Home > Networking Hub >

Testing in mastermind (5)


02-05-2015 07:00 PM #1 jest1234 (AMC Alumnus)
Testing in mastermind

Hi,

can you help me please with simple question. We are 4 people in mastermind, so what is the best way for testing. For example we have 8 banners, 3 LP, 2 offers and we will spend 50 USD each per day.(50x4=200USD)
A) each of us tests 50usd on all 8 banners,3 LPs and 2 offers
B) each of us tests 50 usd on 2 banners, 3 LPs and 2 offers

From easy to find a winner is probably B better, but how knows. Is there any other possibility?


02-05-2015 07:15 PM #2 shakedown (Member)

You guys are doing it quite different than most masterminds by sharing a single set of creatives. It makes it kind of tough to give advice. Do not do plan B though! 2 banners per campaign is a sure way to fail. Plan A sounds better although its still not ideal.

I would assign each person 1 split test and share the insights. Since you guys have all the same creatives, you can test stuff like bids,headlines,offers,countries.

Besides that, I don't have much advice as I like to test angles first. When an angle loses, all the banners and landers associated with it are gone.


02-05-2015 07:23 PM #3 jest1234 (AMC Alumnus)

So you would advise us that we all test let say 2 different angels(thats 8 angels) and than pick two winners and test it from there.
In this approach we should let say mad 8 banners each, thats 32 banners. Is this correct or i am on the wrong way.


02-05-2015 07:39 PM #4 jest1234 (AMC Alumnus)

What is the best way to use mastermind for testing? We are all noobs


02-05-2015 07:48 PM #5 shakedown (Member)

Quote Originally Posted by jest1234 View Post
So you would advise us that we all test let say 2 different angels(thats 8 angels) and than pick two winners and test it from there. In this approach we should let say mad 8 banners each, thats 32 banners. Is this correct or i am on the wrong way.
Yeah, now you are on the right track. Testing 8 angles and 32 banners between you guys will have much better results than the previous plan.

There is no one right answer but this is how I would do it:

Run that test until you get statistically significant data. I like to use this. If 1 angle is the clear winner, then you guys can all work on that angle and move on to testing landers,banners,offers etc like you had planned. If some of the angles show promise, you guys can split the ones who show promise and try to get one of those profitable.


Home > Networking Hub >