How do you tell if a campaign has 'potential'? (7)
Hey guys,
I've been testing a few new mobile campaigns, with some clear winners and clear losers, but I have a question on that large grey area between winners and losers. I want to know when you have a 'meh' campaign, how far down the rabbit hole you should go? Here is a very typical example of what I've seen:
One test campaign is as follows:
Method: Direct Link, Android only
Payout: $2.25
Clicks: 291
Conversions: 3 (across different carriers/models)
Earnings: $6.65
Cost: $20
I'm really stoked that I'm getting conversions, but with only 3 conversions I don't feel like I have enough actionable data to start optimizing carriers/handsets/etc in any meaningful way. At this point if I can't make any optimizations according to these metrics, I see 3 options available to me.
1. Assuming the same conversion rate, I could just leave it alone and let it run all week with another $100 so I may get a larger sample size to optimize. The only question I have here is how much in the red should I be before calling it a day, and moving on to options 2 and 3?
2. Split test in a Landing Page - this is what I will probably end up doing, to try to align my creative with the offer.
3. Bail on the campaign, as there isn't enough potential.
As I see it, those are my 3 options after an initial test. So which option would you guys take at this point? If there is an art form in this business, it's determining which campaigns even have 'potential' and I feel like that's a skill I'm lacking. Thanks! 
~Ian
With mobile you really need to test much deeper than that.
291 clicks really don't mean much.
I would focus on the winners and leave the meh campaigns. But that's only because I have a lot of campaigns running and simply do not have the time to spend on these meh campaigns.
The time lost on trying to get those to work could also be spent on finding more winners and pushing your existing winners harder.
Time is money in multiple ways when you run mobile. Lots of offers have a short life span and they could be pulled any time. This goes for the winners and the meh.
Examples of what I experienced:
You could be mehing around while your winner gets pulled. You could have gained more profit pushing your winner harder.
You could be mehing around while the actual meh offer gets pulled. You lost your meh investment plus losing out on extra profit on the winners.
After mehing around you finally start seeing a profit. Great!! Only need another week at this rate to make your $500 investment back.. in the meanwhile offer gets pulled.
instead of spending $100 in a week, why not to spend $100 in a day and optimize campaign quickly?
How do you tell if a campaign has potential? Through experience.
I have held onto losing campaigns for a longtime because I've made similar campaigns work. I've dropped campaigns after a day of testing because I just wanted to throw out a random idea I've never tested before.
There are so many factors it's hard to give you specific advice.
However...if I were you I would of started my campaign with some big ad split tests. 5 ads using different angles. Spend around $100. Then analyze the data campaign wide, as well as by ad. If I'm losing by -30% or better I would keep plugging. If not, I would probably let it go...unless you know other people are making it work.
Thank you for the awesome advice everyone I am going to pop some money into this, get a few more angles up and give it a go!
Basically, what Mister Green said.
I try to at least break even and track everything from the start so I can step back and say, okay, I'm not making money with this angle but I can reduce my losses and take in some profits by getting rid of bad traffic sources...what are the bad traffic sources?
Then just go through and bust up whatever traffic sources are clogging your servers while testing other angles/landing pages/headlines/images until you are at around 50-100 percent profit margins.
If a campaign doesn't convert at ALL or hardly converts after spending around 100 bucks or so...let it go, especially if it's CPA. If it's CPS, maybe hang on and try 500-1000 dollars worth of testing to see, CPA you will know pretty quickly though.
Home > Paid Traffic Sources > Mobile