POF N3WB Question:

[niftyg]

Hey everyone,

I'm preparing to launch a POF campaign and have some questions as I am partially lost, Please let me know what is the best course of action.... and on with the q's.

Here is what the current plan is.

I have what I believe is a solid landing page that I ripped off (very solid) and a solid offer, For this round as I am a very beginner and somewhat confident in these two and wouldn't like to change them. Let's just focus on the below as I am new to split testing.

My plan is to...
Target many small groups with a few creatives. Here is an example:
US Male 21-27 (4 Creatives)
US Male 28-34 (4 Creatives)
etc, etc (more countries, more genders.)
....

Now here are my questions if my targeting are right or not.

1. Target multiple groups like I have up there to figure out which creatives work or not, or go broad? I hear broad burns through your budget faster, however with lots of campaigns I don't know.

2. What are good countries to start targetting, how many, and how do I figure this out?

3. What sort of budget for each campaign to get proper results, and how much CPM bid should I start at?

4. How many ad creatives should I start per group?

5. What OTHER attributes to biz op adds usually use if anybody knows off the top of there head. I am not starting with dating as I have had success with other types of offers on POF so far.


This covers most of my questions for now


I very much appreciate any response and look forward to them

Talk soon.

Grant.

 

[BenPOF]

What's your budget Grant? One thing newbies to is blow their budget trying to test too much at one time. You're much better off fully split testing and learning about a few demo's than half assing and learning a little about a lot of demo's.

Feel free to hit up myself (ben@pof.com) or Cyndi (cyndi@pof.com) if you want to talk about details that may not be suitable for public eye

 

[android2102]

Im new too so this could be a case of the blind leading the blind but this is what makes sense to me. Maybe some more experienced guys could chime in.

1. As Ben said, definitely limit demos when you're new and then when you find success with niche demos broadening the targeting could be an option for scaling up. I consider what you have listed, US Male 21-17 as broad targeting, maybe if you added ethnicity or something you could tighten up your targeting.

3. When testing it's easy to get real complicated and when I start a campaign I like to vary only a few variables at a time. I choose a CPM at the higher range when I'm focusing on improving ad CTRs. Later when I know CTRs are not an issue (.1 or higher if you want to be safe) I worry about profitability which means adjusting CPMs. The CPM range varies a lot based on your demos so if you have no idea where to start, try alternating bids every hour (change bids to .30 at 1pm then to .4 at 2pm then back to .3 at 3 pm and then .4 at 4pm) and record how many impressions your getting in each hour. You'll be able to tell approximately how many more impressions youre getting for the extra 10 cents. If the two bids are getting almost the same amount of traffic you'll know that the higher bid is a waste of 10 cents and then you'd test .2 and .3 to see if those bids get you different traffic volume. (*these numbers are made up examples*)

4. I find that choosing less than 10 ad creatives doesn't give me enough chance to find winners. Meaning the chance of all 10 ads being a flop is a lot higher than the chance of all 100 ads being a flop, because you have more images you will have a greater chance of finding several winning ads. On the flip side you could end up paying a lot of money waiting for 100 ads to gain enough impressions to know which ads are the winners. Maybe a number between 15-30 ads at a time. then when it's clear which ads are out performing others cull away the bad ones and test new ones. You need to have several (6 or 7 at the least) successful ads running or you might burn them out and their CTRs will plummet

I don't have enough experience so I'm not going to BS an answer to 2 and 5

Andrew