Trial and Error-ing my way through Mobile Pops (48)
Hi everyone,
Just a little background with where I am at.
I haven't had a lot of success yet in my campaigns and from what I've read there are a few things I am going to be doing differently.
Mistakes I was making
-Only running one offer per geo through my funnel and then if it didn't do well, I would try to get a bunch of new landers and run in a different geo.
-Changing too many things while the camp was running (like adding new landers)and not letting it run it's course
-Having my tracking server based in US when I was running international offers. (speed)
-Not optimizing my pages for speed
These are just some of the mistakes I have been making and I am going to work on correcting them.
I'm sure I can improve in other areas too, which is why I've created this follow along to get your advice!
I was using Funnel Flux which I really like and using amazon cdn, but I don't have enough confidence in my technical abilities to optimize speed this way until I've gotten a better handle on affiliate marketing overall.
I'm going to sign up for a vps/cdn and
The other big thing I am going to focus on is getting 2/3 offers per geo (2/3rd tier geos) with 3 camps staggered bids/10 landers, which I saw Vortex give the advice to do on another follow along.
One of my biggest weakness in life is consistency with regards to my personal goals, which I'm sure others can relate to, and I want to tackle this not only by working on am every day, but also by updating this thread at least twice a week (if not more).
Here are my stats so far this year 
although this is a little skewed in the positive direction and is closer to -30% roi.
Anyway, today I'm going to get
I applaud you for taking the time to sit back for a while and analyze your own mistakes, it's always a good thing to do - after all you paid for this, so make sure you learn from them
Speed is important, maybe the most with mobile pops, so it's a good idea to optimize it as much as possible. I'm not sure how big impact the switch to
Testing several offers from the same GEO is a must, do not limit yourself to 2 or 3 either, test everything you can find. Having the right offer is one of the key factors these days. Staggered bids and multiple landers is a good approach too, but maybe don't start with that many at once, you would need to spend a lot to properly test every possible funnel.
Changing variables when a test is running is a NO-NO, you properly identified this problem. You must "give" enough traffic to every change and it's always the best to change one thing at a time.
Your plan looks solid, looking forward to your first results
Originally Posted by matuloo
In your opinion what do you think is a good middle ground for the amount of LPs/bids I should test? How many geos do you think I should test at a time? I don't want to over spend foolishly starting out, but I also want to make sure I'm testing well. Let me know what you think!
I have found a ton of potential offers across small geos from my networks. I am hitting a little snag there with almost too many choices.(paradox of choice) I'm going to be looking at what geos are performing well across networks to see if I can test those campaigns first and speak with AMs. Today I will pick at least one geo and then the offers/lps as well and get things rolling! Stay tuned for the STATzors.
Edit:Got 1 camp up with 3 offers, launching 1 more camp same vertical same geo. Then setting up 1 in diff geo.
Stats tomorrow morning!
Originally Posted by beerbelly
Start with some Tier 3 geos imo, always good to make sure everything is setup correctly, then you can work on optimization process with feedback here from the forum.
Originally Posted by beerbelly
In your opinion what do you think is a good middle ground for the amount of LPs/bids I should test? How many geos do you think I should test at a time? I don't want to over spend foolishly starting out, but I also want to make sure I'm testing well. Let me know what you think!
I would recommend to test in batches and start with a low bid tactic, without going for staggered bids immediately. Take 3-5 LPs and about the same amount of offers and try to pick the best combination from these. In case you succeed and find something that is giving you stable conversions, isolate that funnel and test staggered bids to figure out what level gives you the best results. So first goal is to start getting some conversions, then test what bid level works best.
Once you have this, add more LPs and more offers and keep running at the bid that showed you the best results.
As for the GEOs - you should stay in 1 or 2 until you get a good feeling about what you are doing, then advance to as many as you can handle
So I ended up keeping my tracking server in USA and just using the google cdn page load speeds were still under 1s.
I set up my bids just a little below the average.
I launched 2 camps in the same geo/vertical yesterday. I sort of went with mr bruans approach of testing offers with 10x offer payout or at least 10$. I wasn't sure if I needed more since there are 3 offers total per camp.
My screenshots were .....not great I had to put them in photoshop to blur lol
Camp 1 stats
3offers
8landers
payout: .8
spend $10
The stats are broken down per offer
http://prnt.sc/czb3d4
The top line is the total of the 3.
Here it looks like the 3 conversion offer may be the winner but it's not statistically significant since another offer also had 1. I am thinking I will cut the offer with no conversions.
There were only 4 conversions total and spread out between landers so not enough data there.
Each offer got roughly 3.33 of spend
so offer 1 2.40 /3.33 with a ROI of -27% is the winner of the day lol!
Camp 2 stats
offers:3
landers:8
payout:.8
spend:$10
stats are broken down per offer
http://prnt.sc/czb3qd
The top line is the total of the 3.
There were no conversions on these offers so I will not spend any more money here.
I also used a bot test page, but I am unsure if I want to cut any of those placements since I'm sure I did not spend a ton per placement, have yet to check those stats though.
I'm working on setting up some other camps in this geo today and running more traffic to camp 1. I'm going to keep focusing on this geo until I've ran most camps I can find then add another geo to start testing as well.
I've also realized maybe I don't have as many offers as I want, haha yesterday I thought I had too many and today I don't think I have enough. I'm just trying to find things for the same geo a lot more before I jump to the next, and then looking at other geos I don't see as many low payout offers in the same geo spread across the networks that have the most sweeps offers for me. Anyway going to try to get into a few more networks to see if that will help at all.
Launched a new camp in new geo will post results/ going to be setting up more today. I have realized I need to take a look at my budget per camp and decide on a set number. Would you guys recommend the ::: (# of offers)*(#of Landers)*(payout average)*3.
I was doing just 10x offer payout, but am worried I am not getting enough data and dont want to hamstring myself.
Originally Posted by beerbelly
Last night I ran camp 1 again and a new camp
camp 1
I cut the bad offer
offer=2
landers=8
payout=.8
spend=10
conversions=0
Have paused this camp
Tested a new camp yesterday as well in a diff geo
camp 3
offers=2
landers=5
payout=.75
spend=10
conversions=1
With both of those camps doing poorly I am setting up some super low payout camps. I launched one already and waiting on stats will post tomorrow!
Nothing doing with my first campaigns getting back into am but that's ok.
I am really having a debate of whether I should just stay in one geo and test like sweeps/dating offers/av pins. Or move from geo to geo and just stay in the sweeps vertical. I can launch at least 2 camps a day if I find stuff to launch and for the next couple of campaigns I launch I'll really be focusing on 50 cent or below payout offers to see if I can just get some conversions and get some flow going. Input/advice always welcome
Elloooooo friends
Another conversion came through the france camp --still only $.80 rev on 10$ spend. So for now that offer is staying paused.
The new campaign
Campaign 4:
payout:$.07
offers:1
spend:$10
rev:4.55
This campaign could have some potential, at least there are conversions .
It looks like lander 4 is the winner, I am going to at least cut lp5.
At this point would you go with the winning lander or get more data? 
Every lander has spent 10x the payout.
My optimization Hierarchy taken from @mrbruan
Offer
Lander
Device
OS/version
Browser/Version
ISP
Categories
Placements
So with that in mind should I cut the landers or send more traffic or move further down the optimization pyramid?
Realization:
The thing that takes me the longest is finding more offers to test. Most of the recommended sweeps offers I get are over $1> payout and I want to focus on the super low payout offers, like $.50 and below. If I can figure out a way to speed this up I would be able to get a lot more campaigns up a day.
Goals Today:
Set up another campaign or two today to have some new camps to run traffic to over the weekend
Optimize Camp 4
Happy Halloween!
Over the weekend productivity fell off due to some people coming into town.
Launched another camp with .08 payout only 2 conversions with 10$ spend, gonna double check to make sure that the funnel was working correctly.
Other camp stayed at around -50% roi when got rid of the other 2 offers, went down to one lander, will make 2 variations of said lander, and will update stats when I run again.
When you guys are optimizing a camp do you try to run the offer the same time every day or make the changes then just run it again that day?
Happy Halloween! (even I never really celebrated this American holiday)
I just read through your first post and it's really great that you can honestly say and see where your downsides are.
What you need to do now, is to focus on them and improve them.
I know sounds easy, but basically that's it.
Regarding your campaign, I see you are not doing bad, you had -30% and now you are around -50% right from the start, which are honestly good numbers for someone who is starting out!
You must realise that in this stage you are collecting data and then you can act on it. You will start seeing specific landing pages, offers, zones, browsers, models etc. that are working and vice versa.
Erik
Continued with campaign 4 Today
Set up 3 variations of the best lander
The first lander without alterations performed the best 
.
Spend:$10
Rev:4.97
ROI:-50%
So the next optimization down my list is Devices from that data I don't think I see a clear winner, but wanted to see anyone else's input. The blackberry devices (rim)look good, but at the same time when I checked the inventory for those devices it was pretty low, less than 1k impressions per day.
I took out the 3 variation landers I created and ran a test for 2.20 of a low bid and had an roi of -43% which is a little better, I'm going to need to do more bid testing later, but just wanted to experiment a little.
Maybe cutting apple traffic would be wise?
Spend 2.30
Rev: 1.30
ROI:-46%
I still need to test for os/version and browser/version
Then Isp and then placements, I also have not done a bot test yet. All in all there is some work to do still on optimizing this campaign.
I am going to work on going through at least the entire optimization pyramid I am using tomorrow to see if I can't start finding some greens. Ie. Running 10$ spend after optimizing for each level and going on to the next, I had been sort of doing it one day at a time. I sort of realized speed in pops is important and was wondering if that was the right way to approach it. Instead sort of how @erikgyepes runs his camps all day I'm going to try to optimize continuously during the day tomorrow. Shall report findings
Also for the rest of this week I will launch camps for new offers at a rate of 1 per day the rest of the week.
The reason these tests might look a little weird as far as the spend goes was because I was let the first 1 go for 2 hrs and the second one I let deplete what I had left actively in that budget. Still way more than 10x the offer payout so I thought the data was still reasonable to draw conclusions from.
So I went ahead and got rid of Apple Devices for the next test
payout:.07
Spend:7.91
Rev:4.97
ROI:-37%
Saw ROI was moving in the right Direction and then just checked categories and realized I had left all of the adult categories in
Did a small test without them.
Spend:2:30
Rev:1.61
ROI:-30%
I did notice there was an ISP with about the 2nd highest amount of traffic that was unprofitable throughout the optimization and remained so after the last cut, going to nix it and test.
I set up a new campaign/offer as well and will be posting data for that as it comes in.
Update:
Camp 5:
Offer:AV 1 Click India
Payout:$.37
Spend:10
Payout:0
Ran Camp 4 again
Didn't change anything was just getting a little more data for significance
Spend:9.97
Rev:4.69
ROI:-53%
1 Step forward and 2 steps back
I'm at a loss of what to do next, should I just test new offers, should I go a few steps backward in my optimization.
Could it be the time of day that I ran it?
Could it just be this day of the week does worse for the geo?
Is this traffic source or offer just saturated?
Should I literally just stay in this 1 geo and test any offer I find even if no one is running it lol what's the best path to success here?
haha too many questions
I decided to pull my best performing placements from campaign 4
Lo and behold
ROI:35%
Spend:3.94
Revenue:5.32
Profit: 1.38
However the availability is 5k impressions a day as you can see from this spend lol
My total spend on this campaign has been
Spend:78.49
Revenue:36.26
ROI:-53%
Meaning I'd realistically have to run this camp for 31 days at this profit rate to break even. But it is nice to see some green even if it's just a super small amount.
I think what I'm learning from this campaign is I really didnt have a great lander. I'm gonna try to test some landers on the main campaign because if I just keep cutting I'll be left with something similar on the main campaign(very little traffic).
Are there any benchmarks you use before optimizing past lander?
I think one other thing I may have learned was don't overspend on data, spend until statistical significance and stop, dont just spend to spend. I think I may have been able to make some optimization changes with statistical significance without spending as much.
I also see how potentially I could use different offers on these placements and maybe that will make the data I got pay for itself.
I think I also could have created a whitelist camp of placements converting well with Apple which I'll probably try to do.
It's also been a little frustrating that every other offer I test isn't converting at all.
I really ran into a wall the last 2 days of like "ahhhh im never going to be able to do this" "ahhh I dont want to run super blackhat landers"(I'm not, but worried that's the only way I'll be super profitable) "ahhhh I should get a soda" So yesterday I set up the whitelist campaign and drank some soda....super unproductive LOL. Even my best performing lander currently has that FB look, idk why all of a sudden I'm all worried about it....probably just a success barrier or easy reason to give up, but at the same time just don't want to do anything legally questionable to open myself up to" have a bad time". But ill keep pushing forward!
Anyway I'm gathering myself and trying to launch some new landers/camps today.
Creating some different variations of this lander and I'm gonna continue to try to launch 1 new offer a day.
UPDATE:
Ok so I actually did not have enough conversions with Apple brand to create a targeted placementcamp.
However I did create a campaign with placements that are converting and just want to see if I can fiddle around with that at all to get a little bit higher profit maybe like xx a day would be sick lol
It's just a matter of psychology.
When things work out you feel like a king of the world, but when suddenly everything drops for a longer period, then you think that your life sucks.
Keep moving forward, fail faster to succeed faster.
The thing is that there are 1000s of offers, and honestly most of them suck.
Your job is to find that 1 that will work and help you to become that king of world (at least for a while).
BTW what geos and verticals are you testing?
Originally Posted by erikgyepes
Vertical-- sweeps -- But I am open to testing like low paying dating offers/av offers if they are on top offer list because I just want to find something that works lol
I've only tested 1 AV in India-- I ran like kind of aggressive but instead of saying you have a virus I made it all say "you may" 0 conversions and didnt run the google lander LOL
Ive tested sweeps in Thailand, France, Belgium, Indonesia so far.
Set up a new camp for sweeps in India that I will be report stats from soon as well. I'm trying to stick to the lower tier geos and recommended offers but they dont always align.
Quick Update:
Camp 4 Whitelist
Spend:$3.11
Revvvv:$5.39
Profit:2.28
ROI:73%
It felt nice to see two days of profit even with it being so low. How it went up to 73%ROI I couldnt tell you I changed nothing but I dig it.
Whitelist 2
Spend:$4.52
Revvv:$2.87
Profit:-$1.65
ROI:-37%
Not really a winner with those placements although a few of them were profitable/broke even will try again with them, and really I dont know if I can optimize the main campaign any further unless I try a different traffic source since there is abt 120k impressions left and I already took out my best placements. But you tell me, maybe there is.
Update
Camp 4 Whitelist
11/6/2016
Spend:$4.29
Revv:$7.35
Profit:$3.06
ROI:71%
3 Days of profits in the Whitelist campaign is pretty nice, running it today and running some other whitelist placements in another camp to see if I can scale it up at all. I may also try to run on another traffic source with the optimization I had used just to test.
So far it's not doing well today, but mondays have been bad in the past for this geo.
Camp 6 --India/Sweeps camp/landers 5/ 1 offer
payout:.04
Spend:$10
Revenue:$.68
ROI:-93%
Part of this is due to one of my landers having a mistake I thought I fixed but it didnt update on the CDN, but still doesn't look great. Will probably not run any more traffic may try a different India offer or if I can find some across multiple networks would be even better, my CTR avg on my landers was 20.45% so I think it has more to do with the offer than the landers.
Today:
Launching some new campaigns in MY/ZA/IT hopefully I can find a winner.
Update:
Wow the traffic seems crazy expensive for ZA --recommended bid for propeller ads is $12 CPM ..... with a .2 payout need 60 conversions per 1k impressions or a 6% CRV .... I'm getting like .5% CRV on my other offer lol and even in United States it recommends $7 CPM, the bid prices look a little better on ZeroPark may test there. Going to run MY/IT first now.
My account was also suspended from Zeropark for inactivity and I couldnt reactive with the information I sent them which was basically the same questions from signing up. Maybe I claimed too low of a budget, I emailed them back asking why I couldn't be reactivated and waiting to hear back....that's a hurdle I was expecting to hit lol
Update 2
11/7/2016
Camp 4 Whitelist
Spend:$10.16
Revvv:$15.47
Profit:$5.31
ROI:52%
^changes I made were just upping the bid I was way below average and by doing that I lowered my ROI percent but my overall traffic/profit increased!
I had played with the bid a little bit, but now I'm kicking myself for not testing it more.
Camp 4 Whitelist 2
Spend:$2.52
Revvv:$1.33
Profit:-1.19
ROI:-47%
So these placements that were also converting are still doing poor, I upped the bid on these as well but it didnt help, after a few hours I just paused this camp.
Got some camps with new offers set up just waiting on approval
Campaign 7 --France/sweeps/2 offers/7 landers
Payout AVG:$.57
Spend:$25
REV:$2.75
Profit:-$22.25
ROI:-89%
This offer did poorly if an offer doesn't do at least -50% at the beginning it's not worth doing is it? At least for a "noobie" like myself?
Camp 4 Whitelist
Spend:$14.54
REVvv:$13.79
Profit: -.75
ROI:-5%
I wanted to test the higher bid earlier in the day and it did not work at all, once I put the bid back down for a little bit it gained ground on the losses it had acquired.
A camp I had set up I couldnt get approved to since the cap is already having issues. Having a hard time finding offers in tier 2/3 geo with low payouts to test. I am setting up an italy campaign currently then I might set up a campaign in spain. I am wondering if I should start testing higher payout offers.....
I'm also working on setting up 3 new campaigns ---should i test two traffic sources for each campaign or stick with the traffic source ive been using and split test carrier vs mobile traffic-- i had been keeping that in the same campaign--- which of those two tests would be the bigger bang for the buck so to speak
Nice to see your first greens!
I had played with the bid a little bit, but now I'm kicking myself for not testing it more.
Originally Posted by erikgyepes
I think I'll set up a 2nd camp and run two bids and see how it does...I may have burnt out the placements though lol. I'll stay with the lower campaigns as well I want the data, it just seems like there arent that many recommended campaigns that fit the criteria. Should I start doing offers that are not recommended?
Here's results from the Newest test
Camp 8 India- sweeps 2 Traffic Sources / 2 Offers / 11 landers
Traffic Source 1
Spend:$15
Rev:$.99
Profit:$-14.01
ROI:-93%
Traffic Source 2
Spend:$10
Rev:$ .17
Profit:$-9.83
ROI:-98%
There was 2 more conversion that registered as organic traffic that I believe are from TS 2 but I counted them in the revenue.
Offer 1 REV: .76
Offer 2 REV: .33
I tried to run an angle on the fact that there RS 1000/500 currency was discontinued, but it didnt even outperform anything, haha still thought it was a cool idea!
Biggest takeaway from this test is there is a big diff in traffic source and offers between networks which is said a lot, although would need to spend more to get that 95% stat sig.
Kinda wish I would have named my follow along - Trial by FIRE (-99% ROI)
I may try to see if there are any profitable placements I can run traffic to for the India offer.
One more question I have.....do you think it's better for me to just stick to one campaign and just run the hell out of it to try and get that 1 campaign to work...or is it better to just be testing as many offers as I can. I'm guessing they're may not be a right answer to that question, just been thinking about it lately from the mindset of what can I do differently.
Update:
White CAMP:
Spend:$6.24
Rev:$4.41
Profit:$-1.83
ROI:-29%
2 days of losses
haha. Could be multiple things, could have burnt out placements, could just be bad days. Either way I'm duplicating this camp and doing a bid test tomorrow to see what's better a low bid or high bid.
Also will try to test it on another traffic source.
When I started this FA I talked about how I wanted to test more offers in a geo before moving to the next one. I am not doing that well, and not sure if that's the best thing to do with the data I've been getting and the offers I can find.
I have not changed my tracker still using server based in US(trying to figure out best move for that still, like if there really is a big difference in speed)
The one thing I've done a lot better this go around is letting my tests go all the way through without changing them, I used to like change landers on the fly if I saw one doing well LOL. I do play with bids but I think that's pretty normal.
I'm sort of throwing darts at the wall with a blindfold on trying to hit a bullseye and expecting success/to hit a bullseye just because I'm throwing them, I'm still grinding out putting together camps and taking action, but I think I can be doing it smarter.
I think overall I need to get a little more methodical in my testing methods and my analytical methods, going to spend the day outlining what that might look like. And I have a new camp ready to launch as well.
Alright Alright some updates
So I did do a bid test on those placements....maybe not the best bid test in the world I realized after but still two data points were had
Camp 4 Whitelist:
High Bid
Spend:$7.38
REVV: $3.36
Profit: $-4.02
ROI:-54% (Ouch)
Low Bid
Spend:$4.41
Rev: $3.78
Profit:$-.63
ROI:-14%
Low Bid worked better here. Both campaigns had the same CRV .45% so was kinda curious. Also overall looks like this camp may have gotten saturated/placements saturated/ or something in between. The offer does have a cap on it as well so makes sense, and my AM told me that one of his pubs was doing 12k a day and the limit now is 5k. Running one more time just to see, and working on some other camps.
I ran the Camp 4 Whitelist campaign(lower bid) just a little bit more of traffic
Spend:1.37
Rev:1.33
Profit: -.04
The issue is the majority of the conversions came from computers which I'm not even targeting....has anyone else ran into something like this?
Got a Brazil campaign up and running 2 offers/7 landers/ 1 TS / payout - avg $.11
Was able to get the best offer off first test and the best 3 landers off of second test.
Letting it run now and will see how it goes.
Camp 9
For better or worse somehow my traffic has been showing up on computers I guess.
Spend:$8.48
Rev:$2.6
Profit: $-5.88
ROI:-69%
First thing I optimize after offer and landers is device. On my traffic settings I had selected smartphone/tablet, yet somehow I'm get computer traffic. Not necessarily a bad thing since the offer does accept this traffic and it's performing better. However, not able to utilize this data to optimize my mobile traffic.
Not sure what to do with this brazil camp I might try one test of just desktop traffic and see the results.
Ok im starting to just completely doubt everything I've done the last 2 weeks LOL my previous best performing sweeps camp is showing most of it's conversions coming from a computer as well....which is ok for both of these offers but still not exactly what I thought I was doing. I am starting to feel very foolish here.....everything I believed in was a LIE!! lol
I never checked device type because I assumed it was only running traffic to smartphones like I had chosen. Lesson learned I guess, although the computer traffic is outperforming my mobile traffic which could be the way I optimized or could be that traffic is better/less saturated.
Update: I'm going to get 5-10 campaigns set up in different geos and run them all on the same day, that's my next plan of attack, I'm working on getting those campaigns ready. Hopefully I'll find a campaign worth optimizing!
Well one problem was solved!! The circle of lies continues.... The lie that I thought I had been living with my traffic coming from computers is a lie!!!
I guess funnelflux was mislabeling some smartphones as computer, going to try to use this data to optimize the brazil campaign as far as OS goes.
Camp 10:
Ecuador / sweeps /4 landers/ 3 offers /payout avg $1
Spend: $33
Revenue:$4.35
Profit:$-28.65
ROI:-86%
Worst part about these stats is the conversions are spread across the offers and the landers.
Anyway I'm going to start working on my mass campaign launch, I did launch this camp in ecuador--- and may try to further optimize the Brazil camp.
I did run the brazil camp - Camp 9 again at -70% after blacklisting the biggest unprofitable placements pretty much right in the same place, although one ISP looks like there may be potential.
I've got 9* campaigns in same vertical I'm going to launch at the same time across multiple geos, I've got 4 in the starting gate and just waiting on getting the rest approved. My first mass test of sorts and I think I've learned how to launch a little faster just from setting it up.
Serious question for the brazil camp I haven't split test my winning lander with any variations...is that something I should wait to do once I get in the green or at the very beginning when I'm at the lander test?
I wanted to mass test the offers, but I still never got approved for them and was out of town over the weekend. I just need to hit up my AM and make sure they get approved. I did run 2 of them though.
1 click offers -- Gameloft
Ran landers like -continue here/gift/a download/a picture of the game
Ghana
Spend:$10
Rev: 0
Nigeria
Spend: $10
Rev: 0
The worst part about running these, besides the 0 conversions was that I had forgot to set up my links correct me and it cost me an additional 10$ spend on each.
I am also unsure if I should test more 1 click offers....stay tuned.
Now looking for more offers to test.
Regarding recommended offers - you can of course always choose offer by yourself. AMs just trying to help you based on your preference that you say them
Thanks for replying erik you keep me going haha!! I've been spending the last few days working on a little ecommerce site just to keep grinding but switch gears a little bit, but plan on getting some more campaigns going here within the next 12hrs.
Haha, I'm watching you!
Sure take your time!
Originally Posted by erikgyepes
I've got some campaigns that I'm just waiting for approval on #weekend. How many rounds of optimization do you think I should do before I decide if a campaign is worth it or not?
I'm going to be trying to optimize them in the same day after each round of spend.
I guess my question is what sort of ROI do I want to meet after optimizing landers and offer.
-50% there with two rounds of spend? or do I want it to be higher for the campaign to be worth it?
I think this is something I am struggling with, im not sure if I should continue trying to optimize a campaign or just test more. I think it's somewhat obvious I need to be testing more overall since I've only launched 10 camps roughly in 3 weeks, but just wondered what you thought!
Camp 31 looks like it has potential.
Camp 24 you should be able to get profit within the next day or two max.
Show the stats for Camp 24 for the each lander since you are only running the 1 offer. Also, are the conversions coming from a bunch of placements or just a handful?
I can test more landers, lp2c was the continued winner for this offer and after testing the landers on 1 offer late yesterday I ran traffic just to that lp today.
1-1 camp24 Landers
1-2 Camp 24
Conversions are spread out among placements, one of the top traffic placements as far as visits on my tracker is converting, when I checked traffic it said 6k impressions per day, and it's converting a little above 4% so might be worth white list testing but it would be just another one of my dumb little camps eating up a testing budget
Camp 31
I'm going to grab some more landers to see if I can't get anything converting a little better.
Since I am still at -50%, I might need to spend more on that offer overall. It's the first offer I've used with payout over 1$ that's actually been converting so I'm a little loss without as many conversion points to look at.
#1 Mistake you just made.. you cut the landers for Camp 24 before you even gave them enough time to show results.
Just like you see when you continued running traffic to lp2c and comparing Jan 1 to Jan 2, the results went from 25% ROI to -22% ROI... one of the other landers could've had the opposite effect.
You need to run the test all over again for each of the landers and including lp2c. These are the types of mistakes, assumptions and early choices that will kill every good campaign you start to find.
**This is a prime example of why having a system and a process in place for testing is extremely important. You made a choice based on your gut (which is inexperienced at this point) and you failed to make a factual choice based on data.
Andrew
hmmm is this in part because I need to run all the landers over multiple days or I should just make sure I get that 95% statistical significance and the days don't matter?
I was trying to over speed this campaign up and being irrational.
The multiple day thing helps but isn't necessary for low payout offers.
You need to give all landers a reasonable amount of exposure to each placement sending traffic, since each placement has a different CR and the time of day will also greatly impact the CR on a campaign.
Your offer is .07 payout, so you need to be spending $3-5 per lander to determine a real winner using statistical significance. You are no longer in the "initial testing phase" so adhereing to small budgets for testing needs to be adjusted to fit the situation and find a clear winner.
Andrew
Ok yeah I was definitely still behaving in a manner that was akin to to the "initial testing phase" . Realizing as soon I started to post stats that I should not have cut the landers so soon and that I was being dumb. I reread through my entire follow along and have been realizing I really need to apply the advice you and others have given me better. I sort of follow it, but I think I may also come across like i'm just doing my own thing. AM has made me realize some of my lazy habits and my tendency towards shortcuts, that I was previously blind to.
I know it is in your post, but for the initial phase that's really just getting an offer that's converting?
If only one offer is converting or only 1 lander when you're just doing the initial phase it's ok to cut the other stuff? Then you'll want to get more landers or more offers for testing onece you have found some that are converting? If multiple offers are converting then you need to test those to statistical significance and then test landers to stat sig as well? Is this more correct?
Yes.
First goal is to find an offer(s) that are converting. Usually you want to take all of the converting offers and run it against many landers to find the best lander, so that enough traffic is going through to the offer page. Then after those two tests, you want to test as many offers as you can find. Then cut back targeting segments/placements, etc. Then profit.
Ok, well i'm running the offer tonight and will run 3-5$ to each lander.
Just set up another SOI offer with 7 lps and 5 offers and sending that to AMs to be checked -- so the goal of at least 1 camp a day has been met on day 1.
Still working on grabbing more landers for camp 31 and I'll check if it's on any other networks.
My am from avazu told me camp 24 is paused, I was running traffic and it was converting when I was told that, as well as the offer is still available and just says it is capped at 1000 leads? I have paused that offer for now, and have 2 I am just waiting on hearing back about landers. I had to make some changes and double check everything said chance to win, because I quickly threw some up and they said "had won" or "you are a winner". One of my ams had mentioned not to get caught on adplexity with those landers, did they just mean the offer owners check adplexity for violating landers or something else?
Also to your credit Andrew and to highlight my poor decision different landers were out performing my chosen lander. I didn't complete the spend on the offer after hearing that info from AM, but man did I want to.
Camp 31 I will be running more traffic to today as well.
Since I am waiting on various offers getting approved I will probably have a multi offer test day wednesday or thursday of this week.
The offer for Camp 24 comes and goes very often, so keep your landers and campaign setup for when it comes back. You can still run it now and just use the 1k lead cap, just resort to juggling the best 2-3 offers on the appropriate networks and you should be able to still make money just have to monitor when to switch the offer out. (Keep in mind, the same advertiser owns the offer, it's just rebrokered from network to network at a different payout)
And for the lander test, only $1 spent is still too early to tell. You will need atleast $4-5 spend per lander to get a clear idea of a winner, but atleast you can see what I mean about the difference in performance with a small sample of data.
Andrew
Here is a new camp I am testing.
Camp 32
Offers
Landers
I will get rid of the offer that has only 1 conversion should I just use the best lp to decide best offer or just run more traffic overall.
I have a pin submit iphone 7 camp set up just waiting to run traffic to it. I may revisit some of the campaigns I have tested in the past that converted with more offers split tested against them and landers.
Originally Posted by beerbelly
Camp 32
Offers
Landers
I will get rid of the offer that has only 1 conversion should I just use the best lp to decide best offer or just run more traffic overall.
I have a pin submit iphone 7 camp set up just waiting to run traffic to it. I may revisit some of the campaigns I have tested in the past that converted with more offers split tested against them and landers.
Got your PM and finally checking in. It looks like Andrew's been giving simply marvelous advice here. I haven't read through the entire thread, so would like to apologize in advance for any overlap in suggestions.I would suggest running more traffic before cutting stuff so you'd have more data to look for trends. Also would be good to drill down into Offer > Lander > Major Segments (e.g. OS/mobile carrier/browser...) to look for any green. Also - will there be additional offers you could test later or are these already all the ones you can find for this geo and offer type?
Revisiting old camps may be a good idea. You may already have a decent lander from a previous split-test, and may have already cut some bad placements. Taking advantage of data you've paid good money for is wise. Testing new geos can be exciting and may result in more profits, but then you'd need to start testing from scratch. Pros and cons.
Amy
Hey Vortex, thanks for the reply! Andrew's advice has been extremely helpful/useful. Things have died down in my follow along a little bit, but I'm still working on getting new camps set up, I may try some AV offers in the near future as well since I've hit a lot of the sweeps offers below a 1$ payout in the lower tier geos. (although I think I could have definitely tested them better like you & I touched on about going back to retest some previous camps)
In regards to camp 32
There are some landers that are green depending on the offer, and therefore some carrier/os greens but I guess the data is too spread out to really draw any conclusions. I'm running more traffic to that offer. There are also some other offers available in this geo from soi to pin that are the same vertical of sweeps but just a different prize.
Do you think I should test those other offers instead of beefing up the spend on this one?
I launched another camp I've had set up for a little bit.
Camp 33
Offers
Landers
It's a pin submit offer, so higher payout, and less traffic. I've stayed away from pin submits for this reason. The traffic source I've been using here only has 11k impressions at top bid for this geo/carrier combo. I'm able to get 6k. Is that worth testing on that traffic source further?
I have set up this offer on another traffic source with >100k impression potential and will follow up how that goes.
In conclusion, I've failed at launching a campaign a day, I'm getting a little demotivated due to my lackluster performance, and I'm running out of testing budget. However, I do feel that the hardest thing in a lot of areas is getting the first success, so I'll keep pushing.
Do you think one of the hardest things about AM is getting the first profitable campaign?
It wouldn't be necessary to test and cut the same offers and landers on multiple sources - that would be wasting money, which is not recommended especially if you're not working with a big budget.
Making great progress here! There's no need to force yourself to launch a camp a day. If you really don't feel like setting up campaigns, then do some learning either by spying in the wild or on adplexity. See how other people are running their camps to get some inspiration. Go read some success stories but only if doing so makes you feel good. Browse top offers lists from various aff networks to see what types of offers are doing well. Try to keep things light when you're feeling demotivated or frustrated. To me, trying to push through funk is not the best idea, and would more often than not just leave you more frustrated. Relaxing and taking a step back until you feel the inspiration to take further action may be a better way to go.
Amy
Ok the points you made, make a lot of sense. Thank you!
I'm running more traffic to both campaigns as we speak and will update with stats. It's nice to see a RON test perform at -30% roi even if it is on small spend/small amount of traffic.
I was going to edit out my mental headspace at the bottom lol because I felt it was a little bit too emo-ish (for lack of a better term)so I had to edit it back in, but I think you answered those frustrations pretty well. Pushing through the funks doesn't really work for me either lol, I just wish they would end quicker!
It feels overwhelming when things aren't working well but I also kick myself for not following through on my plans...but the past is the past, I think I should start with a more manageable goal for me and then build into 1 camp a day and more. I'll keep this thread updated on what that might look like as I see what I can accomplish this week.
So little follow up, camp 33 did the same with an additional 10$ spend at -30% roi, still only 6 data points because of the higher payout, gonna see if anything sticks out and still running more traffic there.
I also ran camp 33 on another traffic source on ron 20$ spend 0 conversions and paused that, but hopefully I can get some useful targeting on this traffic source and then expand over.
camp 33
offers:
landers:
Not sure what else to do here but run traffic as I don't believe I'm at stat significance yet.
Camp 32 yesterday
Offers:
Some lander combos are giving me -20 and -12 % roi still. And the carrier with the most traffic had positive roi on offer 1 with out optimizing landers which there I'm running 6 of.
Do you think it would be safe to cut the offer that didnt convert today even though it did better the previous day? or do I need to test the best lander against the offers? Or should I wait for stat sig? ,a
Working on some new camps as well, hopefully I can find a winner
Can't see your attachments. Could you please upload the images to imgur.com and post the BBCode links in the forum post? I went over how to do it at the start of this post:
https://stmforum.com/forum/showthrea...l=1#post300912
Will reply to your post after I see the stats in the screenshots...
Amy
Whoops, I messed it up yesterday while I was writing the post I connected to a vpn and it disconnected me. I thought when I checked that the links were working yet lo and behold they were not, sorry about that.
Camp 32 Today:
Offers:
Landers:
The lander performance didn't change that much, but the contrast was there were no offer/lando combos below -50% roi.
This camp performed poorly with more spend trying to get data pts. Offer 1 which was converting didn't convert at all, may just need to test some new camps and see if I can't get some better performance out of the gate.
Camp 33 -- This camp so far has spent 16$ and has 0 conversions so is also performing poorly from last night/today, but the spend has not completed so I haven't included those stats.
Home > Mobile > Follow-along Campaigns