A lot of these mobile ad networks have thousands and thousands of apps and/or mobile sites in their inventory. So, when you run a campaign, initially your ads are going on many of these apps and sites... it's like having a ton of mini-campaigns within one (ie. each app is almost like a campaign in and of itself as opposed to running traffic on just one site or app)... to get any relevant data you gotta have a ton of clicks and eventually conversions on a ton of apps to know what is converting because the ad network rotates your ads through their app inventory... so you get a few here on this app, a few there on that app, etc. This takes gobs of money.
Even if you had 10000 clicks... if the network has 50K apps, 10000 clicks isn't all that much data to work off of because none of the individual apps or mobile sites that you get clicks from is getting enough data to optimize off of. Lets say they rotate your ads just on 10000 of apps... that's just 1 click per app... hardly enough to get good data to work off of. It's basically like doing RON traffic... blindly... with traffic coming in from so many places that no one app or site gets usable data.
I realize you need money to make money but is there a smarter way to go about it?
Another one of the issues is that many of these ad networks don't show you what the app is, they just use ID's and/or you can't target all that well outside the normal device, OS, etc. So, it's tough to make some logical choices in the beginning (i.e. you have a male dominated offer so you don't want to show it on Glamour for example). Traditionally, if you put your campaign say on just 1 site... or even say 10 apps... you could optimize rather quickly... but with these mobile ad networks, it's a bit tough.
Some possible Solutions:
1. Pick an ad network that has really good demographic targeting so at least you get the gender, age, etc. right... and be a little further along in the targeting processing... ie, Millennial Media has some decent demographic targeting (more than others it seems) but they are a little pricey and a bit strict on what they allow.
2. Throw a ton of money at it... which for many on STM (and elsewhere) probably isn't an early option that they can do.
3. Use an ad network that shows the mobile sites/apps that you can advertise on and you can get then target those... and therefore... use logic and pick 5 or 10 sites/apps to target that make sense for your offer and then you can focus on the actual click data and have relevant data to do it... and a lot of the unknown of where the traffic is coming from is now gone.
Given the above challenges... does anyone know of any ad networks that satisfy #1 and #3 above or any other ideas? Any tips or advice?
Thanks a bunch in advance.
PS - I did read the mobile primers on this site and they were well done - thanks!
Obviously the higher the variables involved the more money you need to spend to test properly.
However for mobile, even if there are 50k sites , you won't get traffic from all of them.
Actually from my experience most of the traffic will come from 5-10 placements and some might be profitable right off the bat if you have the right angle.
After you start optimising a bit you should be able to cover bad pubs loses from the good pubs.
So, talking numbers, spending $100 - $200 per campaign for low paying offers ( less than $5 ) is more than enough to get started and know if a campaign will be good or not.
I suggest starting a campaign and moving from there.
Thanks for the response.
I have done a ton of campaigns... and I am seeing stuff similar to this: I just started a campaign today on Airpush... out of just 242 clicks... there are already 60 apps those clicks came from... and I just started the campaign. Not a lot of clicks I know... but the point is... already 60 app ID's from just 242 clicks. Imagine when I get to 10000 clicks?!? This has occurred on other ad networks as well.
What ad networks are only sending you traffic from just 5-10 placements from the getgo? That would be nice.
Also, I noticed you said pubs... as opposed to app's. Shouldn't we focus on the apps as a publisher ID could have a bunch of different types of apps?
Thanks again.
Ok that's a good point, but my question is , how is the traffic distributed across those 60 app ids?
For me the 80-20% rules generally applies, 20% of those sites take 80% of the traffic and that's where I focus on.
If the traffic is distributed evenly in your case the only option would be to spend a bit more.
Yeah, the top 3 app's get much more traffic than the others... I am wondering at what point - ie. # of clicks should I just focus on those ie. black list the rest and just use those? And, do I focus on just those... even if I don't get conversions yet? I guess the logic could be that people from those apps are at least interested in the offer and the theory could go that that once I get enough clicks from those apps - eventually it will convert? What do you think?
I wouldn't blacklist the ones that get lower traffic. Let them run since they don't spend a lot and cut only if you have enough data.
"I guess the logic could be that people from those apps are at least interested in the offer and the theory could go that that once I get enough clicks from those apps - eventually it will convert"
This is not always the case, you might be getting clicks from those because your bid matches with what these sites sell their inventory for.
So what I suggest is to let each site spend 3x-4x ( or you can read Caurmen's optimization guide for more specifics ) the offer payout and decide after that.