When To Stop Testing Your Ads
Often I get the question, "When do I stop my testing my ads?" Here's the short answer.
Before I ever start, I set a goal for the number of sales (or leads) that need to be achieved. The idea of just testing to see how high is high I find to be counter-productive. There should be a valid business reason to try to increase conversions and response. If you know that a 10% increase will put $X to the topline, you then have a "reason why", the "reward" of the test and the minimum to declare success.
If I am trying to get 10 new customers a day for a segment of a campaign and my baseline is 5, obviously I am looking for a 200% increase. Once I have achieved 200%, I usually lock the test and channel down and don't change anything until the market tells me that I need to redo the test to get more out of the traffic. Once an ad is working well and stable. I recommend to move on to the next marketing channel. It is often better to have several channels running well than it is to place all of your faith and income into one marketing channel and campaign. Markets change. I recommend that you spread out to reduce risk and protect revenues.
The basic metrics that need to be monitored for a campaign are cost per new customer, cost per sale, average order, number of new customers, cost of promotion and number of clicks required for a sale. Those six metrics are, I believe, the most important to monitor. Costs down or stable while increasing profits is the ideal situation for a marketer that is truly concerned about driving profits. Too often to purchase (or add) more traffic is the easy answer to falling sales figures.
In essence, most look to overcome low or poor conversions with more traffic and exposure. You can run tests with only 100 clicks per day and double sales in a few short months.
Have an end in mind before you start testing and know what that "end" means to your business. Once you will have a clear target your testing will become more focused and meaningful.
To Your Next Successful Test,
David Bullock
Increasing Conversion Rates One Test At A Time
Before I ever start, I set a goal for the number of sales (or leads) that need to be achieved. The idea of just testing to see how high is high I find to be counter-productive. There should be a valid business reason to try to increase conversions and response. If you know that a 10% increase will put $X to the topline, you then have a "reason why", the "reward" of the test and the minimum to declare success.
If I am trying to get 10 new customers a day for a segment of a campaign and my baseline is 5, obviously I am looking for a 200% increase. Once I have achieved 200%, I usually lock the test and channel down and don't change anything until the market tells me that I need to redo the test to get more out of the traffic. Once an ad is working well and stable. I recommend to move on to the next marketing channel. It is often better to have several channels running well than it is to place all of your faith and income into one marketing channel and campaign. Markets change. I recommend that you spread out to reduce risk and protect revenues.
The basic metrics that need to be monitored for a campaign are cost per new customer, cost per sale, average order, number of new customers, cost of promotion and number of clicks required for a sale. Those six metrics are, I believe, the most important to monitor. Costs down or stable while increasing profits is the ideal situation for a marketer that is truly concerned about driving profits. Too often to purchase (or add) more traffic is the easy answer to falling sales figures.
In essence, most look to overcome low or poor conversions with more traffic and exposure. You can run tests with only 100 clicks per day and double sales in a few short months.
Have an end in mind before you start testing and know what that "end" means to your business. Once you will have a clear target your testing will become more focused and meaningful.
To Your Next Successful Test,
David Bullock
Increasing Conversion Rates One Test At A Time
0 Comments:
Post a Comment
<< Home