“mobile”/ adcombo
Dismiss Notice
Welcome to Our Community
Wanting to join the rest of our members? Feel free to sign up today.

how long should I test a campaign?

Discussion in 'General Affiliate Marketing Forum' started by watermelonman, Oct 18, 2013.

  1. watermelonman

    watermelonman Affiliate affiliate

    hey Charles,
    I am asking this because of several reasons:
    1) most of us have limited budgets and it's impossible to test everything infinitely. How long should I test it before calling it quits or scaling it up?

    2) (this is related to my previous question regarding timing of launching a new campaign) one day's data won't be statistically significant so how long should we keep testing? 1 week? 1 month? 24 hours a day or just certain hours in a day?

    3) Assuming your answer to the above question was "1 week". At the end of 1 week, I'll have two options---either kill the campaign or scale it up. No problem with killing the campaign---I will just launch some more. But if it's a winner and I'm ready to scale it up, other ppl using spy tools can see that I've been running this for a whole week and it's very possible that it's a winner. What if they start copying right away? Wouldn't that have a huge negative impact on my to-be-scaled-up campaign?

    Goodness me, I hope I'm not over thinking and over analyzing again here. But if that's the case, please shed some light!

    Thanks gazillions!!
  2. terraleads
  3. solare1

    solare1 Affiliate affiliate

    Make sure you test each offer for at least 100 clicks before jumping to conclusions.
    I read it from somewhere can't remember it.
    That is with split testing the offer by the way.
    Never say anything about direct linking..

    Maybe I'm wrong..correct me :)
  4. watermelonman

    watermelonman Affiliate affiliate

    thanks I've heard that as well but the problem is that depending on the "timing" of your campaign launching, 100 clicks can be done in a few minutes. Which raises the ultimate question: is it even statistically significant at all???
  5. charles_ngo

    charles_ngo Affiliate affiliate

    Persist or Pivot: Deciding What to do With a Failing Campaign - The Blog of Internet Marketer Charles Ngo - Affiliate Marketing

    Testing an offer is NOT a measure of time. For example one guy could test an offer for a week and only get 50 clicks, while someone else tests an offer for 2 hours and gets 2,000 clicks. See what I mean?

    Don't worry about spy tools, pretend they don't exist. Someone can take your ad/landing page and still not profit. There's so much behind the scenes stuff such as targeting, bids, optimization, offer payouts, that they can't copy.
  6. watermelonman

    watermelonman Affiliate affiliate

    Got it Charles, thanks!

    So what's the rule of thumb? 100 clicks? 500 clicks? what if 100 clicks means nothing and you think it's a loser but after 300 clicks it becomes a winner?

    or maybe it has something to do with the payout?

    is it a science or a game of randomness? Sometimes uncertainties like this just paralyzes me(maybe some other ppl as well)

Top Resources