: Test Your Headlines for Maximum Profits by: George Dodge Professional copywriters do not simply write or select one headline and then hope for the best. They create a number of different headlines; often times as many

Posted in: #Marketing #Headline #Headlines #Test #Response #Testing #Split #Change #Number #Contro

Test Your Headlines for Maximum Profits

by: George Dodge

Professional copywriters do not simply write or select one headline and then hope for the best. They create a number of different headlines; often times as many as a hundred before selecting the headline they think will perform the best.
But they don't stop there! Professional copywriters know that their choice will often not be the choice of the public and therefore they test their headlines in the market place.
Believe it or not, even professional copywriters find that half or more of their ads, sales letters, or campaigns fail to become winners. But, by testing, they are able to cut their losses early and maximize their successes.
The method of testing most commonly used is known as split testing. With headlines, split testing involves the exposing of two different alternating headlines to prospects. This can be done both online and offline, but online is much easier and the testing process is quicker.
Here is the way headline split testing works online. Visitor A comes to your website and sees headline #1 while visitor B comes to your website and sees headline # 2. When visitor C comes to your website, headline # 1 is shown again and the alternating process continues for the length of the test.
After a number of actions have been recorded for each headline, such as clicking through to the order page, the results are compared to determine which headline resulted in the largest number of desired responses (the click through to the order page, in this example).
The larger the number of total responses recorded, the greater the accuracy of the test to determine the headline winner. Once a winner is chosen, that headline becomes the new control. A control is the current best performing item being tested, in this case, a headline.
Now that a control has been established, a new headline is tested against the control to see if the response can be improved. This process continues on until a given headline has been the control for a large number of tests.
Initially the two headlines that are being tested can be quite different. What you are looking for in the early test is what general kind of headline seems to work for the target market. Once a general kind of headline appears to be a winner over competing types of headlines, then you begin to tweak and split variations of the winning headline itself.
Let's take a look at an example.
This headline example comes from Jimmy D. Brown, a very successful Internet marketer. The following two headlines were split tested to determine a winner.
Headline # 1: "The Power of Viral eBooks"
Headline # 2: "How to Create Automated Profit Generators"
The Result? Headline # 2 out pulled headline # 1 by multiple times.
Now to simplify this process, let's assume for the sake of this example that Headline # 2 proved to be the winner, not only of this test, but a number of follow on test against a number of other headlines. So, you have determined the type of headline that seems to work well for your particular target market. Now, it is time to tweak the headline by split testing variations of this particular headline.
When you split test variations of a headline, you want to only change one thing at a time. Perhaps you will change only one word in the headline, or the color of the headline, or the using of quote marks around the headline, etc. But you only test one item at a time. Each item you test is called a variable; it is the thing that varies between the headlines in the test.
Everything in the headline can make a difference in the headline response rate. Things that make a difference in response are called response modifiers. In addition to the items mentioned above, other response modifiers in a headline might include the font used, the size of the font, the capitalization of the first letters of each word, etc. Everything in your headline can and should be tested.
For an example of how a very minor change can make a difference in your headline response, take a look at these two headlines:
Headline # 1: "Put Music In Your Life"
Headline # 2: "Puts Music In Your Life"
The only difference between the headlines is the addition of the letter "s" to the first word. Again the second headline greatly out pulled the first headline. Some copywriters assume that this is because the first headline implies some work on the part of the reader, while the second headline implies that something or someone else does the work. But the reason is not important, only the results matter.
Each tweak of your headline that results in a new control improves the desired response rate of your headline leading to a maximization of profits, if your headline involves the selling of a product or service.
You can never know for sure why one headline out pulls another, but while you can guess, the reason is not important. What is important is the results, no matter what the reason.
Jim and Audri Lanford of netrageous.com, now scambusters.org, once made a change of color to a sub headline on one of their online sales pages and orders almost dropped to zero. When they changed the headline color back, their normal order rate returned. They guessed that this was because the color change may have caused the headline to look more like hype, but as noted above, the reason is not important. What are important are the results.
Are you testing your headlines? If not, you should be. Every little positive change continues to increase your profits and it all adds up in the end. While some changes may only increase your desired response by a fraction of one percent, some people have seen changes of as much as 1,500% with a single headline change while the rest of their copy remained the same.
How would you like to see either a single or cumulative improvement over time of 1,500% in your desired response rate? I thought so! Then go forth and test, test, and test some more!