Typical stats for mobile popunder campaigns (6)
Since I am new to mobile popunder traffic I wanted to ask, how the stats should look like for a promising campaign. I tried to apply the rules I learned in the cookbook, but given the traffic is more expensive and CPAs are usually higher, I wanted to make sure before cutting placements too early/late.
1. How many Visits do you let a lander run, to be 95% sure you have enough data? For Banners I used to take at least 2000 visits. Logically this should be the some for pops, right?
2. What is a decent CTR? I have currently ~20 landers varying from 1% - 20% with low xxx visits. For which numbers should i look out, after 2000 visits are reached? (I know that CTR is not as important as conversions. If CR and CV is okay, Ill keep them even if the CTR should be low).
3. As you cut placements after 2-3x CPA spent, can I use this thumbrule also for poptraffic?
4.How many angles/landers do you test when starting a campaign for poptraffic. I started out with 10 different angles, but not sure if this is too much for a first run. I am asking, because I currently run traffic in ZA and they have low volume available, which results data running in slowly spreaded on so many different landers.
5.When runing RON it is recommended to create a new campaign, targeting only promising placements. What excatly do you mean with placements? Only site IDs, or do you also talk about the more promising landers, OS, carrier and even device models? Isnt it too specific and you barely get volume because the targeting is so tight including all of these?
great question, I'm also curious!
me three! Just getting my feet wet on this traffic type. If you two have any other really useful threads , pls share!
@aaaart; Let me try to answer your questions. Please don't shoot me if it's not working the same for you, this is just based on my personal experience.
1. How many Visits do you let a lander run, to be 95% sure you have enough data? For Banners I used to take at least 2000 visits. Logically this should be the some for pops, right?
Normally I would spend 50-100$ a day on 1-3$ cpm geo, and run it 3-5 days, and start optimizing from there. I would kill the campaign if it gives me ROI -80%.
2. What is a decent CTR? I have currently ~20 landers varying from 1% - 20% with low xxx visits. For which numbers should i look out, after 2000 visits are reached? (I know that CTR is not as important as conversions. If CR and CV is okay, Ill keep them even if the CTR should be low).
I dont really look at CTR, but 20% sounds reasonable.
3. As you cut placements after 2-3x CPA spent, can I use this thumbrule also for poptraffic?
Depend on what type offer you are running. If you are promoting .30 CPI offer and you cut placement after you spent .60 cents in one placement, that doesn't give you any data. EPV is actually your friend in running pop, if your EPV is lower than your CPV (which is your CPM bid) then you can cut the placement.
4.How many angles/landers do you test when starting a campaign for poptraffic. I started out with 10 different angles, but not sure if this is too much for a first run. I am asking, because I currently run traffic in ZA and they have low volume available, which results data running in slowly spreaded on so many different landers.
I prefer testing different lander when starting new campaign. If I find winner, I test with different angles. But yes, go with multiple lander straight away but don't test too much lander/angle with small daily budget and over short period. Personally I think 10 is too much, maybe try 3-5 different variation.
5.When runing RON it is recommended to create a new campaign, targeting only promising placements. What excatly do you mean with placements? Only site IDs, or do you also talk about the more promising landers, OS, carrier and even device models? Isnt it too specific and you barely get volume because the targeting is so tight including all of these?
I think in pop, placement is non other than site id, I called the other targeting options.Talking about traffic source (TS) optimization, of course you do whatever it takes (I mean WHATEVER) to make your campaign profitable. Some TS provide carrier targeting, some not, you just have to make use what options available. Don't cut site id too fast if the TS provide many targeting options. Yes, over-optimize will result in small volume, but I will run it as long as the ROI looks good. You can also try to scale your winning campaign to other TS.
All the best for your campaign buddy!
Nice post. Nice answers.
I'm just starting to get into pops and am wondering the same things.
Originally Posted by nuel_022
Originally Posted by nuel_022
I'd be interested to hear what kind of conversion rate people are happy with on a low payout offer. I'm thinking 3% - 10% is a good range on a campaign level.
Originally Posted by nuel_022
If you can narrow down to a proven LP and a proven offer - so the only dynamic variable is the placement - I'm also wondering how long people are prepared to let things run before they've seen enough?
So you're spending $200+ before optimising anything at all?
Yes, because I have already know that my LP performing well and the offer is converting.
No, if I just started with new traffic source. Reason is "click-loss", because some TS has certain zone that will just sending you bot traffic, so you kind of watch it closely. There can be 1 zone that eating all your budget because of this very reason (happened to me).
----
From my own testing 20% would be excellent. More regularly I'm seeing between 7 - 15%. At the moment I'm happy if I'm around the 10% mark. Of course, as is already mentioned, the CTR isn't really the important metric however.
I'd be interested to hear what kind of conversion rate people are happy with on a low payout offer. I'm thinking 3% - 10% is a good range on a campaign level.
I 'm not sure what is my CR. I'm more interested in CV, 1% is GOOD, 2% is MIRACLE (in this case CPI offer). Depend on how aggresize your LP is, hahaha.
----
You're right that pops make things pretty simple when it comes to spotting winners simply by comparing EPV and CPV - but you still need to make that call based on a certain amount of data otherwise it wouldn't be reliable.
If you can narrow down to a proven LP and a proven offer - so the only dynamic variable is the placement - I'm also wondering how long people are prepared to let things run before they've seen enough?
For testing purpose, I run it 3-5 days over the weekend. Set the frequency cap to 1, so it won't burn your daily budget.
Home > Paid Traffic Sources > Mobile