testing quickly, reasoning? (13)
hi,
in the 40 day marketing guide it says this: "Making the decision to stop a campaign is important but depends on which offer you’re running. Particularly with email and ZIP submit offers, it’s better to test quickly rather than run up a net loss of $50. Offers move on quickly and wasting time on unprofitable ones wastes the opportunity to succeed in another."
might be a VERY STUPID question... but i guess is why i put it in the newbie q and a.
- why test quickly?
- why do offers move on quickly? is it because so many affiliates are pushing it and people get accustom to that particular offer?
- why do campaigns on average have short longevity? is it because other affiliates/ offer owners/ networks are spying and doing the same thing, which drys it out?
Thank you very much for answering!
1. Fail quickly -> find success faster.
2. Offers go down due to advertisers running out of budget, finding it's not profiting enough, dries up due to lots of people encountering it, death by legal reasons, spin/angle becomes irrelevant (e.g. sporting event over), etc.
3. You mean your affiliate campaigns? Depends on what traffic source, offer, angle etc your doing. Often if on FB your CTRs drop away or you saturate your demo, push lots of traffic very quickly then find it doesn't sustain, etc. Broad demo campaigns and campaigns you replicate for other demos can last for months and months, it all depends really.
You get a feel for when to kill things off once you learn a traffic source and niche. I'm getting much better and culling off my FB campaigns, images and so forth very quickly to minimise losses. Sure you can miss winners when you don't put in enough time to optimise etc but you get a feel for what does and doesn't work and can spot weak campaigns pretty early on.
I'll add another point to that. You only need to spend $5-7 on a zip submit to know whether or not it converts. So, "test quickly" in this context means keeping a close eye on your campaign data and ruthlessly killing off those that don't work. You won't learn any more spending $50 than you would by spending $10. Instead, you just lost $40 you could have used to test 5-7 other campaigns.
^^^^this....esp with email/zips.....they either convert or they don't.......if you have spend $10 on a email/zip sub and have 0 conversions just move onto the next...
the thing with email/zip submits is they scrub like crazy, so the data you do get is rarely consistent. On Day 1 you have a $0.35 EPC. Then on day 2, you have a $0.05 EPC. With most other offers, your data will be much more stable because they don't shave and scrub to the extent that email/zips do. Not all email/zip submits are this way, but most are.
Hey guys,
Sorry to highjack the thread, but this is to do with testing so hopefully Mr Potter you'll find this useful too lol.
I've got a game offer I literally just started running on FB atm. It's payout is $1.15/lead and I spent $10 on it today and got 3 conversions (although that was over 2 networks so ave payout was $1.06 - the second network bumped the payout mid-testing lol).
My question is, am I on the right lines... I've read that you should spend between 3-10x the payout on "testing" a campaign, but I've never been too sure what that meant. Reading above, I think it means just seeing whether or not you'll pick up some conversions(?).
So now, in my case, now that I know this thing converts, I'm going to start testing a ton of images and try to drive down cpc. How much should I spend per day on this phase (optimizing)?
Cheers in advance
It depends on what the CPCs are and how many images/ads you have. With a gaming offer don't be afraid to run at a loss for a while to find your best images. If I was running at $0.10-0.20 clicks i'd put in $25 to $50 a day so that I could pull in at least 100 clicks across the ads in a day. Any high CTR ads are likely to stick and drop in CPC.
What ballpark CPCs/bids are you looking at?
Cheers for getting back Zeno,
Fb was a little messed up with the suggested bids (said $.08-.11) but I didn't get any significant traffic till I ramped it up to $0.42.
I've got 23 ads running atm - I'll attach a screenshot of the campaign so far otherwise I'll do a shit job explaining!
I've paused two of those ads already (all are shown paused atm though) as they had ctrs of <.08% after over 5000 impressions.
I've noticed this as well lately - suggested bid of say $0.20 but no traffic (and advert not delivering error) until getting a lot higher. Oh well, you bid what you need to bid in the end anyway.
Ok, so I've been doing a lot of gaming campaigns recently (actually broke my first $1k profit week, still have Sunday too!) and I have a specific regime I follow:
1) let em all fly
2) when ads stick and run away with impressions throttle down the bids to let the others have a chance
3) if an ad has >2 clicks, over 1000 impressions and the CTR is less than half of what I want I pause it without remorse
4) repeat, let high CTR ads trickle while others get delivered, all the high CTR ads together should take no more than 1/2 of the impressions of the campaign (no point having other images to test if they don't get seen!)
Point 3 may sound rough or contrary to what other people say but in my experience you need the ads to stick to perform well, and you can see very quickly if they are not sticking. If an ad gets to that state and has <1/2 the desired CTR then it's pretty fucked, I have never seen any suddenly jump up and stick.
Now what you need are clicks to figure out your approximate EPC. That will dictate what kind of CTRs you need to obtain to make this profitable. My gaming campaigns at the moment are running at ~$0.20 EPCs and I have to get CTRs above about 0.1% to break even. In saying that though I have dozens of ads running with 0.2-0.5% CTR so it becomes more of a game of fine tweaking and hoping the campaign won't die too soon!
Zeno, you're a ledge!
Thanks for all of that, I'm gonna follow that to the letter - you're obviously not doing too badly for yourself lol.
Couple of questions though:
1) How much do you throttle the bids down by and when do you judge it appropriate to do that?
2) How much do you reckon I should spend per day based on the image I attached last time?
Cheers (again) in advance man
overall the reason for test quickly is to ensure you dont build up losses that are too high. the key is to find the offer that converts and hasnt got too much scrubbing. then to optimize and blow it up as quickly as possible.
1) test quickly because as a new affiliate running 10 campaigns of which half end with net $50 loss could be the end of your willpower or desire to keep going at this. And as some have said above, on email submits after spending $7-$10 you have enough data most of the time to make an informed decision.
2) yes and for a whole host of other reasons many of which are to do with the advertiser and their funnel.
3) targets get too competitive, offers disappear, networks/advertisers scrub, etc etc.
also
overall
recovering from a $50 loss is a lot harder on a campaign then starting one with a $50 profit.
Originally Posted by el_jefe
Imagine a graph of average CPC versus time. For a high CTR ad it will start high, then rocket down a steep slope and then level off somewhere. You want to change your bid at least an hour after it has levelled off, so not so much dependent on number of clicks. If your ad has had 10 clicks over 3-4 hours and levelled off at a rock bottom CPC then so be it. If it has taken 100 clicks to do that, same deal. You want the CPC to lower naturally to that minima first. If the ad loses all impressions then just raise the bid a bit, your letting the other ads be tested though so it's sort of a "wait there, i'll get back to ya" system. Once you filter out poor performers and cull off ads the good ones that you've lowered bids on will start delivering again.
With those CPCs and that many ads I would budget $50 a day. You will do fine with $25 but with $50 you will get more impressions across all the ads. Limiting your budget can be counterproductive to testing as you won't get the impressions smeared across all the ads as readily. Just watch it like a hawk, check back often and mercilessly kill off the weak. Remember to take into account the EPCs on the network side though! If all the ads fail then start again. New images or angle. Don't get attached to the campaign if it's consistently underperforming.
There's also a psychological factor with testing quickly. The longer you take putting up a campaign, the longer you want to wait it out to see if it works. On the other hand, if you throw a bunch of campaigns against the wall, it's a lot easier to kill those that don't stick. Of course, there is a balance. My experience though, is that money favors action.
Home > Questions and Answers > General Questions