a hundred years. And when this marketing secret is combined with the power and speed of Internet marketing, the results can be phenomenal. It is not an exaggeration to say that you could potentially double your sales overnight with this one Internet marketing secret.

What's the secret? Split testing. This is done by using a "key" to associate a sale with the ad or sales letter that produced the sale so that you can measure the results. In a print ad, you might provide a phone number with an extension. The extension is actually the key. You would have two ads or sales letters, each with a different key, and would then track the results using their respective keys. Internet marketers typically use tracking software to associate a key with the resulting sales, although it can be done as simply as having two sales pages and tracking the sales from each.

For accurate results, you need a test that is big enough that you can rely on your statistics. For example, with a direct mail sales letter this is generally considered to be a mailing of at least 1000. Internet marketers can substitute 1000 page views. Either way, once you know what results you can consistently expect from an ad or sales letter, it becomes what direct marketers call your "control." You always need to have one control that produces known results. Then you can test other versions against the control to see if you can consistently beat the results of the control.

Suppose that you have a Web sales letter that consistently produces an average of 10 sales for every 1000 times the page is viewed. This is a 1% response, a figure that is often cited as average for a direct response sales letter.

Now you need to test one variable at a time to see if you can increase the percentage of sales. For example, you may want to test the sales letter with a different headline, a different graphic, a different price, different bonuses, different text, etc. But always only test one variable at a time. Generally, the headline is considered to be the most important element to test first.

What you want to determine is whether any change in a single variable will consistently produce a higher response. The payoff can be quite significant. Imagine if a different headline was determined to produce a 2% response. That doesn't sound like much of a change - only 1% more. But this would double your sales! Instead of making 10 sales out of a thousand, you would be making 20 sales.

This is a very realistic scenario, so you always want to test against your control. The better your control, the more tests will fail to beat it. But once you find a sales letter that consistently beats the control, then you make it the new control and continue with testing another variation of the headline or a different variable. Perhaps you might next test whether the color of the headline makes a difference. Suppose that changing the color of the headline resulted in another 1% increase in sales to a 3% response rate. This would be another 50% increase in sales and would be triple the number of sales averaged from the original sales letter!

It is very easy to be complacent with a working sales letter and not test further, but this can be a costly mistake. Any change in a variable could result in higher profits. For example, suppose that you determined that you consistently had a 1% response rate when selling your product for $97 but a 2% response rate when selling the product fo

it testing. This is done by using a "key" to associate a sale with the ad or sales letter that produced the sale so that you can measure the results. In a print ad, you might provide a phone number with an extension. The extension is actually the key. You would have two ads or sales letters, each with a different key, and would then track the results using their respective keys. Internet marketers typically use tracking software to associate a key with the resulting sales, although it can be done as simply as having two sales pages and tracking the sales from each.

For accurate results, you need a test that is big enough that you can rely on your statistics. For example, with a direct mail sales letter this is generally considered to be a mailing of at least 1000. Internet marketers can substitute 1000 page views. Either way, once you know what results you can consistently expect from an ad or sales letter, it becomes what direct marketers call your "control." You always need to have one control that produces known results. Then you can test other versions against the control to see if you can consistently beat the results of the control.

Suppose that you have a Web sales letter that consistently produces an average of 10 sales for every 1000 times the page is viewed. This is a 1% response, a figure that is often cited as average for a direct response sales letter.

Now you need to test one variable at a time to see if you can increase the percentage of sales. For example, you may want to test the sales letter with a different headline, a different graphic, a different price, different bonuses, different text, etc. But always only test one variable at a time. Generally, the headline is considered to be the most important element to test first.

What you want to determine is whether any change in a single variable will consistently produce a higher response. The payoff can be quite significant. Imagine if a different headline was determined to produce a 2% response. That doesn't sound like much of a change - only 1% more. But this would double your sales! Instead of making 10 sales out of a thousand, you would be making 20 sales.

This is a very realistic scenario, so you always want to test against your control. The better your control, the more tests will fail to beat it. But once you find a sales letter that consistently beats the control, then you make it the new control and continue with testing another variation of the headline or a different variable. Perhaps you might next test whether the color of the headline makes a difference. Suppose that changing the color of the headline resulted in another 1% increase in sales to a 3% response rate. This would be another 50% increase in sales and would be triple the number of sales averaged from the original sales letter!

It is very easy to be complacent with a working sales letter and not test further, but this can be a costly mistake. Any change in a variable could result in higher profits. For example, suppose that you determined that you consistently had a 1% response rate when selling your product for $97 but a 2% response rate when selling the product fo

can substitute 1000 page views. Either way, once you know what results you can consistently expect from an ad or sales letter, it becomes what direct marketers call your "control." You always need to have one control that produces known results. Then you can test other versions against the control to see if you can consistently beat the results of the control.

Suppose that you have a Web sales letter that consistently produces an average of 10 sales for every 1000 times the page is viewed. This is a 1% response, a figure that is often cited as average for a direct response sales letter.

Now you need to test one variable at a time to see if you can increase the percentage of sales. For example, you may want to test the sales letter with a different headline, a different graphic, a different price, different bonuses, different text, etc. But always only test one variable at a time. Generally, the headline is considered to be the most important element to test first.

What you want to determine is whether any change in a single variable will consistently produce a higher response. The payoff can be quite significant. Imagine if a different headline was determined to produce a 2% response. That doesn't sound like much of a change - only 1% more. But this would double your sales! Instead of making 10 sales out of a thousand, you would be making 20 sales.

This is a very realistic scenario, so you always want to test against your control. The better your control, the more tests will fail to beat it. But once you find a sales letter that consistently beats the control, then you make it the new control and continue with testing another variation of the headline or a different variable. Perhaps you might next test whether the color of the headline makes a difference. Suppose that changing the color of the headline resulted in another 1% increase in sales to a 3% response rate. This would be another 50% increase in sales and would be triple the number of sales averaged from the original sales letter!

It is very easy to be complacent with a working sales letter and not test further, but this can be a costly mistake. Any change in a variable could result in higher profits. For example, suppose that you determined that you consistently had a 1% response rate when selling your product for $97 but a 2% response rate when selling the product fo

a different price, different bonuses, different text, etc. But always only test one variable at a time. Generally, the headline is considered to be the most important element to test first.

What you want to determine is whether any change in a single variable will consistently produce a higher response. The payoff can be quite significant. Imagine if a different headline was determined to produce a 2% response. That doesn't sound like much of a change - only 1% more. But this would double your sales! Instead of making 10 sales out of a thousand, you would be making 20 sales.

This is a very realistic scenario, so you always want to test against your control. The better your control, the more tests will fail to beat it. But once you find a sales letter that consistently beats the control, then you make it the new control and continue with testing another variation of the headline or a different variable. Perhaps you might next test whether the color of the headline makes a difference. Suppose that changing the color of the headline resulted in another 1% increase in sales to a 3% response rate. This would be another 50% increase in sales and would be triple the number of sales averaged from the original sales letter!

It is very easy to be complacent with a working sales letter and not test further, but this can be a costly mistake. Any change in a variable could result in higher profits. For example, suppose that you determined that you consistently had a 1% response rate when selling your product for $97 but a 2% response rate when selling the product fo

the control, then you make it the new control and continue with testing another variation of the headline or a different variable. Perhaps you might next test whether the color of the headline makes a difference. Suppose that changing the color of the headline resulted in another 1% increase in sales to a 3% response rate. This would be another 50% increase in sales and would be triple the number of sales averaged from the original sales letter!

It is very easy to be complacent with a working sales letter and not test further, but this can be a costly mistake. Any change in a variable could result in higher profits. For example, suppose that you determined that you consistently had a 1% response rate when selling your product for $97 but a 2% response rate when selling the product for $67. For every 1000 page views, you would make $970 at $97 per sale but $1340 at $67 per sale.

Always track your response rate, always keep split testing against your control looking for a new control, and always consider the bottom line - your net profit. Many tests will fail - they will not beat your control. But once you have a successful test that consistently beats your control, you will likely have significantly increased your income. Without split testing, you will never know how much money you are losing. But with this Internet marketing secret, you can potentially double your sales overnight.