Mobile Follow Along : Android App Install (34)
Hey Stackers
Tools:
STM Mobile Tracker
Dedicated Server: located in Philadelphia
My Brain
My iMac
WRW
Numerous posts from this forum.
I am currently promoting an Android app install in Mexico. The offer accepts multiple GEO's so there is scaling potential.
Offer payout: .22 cents. I am sure most of you know what this offer is.
I ran 3 different angles A, B, and C. With 4 different banners per angle and one LP per angle. I used 1 RON to target.
I bid $1.00 so i could get data quickly.
I threw $20.00 at each angle to find a winner.
I found my winner ( most conversion/best ctr ). 30 leads total. 18 of which came from my winning angle.
I spent $21 on the winning angle and made $3.60. whoop whoop. -83% ROI
Now. I know I need to lower my bid. But I am unsure what I should lower my bid to. My thought was .30 cpm. My ctr is currently around .9 to 1% .
I have also been tracking placements.
But using RON there are many different placements. Likely over 100.
I am having a hard time wrapping my head around my next steps.
Ok, It took some further research but I wrapped my head around it. 
Big thanks to Smaxor for this post on his blog http://www.oooff.com/php-affiliate-s...ng-cpc-part-2/
The CPM bidding can be based on your CTR and EPC.
Since I currently do not have a trusted EPC of my own, I asked my affiliate manager what his top affiliates current EPC is for my specific offer.
He advised me that his top affiliate is has an average EPC of .05
It is equally important for me to know what that top affiliates payout is. Now he did not tell me exactly. But he did give me a 2 cent pay bump. *cha-ching*
So with the 2 cent pay bump it is probably safe to say that I can aim for a .04 epc, because 1. I am not yet a top affiliate and 2. his payout is still probably higher than mine.
So being optimistic I will shoot for a .04 epc.
Now, I am still missing an important piece to the puzzle. I need to know my CTR in order to know what CPM i should bid.
Based off my initial RON test. My ads collectively had a CTR of 1.05% on my winning angle and that was including wifi traffic.
From what I can tell, that is pretty good considering I did 0 optimization.
So being optimistic, I will say that my ads are going to give me a overall CTR of .9% (9 clicks every 1000 impressions). That is what I will strive for anyway.
So knowing/guessing the above information and knowing the math behind it. I know that I can afford a CPM bid of about .40 cents and still have the ability to break even on max spend placements.
Ok, so my next steps....
The RON network that I am using has only app traffic.
So I will ....
1. Create a campaign for app traffic excluding wifi.
2. Create a campaign for app traffic w/ wifi.
Going to split test the same lander w/1 adjustment. ( the key is to do controlled tests 1 change at a time, so you know which specific thing caused the specific outcome)
Actually I am going to further my testing and play with the frequency capping.
3. Create identical 1. campaign with different frequency cap
4. Create identical 2. campaign with different frequency cap
Results coming soon.
Feeling good!
Let's do it.
Thanks for the detailed writeup. Couple of questions for you: So your traffic source doesn't give you a recommended CPM bid where you can bid a few cents over it? Also, since you were looking for CTR, is it possible to ask the traffic source what the competitive CTR is there?
Originally Posted by hankqny
To your first question, yes. They give me a recommended CPM bid. However, it does not necessarily mean I can afford that particular bid. In the beginning you need to figure out a realistic CPM you can afford based on your particular offer you are promoting.
To your second question. Yes, actually that is a good idea. I could ask my account manager what a good CTR is for the particular offer I am promoting. They may or may not tell you that information, it would be based on the type of relationship you have built with your contact at the traffic source. From what I have read from Mr. Green here on the forum you want at least a 1.0% CTR for mobile 320x50 banners.
How's the data looking from your respective placements? In particular, are there any placements that have eaten a lot of the budget ($1 or more of the clicks)?
Finally! An app install follow along. You seem like a very thorough guy. Best of luck on your journey, bro. subb'd!
Originally Posted by caurmen
Slowly but surely. I am running campaigns 1. and 2. as we speak. As the placements hit $1.00 spend I refer to the cpc if cPC is over .06 I block it. I chose .06 as the cutoff because I was testing some other ads I made and they are about .80% so I don't want to cut a placement too soon, just in case the lower CTR ads have affected the total cPC of that particular placement.
Ok.......*slams head on keyboard multiple times*
Just realized that my new campaigns are not passing the unique dynamic variables from the traffic source i.e. site id, creative id, etc. Instead they are showing the literal dynamic variables themselves i.e (siteid), (creativeid).
In my initial test campaign the dynamic variables worked fine. The only difference between this campaign and my previous campaign is that I am using a rotating lander and I have cloaking on now.
I am almost certain that the reason my dynamic variables are not passing is because of the rotating lander link.
Can anyone shed some light on this and hopefully provide a work around other than having a separate campaign for each lander.
For now I have paused all campaigns until I figure this out as I have no real way to optimize placements without the conversion information from the traffic source.
So the tokens are coming through as (siteid) for example rather than blank? That must be a problem with your ad URL or at the traffic source - what's happening on your end will have no effect on whether those dynamic tokens turn into real values.
Make sure the way you have put the tokens into the ad URLs is exactly as the traffic source specifies! I.e. {blah} not (blah), or {{blah}} not {blah}, etc.
Originally Posted by zeno
Thanks for your reply.
Yes, the tokens are coming through in there literal form (site), (creativeid)
When creating a standard non-rotating lander link in the STM tracker the link spits out the C variables with the proper formatting for the traffic source in the link itself ex: trackinglink.com&c1=(siteid)&c2=(creativeid) .
But, when I create a rotating lander link in the STM tracker the link spits out with the C variable looking like this ex: trackinglink.com&c1=%siteid%&c2=%creativeid%
So instead of it showing the brackets in the link it shows the percentage symbols.
Yes, that is it URL encoding it to replace unsafe characters, i.e. symbols which are not always accepted in URLs.
You have pasted this URL into the advert URL at the traffic source right? That's the problem. The traffic source replaces e.g. {siteid} with 123456. It doesn't replace %7Bsiteid%7D with 123456. Tracking systems, networks etc. will often to this to URLs to make sure they behave properly. They can't be expected to know that a traffic source intends to replace abnormal characters with simple text.
Correct the tokens in the URL used in the advert regardless of what p202 spits out. It might be easier just to add them yourself. I.e. generate the p202 rotating lander link without any c variables. Then just add &c1={siteid}&c2={creativeid} and so on.
Had to change up my links for my campaigns so tracking would work, which put all camps into a review queue for the weekend because traff src doesn't approve camps on weekends.
So I will be building landers and ads to split test in the meantime.
Staying focused.
These are my stats from yesterday
http://imgur.com/l4dmiCB.png
Pretty shitty.
I have been cutting placements based on cpc. But even the ones with low cpc kill me because my conversion rate is low.
After referring to my tracking software I can see only a couple of the spots have been profitable.
The RON exchange that I am using is only app traffic. I am thinking that app traffic is not the way to go with this particular offer.
I could be completely wrong and just have a really shit sales funnel.
Wifi and w/o wifi is relatively the same.
This stings.
Next steps. Launch a campaign on Web Only RON src.
Did some more critical thinking while at work today.
I decided I am not going to give up on the current RON network.
My mindset has shifted and I now have a better understanding for campaign optimization.
What do you say we have some story time to make this more fun to read?
For the sake of the story,
Let's pretend I am a fisherman and not an affiliate marketer.
Let's also pretend I am trying to catch fish and not obtain leads.
Let's pretend my bait is my angle and offer.
and finally that the ocean is my traffic source.
So my current mindset is that I am a fisherman, and I am looking for fishing spots that like my particular bait. Let's say I am using the "boost your battery life" bait.
So I find all the spots in the ocean where I catch some fish with my bait. Then I keep track of all the particular spots where I caught some fish.
Then I come back into bay and I evaluate my results. Now I caught some fish but it wasn't really worth my time because the gas in the boat cost more than what I could sell the fish for. But, not all hope is lost. Maybe I can improve the bait I used in these places to see if I can catch more fish the next time I go out. I will use similar bait because I know it works. But what if instead of the "boost your battery life bait" I use the "Make your battery last all day bait".
So I go back out to those spots and throw out my new and hopefully improved bait with the hopes that it will catch me more fish and make it worth my trip out there.
I repeat this process until it is worth making the trip or until I realize the bait I am using just isn't tasty enough.
Wow! That was fun!
...time to go catch some fish.
^^ Haha I love it!
Starting over with a new offer in a new country. ( keeping this thread going though )
Getting the feeling that the current offer I was promoting had been saturated and lost steam and the payout was almost too low to get any decent volume IMO.
The new offer is in the same category though (utility/productivity app) Just has a higher payout but not too high.
Things that I learned from the first launch.
1. Tracking C variables with rotating lander (thanks Zeno)
2. Optimization ( Cut the shit, whether it be placements, carriers, etc) Whatever is eating up your budget and not producing results.( thanks Stackman and Angry Russian)
3. Figured out a more efficient way to test in terms of where to start.
* After your initial scatter-shot RON test. You should have a couple different placements that seems to like the offer you are promoting. Now this specific placement doesn't necessarily need to be profitable. It just needs to be a place with decent volume that has a decent amount of conversions for your offer.
Then you pull those two placements aside (separate campaign) and set a personal goal in terms of ROI. Maybe 50% to start.
Then you keep on optimizing those placements until you hit your ROI goal.
Once you hit your ROI goal on those placements It is time to scale to other placements (still on same traff src). Hopefully you are able to grab more placements on your traff src that convert the same or better.
Not sure what would come next as I haven't got there yet. Will let you guy's know when I do.
*Gone Fishing*
There is no losing in fishing.
You either catch fish or you learn.
^ Yessir
subscribed! just now seeing this thread 
Any advice on optimization for app installs?
What are some things I should be looking for when using RON to target? placement>carrier/handsets? carrier/handset>placement?
After looking over my data I have found that a specific carrier is performing quite well. So today I launched a new campaign targeting only that carrier with the top performing ads. I also created a separate RON campaign targeting that carrier to find additional placements.
Cull placements but look in the carrier/devices tabs for outliers. However, be smart about it and make sure that one placement isn't dominating the traffic and skewing the results.
Originally Posted by zeno
Had to look up the word cull. So cull placements means to select a lot of different placements (use RON or manually select them) then look for the best carriers and devices. While also keeping an eye out for placements that are getting a lot more traffic than others and try to balance it out so that the traffic is distributed more evenly throughout my selection.
Cull just means selectively kill off/remove. So, don't aim to remove lots at a time, just remove specific placements that are performing poorly. At the same time, if a specific carrier or device is performing poorly AND it's data (clicks, imps, etc) comes from many placements, then you can pause that device/carrier too.
In other words, if you look at the placements/apps/sites and 95% of the traffic is coming from a single placement, DON'T kill off devices and carriers until it has been removed from the equation.
Today's Stats:
These are the stats on my 3 different landing pages.
http://i.imgur.com/c4EHUwj.png
All very similar to one another.
Ignore the CPC as I messed that up and for some reason I can't get the tracker to update the cost.
The cpc is around .03
Overall Today I spent 100.08 and made 48.00.
Plan is to:
Cut placements that received at least 100 clicks and 0 conversions.
Continue to optimize my landers until I find one with a conversion level over 9000!!!
Wow that's a monster LP CTR - almost unbelievable in fact! Finding that balance of CTR and CVR will of course be key.
-50% ROI is good given those stats - should be able to optimise the campaign quite effectively.
Looking good - keep us updated!
I spent another $100 today testing out LP variations.
Yesterday my best LP cvr rate was 4.16%
Today my best LP has a cvr rate of 6.64%
This LP was the One hour translation of yesterdays LP that was google translated.
Here are stats from today, again please ignore the CPC as it is messed up.
http://i.imgur.com/KnLCbsE.png
Overall I spent $100.15 and made $61.92
So -39% ROI
But thats an 11% jump from yesterday : )
The best part is I haven't cut any placements yet and I am running RON.
My next steps:
Cut any placements that received 100 clicks or more and have a cvr rate of 1% or less.
Try out more variations of my winning LP.
Staying focused.
P.S.
Heres a sweet mix I found. No vocals so you can keep concentration.
https://www.youtube.com/watch?v=FGxeTqPssqg
Originally Posted by fatmach
A rule of thumb recommended in this forum is to cut placements after spending 3X payout and no conversions.
I would recommend that you check my post in here for some ideas
http://stmforum.com/forum/showthread...bile-apps-data
Keep going! I'm sure you'll hit profit soon!
Monster LP CTRs! Got any tips?
Yep, definitely agree with Kamaleon here - cut on spend, not on clicks. You may have some placements that are spending tons but generating bugger-all clicks.
Looks like things are definitely going in the right direction! Keep On Split-Testing!
Great follow along, after some optimization you will be in the green.
Also the order you are optimizing is right. Make sure you get the best offer / lander / banner before cutting carriers / os / single placements.
I had some improvement today.
Decreased my bids a bit not much. Still got enough traffic to spend another $100.
My lander from yesterday had a major cvr drop almost 3%.
But a variation I did of that lander performed about 2% higher.
Here are todays stats. Please ignore cpc as it is still messy.
http://i.imgur.com/oisyWun.png
Overall today I spent $100.13 and made $72.00.
-28% ROI
But a solid 13% jump from yesterday.
I cut placements that have spent 3 times offer payout with no conversions. It wasn't many.
Next steps:
Make variations of my lander and spend another $100.
Another chill song I found today if you like music.
https://www.youtube.com/watch?v=kkLTxsAhURw
One important thing I am not seeing in your data is that you should aim for a banner ctr of at least 1%.
no update, maybe he hit already hehe
Home > Mobile > Follow-along Campaigns