Find out how much you should spend based on math rather than hearsay.

Jonathan Beaton · 01-31-2011, 09:46 AM

OK, I always see questions in the forum asking how much you should spend testing an offer.

Rather than basing your budget on an answer from someone who may or may not know what they are talking about, try doing this:

Two Examples.

Example 1- How to determine the confidence interval of an already ran campaign.

Go to: Confidence Interval for Proportions Calculator

Now calculate your Confidence Internal (yes this is math but the website does it for you)

How to:

Step 1: Select your desired confidence level(I usually go with 90%)

Step 2: Let's say you ran a CPC test and had 100 clicks. Enter 100 into the sample size.

Step 3: Now let's say, from those 100 clicks you had 3 conversions.
Enter 3 into the observed survey result.

Now calculate.

In the example above, you can now determine that true population proportion ranges between .19% and 5.18%

Example 2 - How to estimate a desired sample size

Go here: Sample Size for Proportion Calculator

Step 1: Enter your desired confidence level (again, I usually use 90%)

Step 2: Enter your estimated conversion rate

Step 3: Enter your desired confidence interval (I typically use 2)

Now Calculate and you will see the minimum sample size needed.

williamrs · 01-31-2011, 12:42 PM

Great! Thanks a lot for sharing! Math always helps a lot. However, there are many subjective factors to be considered in a campaign, and, if I'm not wrong (maybe I am

), these numbers would apply if we were sure about the quality of the elements that we are using (ads, landing pages, etc), so we would use these tools to determine how much to spend to test the campaign and then we would know if it works or not. But what if the landing page is not good? Then I can follow these numbers and decide that the campaign isn't profitable, but I'd never know that my conversion rate would be better with a new landing page.

This is the reason why I think that tools can help, but they can never do everything for us, a campaign always needs the "human touch", a more subjective analysis.

William

Jonathan Beaton · 01-31-2011, 12:56 PM

Quote:

Originally Posted by williamrs

Great! Thanks a lot for sharing! Math always helps a lot. However, there are many subjective factors to be considered in a campaign, and, if I'm not wrong (maybe I am

), these numbers would apply if we were sure about the quality of the elements that we are using (ads, landing pages, etc), so we would use these tools to determine how much to spend to test the campaign and then we would know if it works or not. But what if the landing page is not good? Then I can follow these numbers and decide that the campaign isn't profitable, but I'd never know that my conversion rate would be better with a new landing page.

This is the reason why I think that tools can help, but they can never do everything for us, a campaign always needs the "human touch", a more subjective analysis.

William

It will tell if you the campaign in its current format is working or not. It is still up to the IM to determine if the problem is the landing page, offer, etc.

Freddie · 02-02-2011, 03:23 AM

Quote:

Originally Posted by SpendTooMuchTimeOnline

It will tell if you the campaign in its current format is working or not. It is still up to the IM to determine if the problem is the landing page, offer, etc.

So, to improve landers, ads and offers we should split test, and to analyze the results accurately - instead of guessing based on hearsay - we can use this little calculator that complements the other ones nicely (courtesy of Perry Marshall).

:::SplitTester.com:::

This one will tell you the confidence level so that you can accurately determine which lander/ad/offer is winning your split test.

Kenster · 02-02-2011, 09:56 AM

I love "scientificating" as much as possible. The more you can systematize your business based on objective inputs, the more reliably you can grow your business.

Of course there is subjectivity all over this business as many aspects involve creativity, but you can still base the subjectivity on formulas!

01-31-2011, 09:46 AM	#1
Jonathan Beaton Internet Marketing Guide War Room Member Join Date: Dec 2010 Location: Tampa, Florida Posts: 256 Thanks: 29 Thanked 56 Times in 32 Posts Social Networking Contact Info	Find out how much you should spend based on math rather than hearsay. OK, I always see questions in the forum asking how much you should spend testing an offer. Rather than basing your budget on an answer from someone who may or may not know what they are talking about, try doing this: Two Examples. Example 1- How to determine the confidence interval of an already ran campaign. Go to: Confidence Interval for Proportions Calculator Now calculate your Confidence Internal (yes this is math but the website does it for you) How to: Step 1: Select your desired confidence level(I usually go with 90%) Step 2: Let's say you ran a CPC test and had 100 clicks. Enter 100 into the sample size. Step 3: Now let's say, from those 100 clicks you had 3 conversions. Enter 3 into the observed survey result. Now calculate. In the example above, you can now determine that true population proportion ranges between .19% and 5.18% Example 2 - How to estimate a desired sample size Go here: Sample Size for Proportion Calculator Step 1: Enter your desired confidence level (again, I usually use 90%) Step 2: Enter your estimated conversion rate Step 3: Enter your desired confidence interval (I typically use 2) Now Calculate and you will see the minimum sample size needed.
	My Internet Marketing Blog

01-31-2011, 12:42 PM	#2
williamrs Senior Warrior Member War Room Member Join Date: Aug 2009 Location: Brazil Posts: 3,496 Thanks: 174 Thanked 1,159 Times in 614 Posts	Re: Find out how much you should spend based on math rather than hearsay. Great! Thanks a lot for sharing! Math always helps a lot. However, there are many subjective factors to be considered in a campaign, and, if I'm not wrong (maybe I am ), these numbers would apply if we were sure about the quality of the elements that we are using (ads, landing pages, etc), so we would use these tools to determine how much to spend to test the campaign and then we would know if it works or not. But what if the landing page is not good? Then I can follow these numbers and decide that the campaign isn't profitable, but I'd never know that my conversion rate would be better with a new landing page. This is the reason why I think that tools can help, but they can never do everything for us, a campaign always needs the "human touch", a more subjective analysis. William
	>> STOP... << Stop struggling alone. Stop wasting time. Stop being scammed. Only 1 SPOT LEFT - CLICK HERE

02-02-2011, 09:56 AM	#5
Kenster Krazy Kenster War Room Member Join Date: Jan 2009 Posts: 5,111 Blog Entries: 11 Thanks: 672 Thanked 2,009 Times in 885 Posts	Re: Find out how much you should spend based on math rather than hearsay. I love "scientificating" as much as possible. The more you can systematize your business based on objective inputs, the more reliably you can grow your business. Of course there is subjectivity all over this business as many aspects involve creativity, but you can still base the subjectivity on formulas!
	Six Figure Alliance Training/Coaching Course... Kenster's Six Figure Mission Blog \| Kenster's Six Figure Alliance \| WarriorJV - WSO Launches