Should I "kill" a target once i spend up to the payout amount of the offer?? (22)
Should I "kill" a target once i spend up to the payout amount of the offer??
Well, Ive got sort of my "own" system going whereby I let a high traffic target run only until it reaches the payout amount of the offer until I deactivate the target. So if amazon.com as a target accumulates adspend up to $40 all by itself, and my diet offer pays out $40... I typically deactivate once the target reaches that amount in adspend. ive been kind of fumbling in the dark with all this and have gotten some good results here and there, but IM wondering if its pure luck and perhaps I should follow a system for really optimizing my campaigns.... can you please let me know what your rule of thumb is when deciding a target is a goner? Thanks
I am also very interested in this.I pause target once it reaches the offer payout like you do. Also i use this calculator http://www.peakconversion.com/2012/0...al-calculator/
Depends on your offer payout man, if under $5-10ish I would test 1-2x offer payout, if $30 payout I would test maybe only 1/2-3/4 offer payout, if somewhere in between 1x worked fine for me
Thanks .. Im looking at my cpvlab and it shows CTR percentages. Ive basically ignored the ability of even USING the CTR feature but its high time I stopped ignoring it .... Now Ive always had a bit of a phobia for crunching numbers cuz i sucked at math.. however, I need to break through and make use of the data otherwise Im losing money. My question is, can I get a sense for which target to kill just by looking at the % CTR?? Does CTR determine likelyhood of conversion? If so I could kill off targets alot sooner than waiting for a full $40 per target assuming Im using the 1x payout rule. If this was your campaign.. on a $40 payout... .which of these targets would you kill? Again....Im assuming CTR does tell us what will be more likely to convert, however...if this is not true and even a LOW CTR target converts better than a HIGH CTR target...what formula again would you use to kill off targets by using cpvlab to save the adspend in a more timely manner? I have a big gap in percentage figure for the conversions on this screensot....so for this example I can see that 10% CTR is certainly getting the conversion, but I DO have as low as a 2.63% CTR which brought a conversion as well.. so I may have answered my own question....but in this case is there a "low" CTR % that I know for sure I should kill off like 1% and under or something? http://awesomescreenshot.com/0a6237p3f9 BTW I am basically doing competitive intelligence and using landers that are similar to whats already working, so in that respect Im assuming that the landing page is already a converting lander. Thanks so much!
Update: Ok, I just realized that a 13% CTR is awfully high in comparison to the others.. I feel like I should rethink the idea of "this lander converts" as a generalization towards all targets.. should I be trying to focus on optimizing those really high CTR targets by creating a "special" lander geared towards that target (branding, etc. I do cloak) ....if so whats a good CTR which tells me for sure I should focus on that target?
Unfortunately this is one of the problems with high-payout campaigns: you need to spend a lot more to test them. There's not really a shortcut: you could try cutting on CTR, but until you have a strong sense of how CTR and CVR are likely to be related for this niche (which you'll only get from, yep, a lot of testing) that could well hurt more than it helps.
When cutting out ads / targets you don't want to look at the CTR.
All you need to look at is the EPC per target. You can come up with super high CTRs but if it doesn't convert than it's useless.
As for how much to spend , depends on the offer. If it's a proven converter ( you spied around / AM told you ) I could go spend 3x-4x the payout.
If it's just normal testing , 2x is enough.
Hey guys, ..I have another question for ya. Some targets seem to be getting like 50-100, and 200-500 impressions per day. I have a couple thousand keywords, and with hundreds of them doing 100 impressions per day...my patience gets weakend by the end of the day. Thats cuz the high traffic targets that get 1k-2k impressions per day are easy to spot and optimize quickly enough... So I just wanted some advice on how you normally optimize the turtle targets that are taking their sweet time. ? Basically when a target reaches closer to 2k impressions is when I hope to see a conversion (skin rebill), so if I dont get a conversion within that time I just deactivate the target.. but the hundreds of targets that seem to take their sweet time with NO conversions daily are having a psychological open wound...or "bleeding" affect on me cuz of the sheer volume of targets. And even if a target sits for a week and hits like 700 impressions (or more or less) and gets a conversion... I also think to myself that this may just be a fluke conversion....and how do I know it will happen again and do I want to WAIT that long to find out.. (which could last a few weeks). I know Im overthinking this as usualy due to being a bit new to all this....but again...its just so daunting to wait it out....while I know there are targets that are converting and making positive ROI which could make my end of day ROI much higher without all these blood suckers. I am sure this is normal, and just part of doing so much volume...atleast thats what I have assumed til now, but I keep wondering if I am going about it wrong and should just stick with high volume targets that are more quickly optimizable. But then I look at some of these targets that got 500 impressions over the week and actually got a conversion and it stops me dead in my tracks knowing I should wait it out. Please confirm if this is normal , I do have the adspend to stick it out btw which is a factor Im sure, so Im not really worried about that.
Originally Posted by johnnybegood
Honestly I'd say this is a psychological thing, and I know different people wait different amounts of time for data.
Personally, when I'm running PPV I'll let targets continue to accumulate data until I've got the campaign profitable or close to profitable, and then kill all the "nickel and dime" targets that aren't in positive ROI just because I can't be arsed waiting for them 
Let's face it, if a target is taking more than two weeks to get one conversion, it's never going to make you rich anyway 
Originally Posted by cmdeal
Originally Posted by johnnybegood
Originally Posted by johnnybegood
Let me provide just a simplified analogy:
Suppose you run a casino, and in your game a punter wins $1 for every die roll of 1 or 2, and you win $1 for every die roll of 3, 4, 5, and 6. Now suppose the punter rolls a 1 and then a 2, and then a 2 and then a 1. Under your rule of thumb, you should shut down this game. But the correct answer, of course, is that you should keep the game running, because as the numbers get large enough you will win 66% of the games. So if you run the game long enough for punters to make 1 million bets, your expected profit will be around $667K
So in this simplified analogy, the casino game is an ad campaign. By making a decision to shut down the campaign by relying on a crude rule of thumb instead of sound statistical methodology, you have just "lost" more than half a million dollars.
Statistical significance: for more details (and some math which will be useful in this specific case) have a read of
http://stmforum.com/forum/showthread...he-Winning-Ads (theory)
and
http://stmforum.com/forum/showthread...ats-Calculator (practise)
OMG fellas.. this is pretty wicked...you got it down to a science.. I will have to admit I always sucked when it came to optimizing and always knew that would be the most difficult part about scaling correctly... dont get me wrong I get the "simplified analogy" via the casino example.. what I dont quite grasp is how to really make a decision when the data is not as cut and dry and seems to be off the beaten path (low CTR has conversions as do high CTR) . Caurmen I clicked on that Calculator.. shit man! are you Einstein reincarnated bro?? Very impressive...yikes its kinda over my head.. I would need the version which reads: "enter cost in this field" and "enter CTR" in that field (instructional version of that page) ...seriously.. cuz Id love to work it , but just dont know how. If you use awesomescreenshot.. maybe you use them too and take a screenshot and can jot down which figures get entered within the calculator. I do have an example of a campaign I had deactivated targets for too early..... yet when I applied CMDEAL and Caurmen's advice on "waiting out the campaign" ...i actually saw the conversions trickling in slow but just the same coming in quickly at times... so I get the "statistical significance" concept in theory, but still not sure how to make it more granular in order to apply it.... or simply...how long do I wait out the stats? Here's my example, guys.. if you can just help me with this example, this ran for 3 days...so what would you do here.. I cant rely on CTR cuz even a lower CTR & higher CTR shows conversions... this offer pays out about $40 and each impression is about .015- .016 for popup traffic. So in this example:
http://awesomescreenshot.com/0e12830oef I have notated my thoughts for each keyword even though I havent shown actual keywords, although we can reference each line by using its impression count , ie. line: 7706 is the first one. Before I let this ride another day, with another couple thousand impressions..Im curious when you would deactivate the first line...as well the other lines if at all possible.. Is my analogy incorrect within the notes? Thanks..Any help with details would help immensely!
and here's a better example which I need the most help with that has been running for a week and is THE one I am super stumped with since I have MANY targets with 5k impressions that I dont know what to do with... I feel that I am only deactivating once I the P/L figure shows RED after Ive spent twice the offer payout (40$)......hence the reason why I keep going back to the "offer payout" crude rule of thumb...which I am trying to break out of very badly... let me know what you would do with these..if you can help, just reference the impression count... Im on my way to profit once I get this down guys!! Im holding my self from deactivating since your last post so I feel Im letting alot of adspend gather enough data...but when is "enough" ?? Thanks again! In his example all are still running unless i noted the line as "deactivated" .... http://awesomescreenshot.com/0b6281kt41
Here's another example guys... I have a few of these campaigns i test out daily.. but again, not sure how much adspend (impressions) to burn through before making that "calculated decision" and deactivate a target... Im bleeding red with so many targets daily which is fine as long as I know what the sweetspot is to kill em'... http://awesomescreenshot.com/081282rdb5 same $40 payout and this has ran for 3 days
another one, by now yu see what the different types results are....so I was going to just kill this whole campaign cuz the CTR on most all of these were low and got frustrated looking at all the red....
http://awesomescreenshot.com/0ec282sw88 but i see 1 conversion with 4.40% CTR so obviously this target is worth keeping even if the volume is not there .. so at this point, I feel like i should gauge all other optimizing based on this CTR.. thats my confusion.. sorry for the long winded posts...but at the end of the day, I realized just now that what I am doing is seeing what the "average CTR" is of the converting targets...and using this as ideal CTR for all others to deactivate really low CTR targets... but again, i get thrown off by even the LOW CTR targets that actually get a conversion and then Im stumped and just leave ALL targets ride until im deep in red with 3k-5k impressions.. so again, am i waiting too long? Instead of waiting for thousands of impressions , in this example can I kill the really low CTR
1,202 22.55% $36.06 $0.030 7 0.58% $5.15 
bump
Originally Posted by johnnybegood
ANyone?
buuuuuuuuump
Break you campaign into smaller campaigns, categorizing the targets into say 10 smaller campaigns. Also if you have any sure winners then get them in their own campaign so your not toggling them. Keep the small volume targets in one campaign, let me sit for a week so they have enough data for you to make a call on. During the time you can focus on high traffic targets.
Home > Paid Traffic Sources > POP / PPV / Redirect