CPVLab Discount: Save $50!
  #1  
Old 03-17-2010, 07:25 AM
Senior Member
 
Join Date: Jan 2010
Posts: 116
Thanks: 0
Thanked 1 Time in 1 Post
Default Testing Scenario

Hey Guys,

I read every post about testing ,but would like to double check on a scenario.

Some say to go by impressions (1000 per offer before ditching or not) and others go by the spend in relation to the payout per url.

Okay, what do you think..

I have spend 20 bucks on an auto insurance url, which is rotating on 3 offers. Impressions come from one url (the others are not doing anything, but I have stopped that sucker) and cost 31. Hence i have gotten 66 impressions without a conversion.

The payout is at $10.

Now, payout wise, if I go for 100% ROI, I should have dumped the url by the time I lost 15 bucks, right? (cause I have three offers, and half the payout would be $5).

But what about impressions? 66 for 3 offers are quite few to make a judgement, no?

How do I make a decision based on this? I mean, we are not even close to having 1000 impressions. But it would cost $310 to test just one offer, and one url (1000x 31cents). So that would be $930 for the entire rotation per url. And IF I had 5 urls to test at the same cost per impression that would be $4650. Or perhaps I am calculating wrongly and it should be 1000 impression on the entire 3-pack rotation? No clue...



What are your thoughts?


Vincent
Reply With Quote
  #2  
Old 03-18-2010, 02:26 AM
Administrator
 
Join Date: Apr 2008
Posts: 1,362
Thanks: 0
Thanked 15 Times in 8 Posts
Default

I always go with dollar amounts. Who cares if it hits 1,000 impressions. Everyone could and you end up losing even more if you go by that. The only time I would ever go by impressions is if I was going to try to optimize my ads and landers big time.
Reply With Quote
  #3  
Old 03-18-2010, 07:40 AM
Senior Member
 
Join Date: Jan 2010
Posts: 116
Thanks: 0
Thanked 1 Time in 1 Post
Default

yeah, you are right. It only makes sense that way. Though I will at some point of time hit check out statistical significance and all that other BS. But first ...more practical :-) thanks Will
Reply With Quote
  #4  
Old 03-18-2010, 10:16 AM
Senior Member
 
Join Date: Jan 2010
Posts: 116
Thanks: 0
Thanked 1 Time in 1 Post
Default

Oh, another scenario.

I now know you go by dollar amounts per url, but what about per offer? I mean, if you have 5 urls and they don't perform at 50% of the payout, then you dump it. What if all 5 urls don't trigger a conversion. How do you decide whether you go for more testing, or whether the camp just doesn#t work...

cause that might be where impressions overall are relevant? I have no clue...
Reply With Quote
  #5  
Old 03-22-2010, 02:27 AM
Administrator
 
Join Date: Apr 2008
Posts: 1,362
Thanks: 0
Thanked 15 Times in 8 Posts
Default

Right but now you're back to the fact that you have to pay per impression, so while yes you could get a conversion at 1,000 impressions, it's pointless. Those 1,000 would most likely cost you far more then you're getting back.

Know what I mean?

Reply With Quote
  #6  
Old 03-22-2010, 07:10 AM
Senior Member
 
Join Date: Jan 2010
Posts: 116
Thanks: 0
Thanked 1 Time in 1 Post
Default

yes, i know what you mean, but what if we did urls in a totally wrong fashion? When do you decide whether to test more or not if your first 5 urls batch is not performing well. Or do you drop it completely if they don't convert, on the basis of the assumption that your first 5 urls are the most relevant type of urls and if they don't convert probabyl nothing else converts.

Or you could think: Okay my first 5 urls didn'T work. No conversion, but perhaps i did the wrong type of targetting, and the next 5 could be all winners?

let me knwo what you think. I am learning to optimized/ make decision with my campaigns that I built.

vincent
Reply With Quote
  #7  
Old 03-24-2010, 02:34 AM
Administrator
 
Join Date: Apr 2008
Posts: 1,362
Thanks: 0
Thanked 15 Times in 8 Posts
Default

Originally Posted by vintres View Post
Okay my first 5 urls didn'T work. No conversion, but perhaps i did the wrong type of targetting, and the next 5 could be all winners?
That is what I think.

Reply With Quote
  #8  
Old 03-24-2010, 07:43 PM
Member
 
Join Date: Nov 2009
Posts: 71
Thanks: 0
Thanked 2 Times in 2 Posts
Send a message via AIM to kevinkimes
Default

I think you need to go with cost per URL per offer. Even if all the offers pay the same, one might convert 100 times better than another.

If on one URL you spend $15, evenly divided among 3 offers, but these offers all pay $20 (assuming a goal of 100% ROI), you've got to keep going until you've spent enough on each offer for that URL. That can get expensive when testing a pile of URLs with multiple good paying offers.

10 URLs, 4 offers, $20 payout, 100% ROI goal.
10*4*10=$400.

However, if all the target URLs are highly similar, and if the offers are all variations of the same type, and if they are getting somewhat even traffic distribution, you might be able to put a test like that on hold if you get nothing back after spending a lot less than $400, since it could be that the offer type isn't going to work at all with that group of URLs. I say on hold, since you could come back later, when you're making more money, to finish the test.

I'm testing with bigger URL groups, and pausing the URLs that get a bunch of traffic right away without conversions. I'm just testing on LeadImpact, so I can't rotate offers. Instead I enter two destinations and use the same URL list for each one.

For me, when I get an offer that seems to convert, I'll test more ways of promoting it instead of splitting or rotating offers for now. When I've got more money to test with, I could probably keep rotating. I feel like the video of testing a bunch of different types of offers at once is somewhat misleading, and not how most of the successful folks here would do their testing in the early stages of doing CPV. That is proabably a great way if you can afford to spend (#URLs)*(#offers)*(payout/((roi/100)+1)).
Reply With Quote
  #9  
Old 03-25-2010, 01:38 PM
Senior Member
 
Join Date: Jan 2010
Posts: 116
Thanks: 0
Thanked 1 Time in 1 Post
Default

Thanks a lot Kevinkimes,

your tips are very helpful!

So you mostly rotate 2 offers? How many urls to test? And how do you scale once you see an offer that seems to convert?
Reply With Quote
  #10  
Old 03-26-2010, 03:41 AM
Member
 
Join Date: Nov 2009
Posts: 71
Thanks: 0
Thanked 2 Times in 2 Posts
Send a message via AIM to kevinkimes
Default

I should clarify that I'm not crazy successful yet at this, like Aaron and Coach. Heheh. But I think I have hold of the basic concepts. I've only made a few bucks so far. Positive ROI on some campaigns, but not even half way to four digits of net income. I was out of the country for awhile, and it was much harder to work on this from where I was.

Right now I am doing one offer per campaign mostly, because my budget isn't big and i'm getting a feel for what works.

I am also focusing on trying to just push stuff out to see what sticks. I'll pick an offer, and think of 2-3 ways to promote it, grab a bunch of URLs related to those promotion ideas, and throw it online.
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT. The time now is 03:44 PM.


Search Engine Optimization by vBSEO 3.6.0