FB Follow Along (29)
Hi guys,
I have never done one of these before so please bare with me.
Here is what i have done so far:
01) Read the getting started guide
02) Read some other links I thought were interesting
03) Selected a traffic source (facebook). I know that in the guide they suggest against this but I feel this has great potential due to the demo targeting.
04) Joined some networks
05) Found an offer (game, direct link to offer). For my first attempt at this and to remove complexity I chose an offer that will not require a landing page.
06) Installed CPV Labs
07) I just created an ad with 10 different images and uploaded it to fb. All 10 have been approved.
The demos I targeted are males 18 - 30 who live in the US and whose interests are mmmopr such as warcraft.
I chose cpm and they suggested a bid of 12 cents. I put the bid to 24 cents and put the campaign budget to 15 dollars a day.
If anyone has questions or any suggestions please let me know.
Since I am using fb instead of pof or adult do the numbers about when I should cut and prune remain the same?
Thanks and let's see what happens.
So after about 3 hours I paused the campagin
It is not so much the budget got spent (it didn't it only spent 4.90 out of 15). It was the fact that all the images had seen over 10,000 impressions and only 2 got click
Impressions Clicks Click-Through Rate Social CTR CPC CPM Spent Reach Frequency Actions
195543 2 0.00% 2.47 0.03 4.94 52440 3.7 2
Not sure how to make the paste in cleaner.
What i have decided to do is do the following:
01) Use racier images (fantasy gamers with big boobs)
02) use a more provactive headline
03) try using cpc budgeting to see what ads are winners before switching to cpm.
Any questions, comments, or suggestions welcome.
Thanks again.
I would like to try to post everyday so I have a complete journal of my experiences.
After running that campaign the AM put me in a skype chat with someone who is more knowledgeable. He gave me a lot of information:
I shouldn't be using google shortened links because
a) they are another redirect
b) it can mess up the stats in cpvlabs (which I see you can reset)
c) people might not because they dont trust the shortened links.
He told me I should buy a domain similar to what I am working on (gaming).
I then bought a new domain and this led to having to have cpvlabs moved to a new dns name (same server new dns).
CPVlabs support has not been very good technically for the most part & I come from a hacking background so I am pretty technically fluent with things like servers, networking & security.
I was having problems with the migration and was going to wipe out cpvlabs and reinstall since there is no data in there but a close friend wisely told me I might have to get another domain when I leave the gaming vertical and I should troubleshoot this now. Good advice. This led to me finding out, after a misconfiguration cut off my ssh access that my vpn had issues. I then fixed the vpn, fixed the cpvlabs and created a process document.
In the meantime I need to have all my creatives approved by the advertiser since i am not using the defaults. I came up with a new 20 images and got stuck waiting and waiting and waiting. I am still currently waiting! I leave for Seattle for 4 days now & it is frustrating since I wont be able to move forward and continue taking actions.
Key takeaways from the last two days:
01) Expect technical growing pains.
02) Create process documents so everything can be handed off and or automated
03) Work on more then 1 offer at once so if one becomes a bottleneck and you are waiting you can work on the other.
Comments, questions or suggestions welcome.
It's old, and also worth a read from chapter one.. but from the bottom of this page is a great read for how Charles got into gaming niche on FB. It's old, and the game has changed, but some of the tactics will still be relevant..
http://www.charlesngo.com/the-rise-of-ngo-ch-4/
Originally Posted by dynamicsoul
And thanks for the suggestion I really appreciate it. I will make this work and be successful just hard to find good mentors.
If you find out how to get your ads running on FB, let me know. I keep getting disapproved. I've tried redirects and iframes to no avail. Do you need landers?
Originally Posted by jdenhaan
Yesterday was a travel day for me and I got info back from the AM about the pictures i submitted. 4 out of the 20 were rejected by the advertiser. I sent the pictures with default file name they had when downloaded so he tells me everything is ok except 4,9,12,13. Huh? How do I know which is which? I also found another offer and started to build another relationship with another AM.
Takeaways:
01) Name your files using batch renaming programs
02) Don't send in 20 pictures. Send in 200. You have no idea how many the advertiser is going to shoot down, and then you also have no idea if facebook is going to approve them. You need to be taking constant action and if there is going to be a lag of a full day for approval send them a ton at once.
03) If you are new work on multiple offers that way you are not wasting time when time when you are on hold. One offer can become a bottleneck.
04) AM's don't give a fuck about you because you are a no body in the beginning. This reminds me of dealing with stockbrokers. It is also a rookie move by the AM because today's rookies are tomorrows superstars. They are thinking battle vs war, which, imo, is short-sighted.
05) Try to get guidance and join as many networks as you can because if you want something specialized a lot might not have it (eg I want only games that can go on fb direct to lander. No installs, no payment information). You would be surprised how few people have this. Even worse the current offer is on another network at a higher payout but the AM from that network won't even add me on skype.
Today I will start working on more pictures for 1st offer and start researching demos for the second offer. I asked yesterday in this thread where people go to find inspiration for pictures and research demos. That thread went unanswered: http://stmforum.com/forum/showthread...ion-for-images
Here is a screenshot of a perry marshall program I was watching that has some good suggestons
As always questions, comments and suggestions appreciated.
Just wanted to post my latest updates. So I am on the road in Seattle but still trying to take action. What i have done is continued to try to nurture relationships with new AMs. One thing I notice are most AMs are shady as fuck so you really have to manage the relationship. It's annoying but I guess the cost of doing business. What's funny is once I hit 4 figures a day I am going to put up a thread breaking down the networks and who were helpful, why and who wasnt. Unfortunately I am not in the position to do that right now.
So the AM from the company that was helping me the most has become unresponsive. No longer responding on skype, email, etc. That is sad because he has been helpful however the offer is on another network for .05 more. I am not super worried about the payouts. What's most important for me right now is I am going through the motions and getting reference experience. I have found 2 other game offers so I will move forward with those. The annoying thing is you need the AMs to approve creatives so you are dependent on them and a chain is only as strong as the weakest links.
Takeaways from today:
01) Keep nurturing relationships (dont put all your eggs in one basket and all that ish)
02) Keep getting other offers (so you are not dependent on any one. Again eggs in one basket).
03) Accept that most AMs have the mentality (as it appears to me, maybe I am unlucky who I have gotten) of a mortgage broker or a greasy used cars salesman.
04) Keep moving forward. Now when I get offers I let the AM now I will be sending them 100 - 200 pictures at a time. How long is the turnaround time to approve that? The more pictures i have approved the more ads and testing I can run, the less I have to deal with them.
That's it. I am now stuck. I am waiting on 2 AM's to get back to me (one was the one who was super helpful who is not responding to skype or email) & the other is trying to get me an offer approved for social (she's actually been pretty helpful so far). The 3rd is super responsive and I need to research demos to come up with pictures for her. Probably no more action will get taken until Monday.
Questions, comments or suggestions welcome.
Thanks for reading.
Hi mate
Just wanted to thank you for your follow along.
I'll keep an eye on this thread!
Good luck!
Originally Posted by vitalis
Today's task: Gather 300 images. 100 for each potential game.
There has to be a way to automate this or buy off fiver. Downloading them is turning into a time sink. Maybe this gets outsourced to a VA.
Hi,
So I lost 5 days because i had to travel to Seattle on business and then I needed a day to recover.
What I did in the time I could not be active with setting up campaigns was:
01) Apply for 2 more networks
02) Continue going through fb products I got from thevault.bz, and taking notes.
03) Continued to shop for offers and I found a second & possibly a 3rd.
Today what i have done is as follows:
01) Gather another 80 images for approval for the first offer.
02) Gather 40 images for approval for the second offer
03) Tried out 3 Windows batch photo resizing apps. I have selected the one I like the most
04) Talked to 2 AM's from different networks.
05) Launched another campaign
The details of this campaign are (same one as before but just to have everything in one place):
01) Game offer direct to lander
02) I will try cpc (as opposed to cpm which is what i did last time)
03) A budget of 15 dollars
04) Run the ad later at night since it is a video game and people will be done with high school & college
05) The demo is 4 million they want 1.12 a click. I bid .50 I have no idea if this is right or wrong.
My quesitons:
01) Is 4 million too much?
02) Is I lower target audience will this bring down the price?
03) Is .50 too cheap? Is it possible they will not even show my ads?
Quesitons, comments, and suggestions are always more then welcome.
Thanks!
Hi,
So I left the campaign running overnight.
Here are the stats.
Results?
29
Website Clicks
Cost?
$0.43
Website Clicks
Campaign Reach?
51,907
Frequency?
1.6
Clicks?
31
Click-Through Rate?
0.037%
Total Spent?
$12.50
While this looks like a failure I see it as a victory for the following reasons.
01) 1st campaign only got 2 clicks for 196k impressions (that was cpm). This one got 29 clicks.
02) Even though it had zero conversions the actual CPC was cheaper (though I ended up spending more).
03) I got data that shows me what ads got clicked on the most and what ads didnt get clicked on at all. This gives me information on what was clearly wrong and what is somewhat right.
04) It allowed me to play with CPC for the first time.
Action steps:
01) Pause all ads that had zero clicks.
02) Look over best & worst performing ads to try to spot patterns.
03) Let campaign run during the day today to gather more data.
04) Try to come up with tighter demographics.
05) Possibly change the angle.
Comments, questions & Suggestions are always appreciated.
Thanks!
Looks good!
You'll probably need to find a way to get more data reasonably soon - maybe slightly higher CPC bidding.
However, you're absolutely right: this is a victory because you're getting data in! Get more, keep optimising, and money will arrive.
Originally Posted by caurmen
I will try to move the budget to 15 or 20 and someone else suggested the same as you: high cpc bidding. As it stands I saw that a bunch of ads were not performing so I shut them off. It is currently running with 4 ads out of the 11.
How does one know when to shut an add off? For me i went with zero clicks or a ctr of under .05x after 5k - 10k impressions.
For the ones that are above do I keep the images and try something else? I am not sure what my next optimization step is besides tigthen the demo (it was 4 million which I will shrink).
Today was a rough day. I continued to let the ads run over night and slowly but surely I started to pause them 1 by 1 today as it appeared they were going nowhere. Still no conversions. I ran into some personal problems that took away time from AM today and while that was happening the offered was pulled. In the meantime I have not heard back from the other AM about the creatives from the other offer. I then talked to a super affiliate who was introduced to me by a friend. He told me the following
01) Use a lander even for games
02) Get a spy tool
03) I should be doing diet / skin and ripping peoples landers and improving them
04) I should be targetting internationally
05) I shouldnt be doign facebook
A lot of this was discouraging but I probably agree with point 2. I think a spy would help me see what is going on out there and help me broaden my ideas.
Takeaways:
01) Try to work on 3 offers at once as at anytime it can get pulled and you are stuck in a holding process either looking for new offers or waiting on
02) Try to use your free time wisely (For example I see that you have to constantly delete the same fields in the power editor so I wrote a quick macro in excel to clean that up for me.
03) Look into spytools
Now i have to find a new campaign. The final stats from today and yesterday were:
Results?
43
Website Clicks
Cost?
$0.42
Website Clicks
Campaign Reach?
61,812
Frequency?
1.7
Clicks?
45
Click-Through Rate?
0.043%
Total Spent?
$18.14
Comments, questions and suggestions are more than welcome
I'd strongly agree with 1, 2 (although remember to innovate on what you see, don't just copy) and 4. I don't run diet / skin on FB myself, but my understanding is that they're considerably more likely to result in account bans than games, so I'd be a bit dubious about 3. 5 I wouldn't agree with - you can still make money on FB, and concentrating on a clean vertical minimises your chances of account bans. FB might not be the perfect traffic source to start with, but I don't think you should jump ship now you've started.
I'd also definitely agree about testing multiple offers. Only having one offer running is asking for trouble, and you'll be amazed at how differently different offers perform.
How did this end up going for you? Gaming on FB can definitely work and it is a great place to start learning the ropes. For future reference I would recommend sharing some of your creatives if you're going in blind to a new vertical/traffic source and so on. They are first and foremost the most important thing to get 'right'.
I intend to craft a comprehensive gaming-on-FB tutorial slash hybrid follow-along in the near future so that may be useful to you, even if only retrospective! :-)
Originally Posted by zeno
I actually stopped updating the thread because I didn't feel like I was getting a lot of help (not that I am entitled to shit) so it seemed pointless since I already journal on my own.
I have been launching new campaigns and optimizing my processes daily. I was even profitable for a bit 2 weeks ago (though I ended up down net 35).
I guess my questions that I am stuck on are this:
01) How many impressions do i need to have a sample size that hits statistical relevance. I talked to a friend who wrote algos at ebay and does machine learning in haskell and he told me without the variance I would need 15k impressions for 95% confidence and 26k impressions for 99% confidence. Is this correct? I still don't get when I should be pulling the plug.
02) Targeting. Should this be super broad or not? I have talked to some super affiliates who are on this board and they told me just do gender broken down by age noting more. This seems super counter intuitive mathematically since that will expose the ad to a fair amount of people who don't give a fuck which will destroy the CTR and blast up the CPC. I have found I have done the best with ultra targeting and scraping uids. The counter argument from the super affiliates is then I have no ability to scale.
03) I still don't understand how to make data driven decisions. The strategy I am working off of at the tail end of this offer is to have 1 ad per campaign group, This ensures the ad gets a fair amount of impressions. I saw last night the idea of letting 3 ads cloned all run in 1 one campaign so it you get triple the exposure if the initial exposure of 1 is not enough. Why 3 why not 10? So I have 1 ad per campaign and each I have managed to get the demos down to 18-19 and 22-23 as the performers. Each ad is the same picture, headline and body just for the different demos.
Here is yesterdays results (attached). Based on them I shut off: Style VIII 18-19-11 as it has 17k impressions and is unlikely to get to .1.
Thanks for taking interest in my thread. I am taking daily action I am just kind of stuck on the next steps (how to figure out proper targeting strategies, broad vs wide & how to interpret the data). I tried to upload a csv but it got rejected so I took a screenshot instead.
Any help with my challenges would be super appreciated. Thanks in advance.
Also when people say CTR do they mean CTR or uCTR? I would think we would want uCTR since people clicking the ad once but being exposed to it again is unlikely to result in them clicking a second time.
Thanks again.
Here is the data sorted by uCTR. That might make it cleaner.
Right, jumping in to start getting you straight on where to go next!
Statistical relevance: Use the calculators listed in parts 10 and 11 of the Getting Started Guide - they'll give you the data you need. Let me know if you need any help untangling them.
Whether your friend's recommendations are accurate very much depend on what you're trying to determine! Was that a CTR or ROI calculation?
Targeting: both broad and narrow can work, but narrow is by and large easier to get to a decent ROI. Broad has a far greater maximum scale, but is harder to get working initially.
Originally Posted by caurmen
I sent my friend your article from the getting started guide: http://stmforum.com/forum/showthread...he-Winning-Ads
I will re-read the other 2.
Do you know when people use the term CTR if they are referring to actual CTR or uCTR?
He was talking about CTR and the odds of something being .1
I'll answer your questions and then give you some further guidance:
With gaming offers you can expect an EPC of anywhere from about $0.05 to $0.50 when direct linking. If an offer is performing amazingly, has a high EPC, then you would be a bit more forgiving with your ad-culling strategy. If the offer is generally giving say $0.20 EPC, then I cull quite aggressively based on what CPC the ads are going to get. Say an ad launches and pulls 0.156% CTR in a campaign at $10 budget. My bid is $0.67. The ad's CPC starts at 0.41 and I see it slowly drops to about 0.31. I got several conversions and the ad ran at about $0.23 EPC. I let it run the next day (but I don't touch budget yet!). It did similarly and had 0.141% CTR, CPC stayed averaged at $0.33, EPC dropped to about $0.18. At this point I would look at my campaign wide EPCs and see that $0.35+ is not happening, so this particular ad is very very unlikely to ever get profitable - I would need the EPC to jump up 50% or CTR increase dramatically -> CPC down. I cull it and focus on improving my ads.
Always break into gender and age groups. Always. The 13-17 age group is cheap but they are the lowest lead quality for advertisers > will get you booted from the offer. Go hard with scraping UIDs - this kind of ultra-niche-targeting gives you a leg up on the competition. Worry about scaling after you've cracked your first successful campaign - no point being able to scale to a 5 million strong demo when you haven't yet got a single ad profitable.
Also, when people say CTR they mean true CTR, not unique. I consider unique CTR as a measure of 'audience engagement', i.e. how much users like your ad - as it represents the % of the audience that have clicked. However, real CTR (clicks vs number of times ad delivered) is the performance measure which also influences CPC and the FB delivery system.
Now, onto some general tips.
1. Make sure you have a well-designed tracking setup that protects your ass from avoidable disapprovals. When first submitting ad I recommend having the offer URL in CPVlab or wherever as the offer page directly. No affiliate link. This way there is no chance of the FB reviewer being bounced by your affiliate link to an unrelated offer that is non-compliant. This is the most easily avoided yet most common way new affiliates get in trouble. Once ads are live, switch to the affiliate link and if possible add a rule to the campaign that says, for example, "if clicker not from US then send to offer page else send to aff link".
2. Try oCPM with default bids. You would be surprised how well these may do compared to traditional CPC ads. Use Facebook's conversion pixel if possible so you can see cost-per-lead values in the reporting system.
3. Put more effort into your gaming ad images. Gamers are easily one of the more visually inclined demographics. I use photoshop. Make a new 100x72 image, drag and drop images in, resize to look decent (go for size = clarity over fit everything into that tiny space). Use unsharp mask at ay 25%/1 pixel if the image isn't very sharp. When done, use File > scripts > export layers to file > JPEG quality 7. No need to make each image 20 KB...
Find relevant gaming images. I strongly suggest not ripping off other games IP. If you target WoW gamers, use WoW images and send them to Wartune then you may get high CTRs and do alright, but the advertiser is unlikely to approve the ads (if creative approval required) and it can get you in hot water. Actually, Blizzard is known to give no fucks, even of their own users (thanks for the oceanic servers for Diablo III, assholes). But other gaming companies WILL get your ads pulled for IP infringement and/or pursue legal action. You become a better marketer when you make legit images work cf. ripping things. I'm a basterd and I report gaming ads that rip other games images because I think these affiliates degrade the platform for people like me who put time and effort in.
4. Be more verbose with your follow along details. What is your EXACT targeting? Show us some of your creatives. We might say don't even bother, they suck, saving you time and money. Don't be afraid to be completely transparent when first starting out. You may think others will rip your ideas but lets be honest, the benefits you may get from that transparency far exceed the potential detriment from swiping.
Lastly, apologies if you felt a bit overlooked during the start of your follow-along - fear not, your questions will no longer go unanswered!
Hi. This was a SUPER helpful post. I do have some follow up questions though:
You say here: With gaming offers you can expect an EPC of anywhere from about $0.05 to $0.50 when direct linking. If an offer is performing amazingly, has a high EPC, then you would be a bit more forgiving with your ad-culling strategy. If the offer is generally giving say $0.20 EPC, then I cull quite aggressively based on what CPC the ads are going to get. Say an ad launches and pulls 0.156% CTR in a campaign at $10 budget. My bid is $0.67. The ad's CPC starts at 0.41 and I see it slowly drops to about 0.31. I got several conversions and the ad ran at about $0.23 EPC. I let it run the next day (but I don't touch budget yet!). It did similarly and had 0.141% CTR, CPC stayed averaged at $0.33, EPC dropped to about $0.18. At this point I would look at my campaign wide EPCs and see that $0.35+ is not happening, so this particular ad is very very unlikely to ever get profitable - I would need the EPC to jump up 50% or CTR increase dramatically -> CPC down. I cull it and focus on improving my ads.
So based on this you are using the CPC vs EPC as a measure of whether or not to kill an ad. Is that correct? It would make sense since that formula would determine your profit. In the above example we assume that you have had ads convert that is how you know the epc? Is that correct? The strategy I have been employing is 1 ad per campaign group. What is the formula to come up with EPC (or is that listed in the getting started guide where Caurmen breaks formulas?) If it is I can look it up. When you are doing the math for CPC are you doing it based on the converting ad (as in per ad) or based on cumulative (say you have 10 ads running and only 4 have had clicks and 2 converted, making these numbers up so not sure if it makes sense). Also why in the above would the EPC drop to to $0.18? Is that because in this example there are less conversions (because you said the cpc stayed the same). Again this brings up what seems to be an age old question for me, how many clicks then is enough to determine that the sample size of clicks is large enough to be stastisically significant or A) does this not matter or B) is this something like one does daily to keep an eye on the ads? Sorry for all the questions. This is what i get for smoking pot and partying instead of attending statistics in college. I knew that would come back and get me one day!
You say here: I target games based on their genre. I would target lots of MMOrpg type games together in one campaign. I don't recommend mixing WoW in with these - do that separately. In my experience the WoW interests bring a large demo but the responsiveness (i.e. CTRs) are low. I would target browser-based games separately, FPS's etc separately, and so on. Don't target WoW, Assassin's Creed, Call of Duty, BF4, DOTA and Final fantasy together. Recipe for disaster.
What I have and am walking away from is a browser based game that is an MMorpg. I looked up the top 10 most popular MMorpgs and put them as the interests. Warcraft was one of these as was League of Legends, etc, etc. These actually, believe it or not seemed best, but i have better ideas to test for the future. The way it is currently running (and being destroyed), is girls (this is a girl game), and 18-19 & 22-23 (most responsive ages) and the same ad and 14 different pictures to each with the only precicese interest being video games).
You say here and I am very confused by: Also, when people say CTR they mean true CTR, not unique. I consider unique CTR as a measure of 'audience engagement', i.e. how much users like your ad - as it represents the % of the audience that have clicked. However, real CTR (clicks vs number of times ad delivered) is the performance measure which also influences CPC and the FB delivery system.
This doenst make sense to me since with smaller demos the same person is bound to see the same ad over and over. Wouldn't the unique amount of people who saw it vs click it be a way better measure? I guess it is moot based on the influening cpc and delivery. This still doenst make sense as that means every ad is doomed to die much sooner then later based on the size of the demo and how well you targeted. (I realize all ads die and there is banner blindness eventually). This woudl make it a never ending race of launching 20 ads a day (random number I chose) and constantly trying to optimize. It makes me think of a boat flooding and all day you are just trying to get water out of the boat so it doenst sink. This is confusing because the ads cant possibly last more then a day or so.
You say this: 1. Make sure you have a well-designed tracking setup that protects your ass from avoidable disapprovals. When first submitting ad I recommend having the offer URL in CPVlab or wherever as the offer page directly. No affiliate link. This way there is no chance of the FB reviewer being bounced by your affiliate link to an unrelated offer that is non-compliant. This is the most easily avoided yet most common way new affiliates get in trouble. Once ads are live, switch to the affiliate link and if possible add a rule to the campaign that says, for example, "if clicker not from US then send to offer page else send to aff link".
I am using CPVlabs and I have had this hapepn. Luckily the network is redirecting them to an ad that is compliant so the only ads that have been rejected have been for being too sexually explicit. How is it possible that you can just switch the link? I see bots CONSTANTLY visting the ads. If it was that easy then it would be super easy to hack peoples computers via drive by downloads and watering hole attacks (I come from a hacking background). Surely facebook must take some kind of checksum of hte offer page (maybe like an md5 hash) and when they see the page is changed have a policy in place to deal with that (otherwise why are the bots constantly visiting the site, can't just be to see whether the link is up or down). The solution sounds too simplistic. It reminds me of the saying if it is too good (simple) to be true..... How do you redirect an entire country? Is that something only Volumm can do or someting i have to set up on the linux box cpvlabs is running on? Can it be done through cpvlabs?
You say (sorry I suck at multi quoting I am actually copying pieces of the thread into notepad and typing this out here before i post it): . Try oCPM with default bids. You would be surprised how well these may do compared to traditional CPC ads. Use Facebook's conversion pixel if possible so you can see cost-per-lead values in the reporting system.
You really trust the oCPM at default bidding? Have you ever done oCPM where you control the bids? I am confused by the fb conversion pixel comment. Do I give that to the network along with my cpvlabs link or does the facebok conversion pixel go into cpvlabs?
You say: 3. Put more effort into your gaming ad images. Gamers are easily one of the more visually inclined demographics. I use photoshop. Make a new 100x72 image, drag and drop images in, resize to look decent (go for size = clarity over fit everything into that tiny space). Use unsharp mask at ay 25%/1 pixel if the image isn't very sharp. When done, use File > scripts > export layers to file > JPEG quality 7. No need to make each image 20 KB...
I just got photoshop a few days ago and downloaded a shitload of tutorials from thevault & bitme on it. I will definitely put that advice to use. I am guessing what most people do is create a shit load of templates (not sure what they are called in photoshop I think actions), save those and then batch the images. They grab 50 images and run it through the effects then take what htey like. Almost like what crazyctr.com does. So size / clarity > fitting the whole thing in? Want to make sure i got that. For the images I did a search for the name of the game on google and used generic fanasty images not from games. I dont want to use the standard images I am given because it appears, for someting that takes hard work, most affiliates are lazy and I am gambling on the fact that they are using them.
You say: 4. Be more verbose with your follow along details. What is your EXACT targeting? Show us some of your creatives. We might say don't even bother, they suck, saving you time and money. Don't be afraid to be completely transparent when first starting out. You may think others will rip your ideas but lets be honest, the benefits you may get from that transparency far exceed the potential detriment from swiping.
I am not worried about getting swiped. That would be to other n00bs deteriments since I am obviously sucking now (and experienced guys have way better shit to do). I wasn't exactly aware of how specific I should be. And i 100% agree, learning the process and what i will get from pure transparency FAR outweighs the risk people jacking me. Like I was telling my boy last night: I have spent almost 200 and I am well ahead of most people who spend 200 (partly is because i have friendshps with some top level guys that date back to before I even entered the industry. Without them I woould be sunk.)
And thanks for bringing my thread back to life dude. If you ever swing through NYC let me take you out and show a good time.
I am sure this is 100% random but after pointing out how a hack could be done by switching the link the offer has converted 4 times. All from facebooks ip space. As in facebook went through and tested the offer all the way to the point of conversion. Wouldn't be too hard for them to do that. They read this thread, saw what i suggested as an attack and then searched the messaging on fb which has the username for this account on it, linked it to my fb and then tested my ads. Either way thanks for the 4 conversions fb!
Ok, first things first - your last post - how are you sure these came from Facebooks IP space? How did you come to this conclusion? Facebook IP ranges very but they will often have an ISP = Facebook, Websense etc. amongst others and it's often through a Linux-based VM. It would be very unorthodox for Facebook to care to that extent let alone signup at the end point of your ads. That aside, I would highly recommend not having anything on this forum that can link you as a user to your specific Facebook account. I'm not sure how you think they tracked you down? If your account email is visible somewhere then remove it.
Right, onto your questions:
1. Yes EPC vs CPC for killing the ad. At the end for the day it's about ad revenue vs ad spend. Indeed you need to know the EPC which requires conversions. EPC = earnings per click = revenue/clicks. EPC is generally one of the stats you can easily see network side, otherwise just take ad clicks and divide by revenue. You need to look at EPC sensibly when it comes to it's significance - most gaming offers have payouts of say ~$2 so if your spend $10 daily on an ad you might expect 2-10 conversions. You really want at least several conversions daily per ad to get a good feel of their performance. You can push for large sample sizes, e.g. 100 clicks minimum per ad spread across several days, and/or at least 10 conversions, but you also need to establish a system that lets you cull low-performers based on how you know the market performs. E.g. if an advert gets 30 clicks and one conversion one day, 24 clicks one conversion the next, but the campaign-wide average is 30/6 then I would hedge my bets and cull that ad before I get to some arbitrary 100 clicks worth of data. You need a strategy when starting out that avoids you bleeding money on flops.
I compare the CPC of an ad to the EPC of that specific ad when it comes to profitability decisions. I then use the campaign-wide EPC as a reference for what the average is for that demographic/angle/ad copy/etc. If the EPC I would need to make that ad profitable is well above the average EPC then I would likely kill it. Note that you should definitely separate these averages when it comes to age, gender, targeting and importantly ad copy.
2. Ok so here you made a grave mistake - don't mix games like WoW with LoL, DotA, etc. They are hugely different, to the point that I would say there is absolutely no correlation between someone playing LoL and their likelihood of playing a traditional MMO. Browser-based is another thing - I would be cautious about mixing browser-based MMOs and dekstop/client based ones. They are very different. Most browser-based MMOs are crap - just micro-transaction riddled low quality games spat out by hundreds of greedy developers. On the other hand, most MMOs start out as upfront costing and/or subscription based games, that then later may go free-to-play. The userbase here has made a decision to invest money in the game and they have different expectations of an MMO.
This is not to say you can't convert both audiences - you certainly can - but you'd be trying to use the same paintbrush on different canvases. Oh and 18-19, 22-23 is probably a bit tight for age targeting IMO unless you have a huge audience - which you probably do because of WoW + LoL, but as I said that's not an ideal combo (actually historically I saw that theses interest-demos responded poorly to MMO-type ads).
3. CTR generally determines/influences your CPC and ad delivery behaviour - unique CTR might have an impact but FB doesn't divulge much about ad delivery specifics. All ads eventually die - or the offer does - but remember that the audience can be very large and changing day to day - not everyone logs into Facebook every day and sees your ad. If your audience is 200,000 people, then 100,000 impressions every day might net you 200 clicks. The next day You get 100,000 impressions, 1/3 of the userbase is different to that of yesterday, people who saw or overlooked your ad previously see it now and click > you get 200 clicks again > repeat. This could last weeks before CTR starts dropping. Remember there is the potential for every user to click on your ad so there is a lot to be sucked out of an audience. I think you may be overestimating audience burn out rate.
4. Firstly, the network sending people outside e.g. the US to a 'compliant' offer is irrelevant - we don't want users being sent anywhere other then our offer. Screw the network, this is your traffic and you will send it where you want it. You say you see 'bots' constantly visiting the ads - this will happen with every ad submissions, but is unlikely after that unless the ads are a bit risky and FB deems them worth keeping an eye on. How are you detecting these bots? Are you sure they are not real user clicks? As for link switching, we don't want to do that we just want to have rules in place that say if country = X send to link A and if country =/= X send to link B. I'm not sure if there is native geo-redirection in CPVlab, but you can certainly do it in
As for your hacking-related comments - you could certainly send people to a malicious page but why? To add 10 out of 200 users to a botnet and simultaneously get your account, CC banned and potentially other repurcussions (e.g. hosting company bans you and other hosting companies also weary of anyone with your name or CC). Facebook doesn't check pages for changes as far as I am aware - landing page rotation and split-testing is an integral part of IM so this would be pointless. They will however catch wind of uncompliant activities, often due to users reporting your ads.
5. Yah oCPM @ default has generally outperformed clicks and actions for me, but this won't be universal. It is the place to start though. Why make it more complicated when you don't yet even know how it works and or performs with your campaign? The FB pixel must be placed on the conversion page so will need to be given to your AM to place or placed in the networks system - e.g. in Cake you can paste it in the pixel box.
6. I don't use batch filters for editing images in the way of filters. I spend more time getting my first batch of ads looking good on a per-ad basis. I just drag n drop them all into Photoshop and resize/move the layers as I want, then flick through them, adding an unsharp mask if necessary, doing some blurring of overly detailed backgrounds, etc. I am very picky with the images I use and how I present them - my main question is always "at 100x72 does this look good and can the user instantly comprehend the image". So, I try to avoid overly detailed images where it's hard to even see wtf is going on. In the end testing and the data reveals all. This industry rewards creativity, perseverance and importantly doing for yourself what others wont - put in more effort than joe average and you usually get more out.
7. Excellent. Keep hustling and learning.
Lot of good advice here, curious to know whether fighting ended up turning this around?
bump 
Home > Facebook & Instagram > Follow-along Campaigns