12 {{ upvoteCount | shortNum }}
12 {{ upvoteCount | shortNum }}

Question about mobile ppc testing

by KINDLUTION
10 replies
I've read that the general rule of thumb is to not spend over 2-2.5x the payout on the offer. On mobile some of these only pay $1.50-2.00/install. So, if I go by that then I'm pausing the campaign at 100-150 clicks. Is that how it goes?

My theory is that since half of mobile clicks are accidental, then we would need to spend 4-5x the payout in order to gather enough data. I'm only testing my 2nd campaign. So, I am no expert by any means.
#mobile #ppc #question
  • Profile picture of the author flinserl
    Originally Posted by KINDLUTION View Post

    I've read that the general rule of thumb is to not spend over 2-2.5x the payout on the offer. On mobile some of these only pay $1.50-2.00/install. So, if I go by that then I'm pausing the campaign at 100-150 clicks. Is that how it goes?

    My theory is that since half of mobile clicks are accidental, then we would need to spend 4-5x the payout in order to gather enough data. I'm only testing my 2nd campaign. So, I am no expert by any means.
    Based on your example I see you're currently spending about $0.10/click with your network. I'm also assuming you are new at mobile CPA. No sure how experienced you are in CPA/CPC campaigns overall, or if this is your first attempt in Internet Marketing.

    From experience you'd need at least 250-300 clicks to get good stats before even starting to do effective split-testing! So: Yes, you'd need to spend at least 4-5x payout.

    That's why I like networks with $0.01/click offers for someone new to mobile CPA. That way you get 500 clicks for $5 vs. $50, which equals the ~2x payout you mentioned earlier.

    Of course: Traffic quality isn't always that good on those cheap networks, but you learn A LOT from your - cheaper - mistakes

    Once you've gotten the hang of testing, trying text/banner ads, choosing different carriers/OS's you can move on to higher quality, "premium" networks and spend $0.10/click then.

    Hope that all makes sense?
    {{ DiscussionBoard.errors[8436454].message }}
  • Profile picture of the author KINDLUTION
    No, that makes sense. I'm actually paying ~.04/click. What networks have .01/click?

    Also, if I get the 300 clicks with 0 conversions, should I just skip the offer? I had a decent ctr on the offer I tested today(.94), but after 322 clicks I had 0 conversions so I stopped it.
    {{ DiscussionBoard.errors[8436470].message }}
    • Profile picture of the author Lewis T
      Originally Posted by KINDLUTION View Post

      No, that makes sense. I'm actually paying ~.04/click. What networks have .01/click?

      Also, if I get the 300 clicks with 0 conversions, should I just skip the offer? I had a decent ctr on the offer I tested today(.94), but after 322 clicks I had 0 conversions so I stopped it.
      It also generally depends on the country you're advertising in. If you get 300 clicks and no conversions, you're either targeting a very bad channel, LQ traffic and a different network may produce different results. It's not the same everywhere.
      {{ DiscussionBoard.errors[8436528].message }}
    • Profile picture of the author flinserl
      Originally Posted by KINDLUTION View Post

      What networks have .01/click?
      BuzzCity would be one of them. This isn't an endorsement, though!
      {{ DiscussionBoard.errors[8443500].message }}
  • Profile picture of the author Greedy
    I've read that the general rule of thumb is to not spend over 2-2.5x the payout on the offer. On mobile some of these only pay $1.50-2.00/install. So, if I go by that then I'm pausing the campaign at 100-150 clicks. Is that how it goes?
    Here is how these theories work.

    Pretend your offer pays out $5, to spend 2-2.5x the payout testing it, you would have to spend $10-$12.50 in ad traffic.
    Signature
    {{ DiscussionBoard.errors[8436707].message }}
  • Profile picture of the author KINDLUTION
    Yeah, I guess I need to choose higher paying offers.
    {{ DiscussionBoard.errors[8436736].message }}
    • Profile picture of the author flinserl
      Originally Posted by KINDLUTION View Post

      Yeah, I guess I need to choose higher paying offers.
      Yeap, it's all a numbers game
      {{ DiscussionBoard.errors[8443495].message }}
  • Profile picture of the author surreal
    Unlikely youll make much of the 1 cent countries. This either says there is very low volume, or no competition indicating low quality traffic.

    Youll need to budget at least double offer payout per source/creative per day for a week. Thats my opinion. You will find small chunks of what works and what does not. If you stop too early, youre throwing away the data youve worked to collect towards being profitable. Spending less may mean losing more!
    Signature
    {{ DiscussionBoard.errors[8443922].message }}
  • Profile picture of the author KINDLUTION
    Wait so that double the offer amount is per day? Not just overall?
    {{ DiscussionBoard.errors[8444704].message }}
    • Profile picture of the author flinserl
      Originally Posted by KINDLUTION View Post

      Wait so that double the offer amount is per day? Not just overall?
      I recommend not to focus on the $/day, but rather on the amount of clicks you need to figure out the profitability of a campaign. The $/day "formula" has a major flaw: it has nothing to do with statistics!!!

      Who says that an $20 payout "demands" you spending $40-60 to test a campaign? What if you "only" spent $30 on 300 clicks and had no conversion? Would it make you feel "better" if you spent $300 for the same 300 clicks to get no conversions?

      It's not the $ amount spent that tells you if a campaign is successful. It's the amounts of clicks you got, and the conversions/lack thereof to validate a campaign.

      If over 300 people cared enough to click on the banner, but the refused to go through with what's been offered: THAT will tell you something about the campaign, not how much money you spent to get the clicks.

      300 clicks/no conversions: that's statistics.

      $200 spent/no conversions: thats a useless, urban-myth based measure that "measures" nothing other than the willingness of one to gamble with his/her money until admitting it's LOST.
      {{ DiscussionBoard.errors[8493010].message }}

Trending Topics