[WAP] App install i want to learn how to test and optimize efficiently, and scale (8)
[WAP] App install i want to learn how to test and optimize efficiently, and scale
Launched yesterday android app install offer in INTL geo CPI is $0.35
So far i am runing Direct linked. CPM, with banners
SPEND: $99.06
REVENUE: $40.25
ROI: - 60%
115 leads
First i uploaded 3 banners which had pretty good CTR 1%-1.2% and 2-3% Conversion rate.
I was bidding around 0.20-0.23 CPM, and getting lots of traffic.
(i had around 40+ conversions) Then i decided to upload 5 more banners and lowered my bid to 0.17 CPM (MISTAKE)
Interesting thing happened and my CTR% and CR% went down.
I think because all my 8 banners were running at the same time inside one campaign they are maybe competing against each other ? (decsive ads)
I havent cut my targeting yet, i am runing on wifi,carriers,apps,sites
So now i am stuck and dont know what to do, not sure what to stop from runing.
What i did is stoped half of the banners so i dont buy worthless data for non performing banners, and upped my bid again to see if i can get my CR% and CTR% up.
How do i decide if a banner is a winner based on CTR / CR before i even have cutted any targeting ?
How many banners is good to have inside one campaign if they are all good and performing, maybe 2-3 ?
How do i go into the green while maintaining volume ?
What should i do first? Test some lps or cut my targeting and by targeting what first carriers,placements,os versions or devices?
If i start with devices after how many clicks/ money spend per device should i cut it?
For example i have tons of devices wtih 5-50clicks without a single conversion, but for me to start cutting them right now it doesnt seems like the data is statiscially signficant.Maybe i should wait until they spend x1 or x2 the offer payout and then i should cut them ? Maybe once i have a lander those non-profitable but still converting devices may become profitable ??
Maybe i need to buy even more data ?
Looking at the OS version data, can i cut OS versions ? 
Durr i am confused help me guys, i am clearly doing something wrong but not sure what 
First I would split this campaign up to target WiFi/Carrier traffic and App/Site targets separately (giving 4 campaigns overall). At the moment your data is very heterogeneous.
Run the banners that have survived this initial broad test.
Tnx Zeno, so i reviewed my Data.
And since most of the traffic available for the country that i am targeting in my traffic source is Wifi and App traffc. I created just one new campaign Targeting: Android > Wifi > Apps
This will help me with the confusion since i will have less variables to look at when optimising.
I also have choosen the top 3 performing banners bassed on CTR and CR using the http://www.peakconversion.com/2012/0...al-calculator/ calculator to make sure my descisions are statisticlly significant and will use only them for now.
They had 1% - 1.5% CTR and around 2.3% - 2.96% Conversion rate. (direct linked to the offer)
I also tested lander vs direct link on my previous camp. Lander performed slightly worse. I have more landers to test i will test them later.
So now with my new campaign, what should be my goal for optimising ?
Should i test more banners to find some with higher CTR / CR% or first take a look at my data and then decide, what i am gonna improve ?
I know that in order to get the most out of the traffic that is available i should optimize MACRO to MICRO, which means i have to improve the things in my campaign that will have the biggest effect.
Looking at my case with Android App install offer (google play), the variables that will make biggest impact are
1.BANNERS
2.LANDING PAGE
3.Then maybe targeting, but what exactly ?? Need some HELP here
I am thinking:What would be affecting my conversion rates for an app install offer on google play, what could stop the users from not converting, and installing the app besides banners and landing page.
My data shows that Wifi 2.5% CR performs slightly worse than Carrier traffic 2.6% CR.So thats not a big factor at the moment.I will run on wifi since it has more volume and once i optimize i will scale to carriers.
Placements,Os versions and Devices are the remaining as factors.
To determine and cut non converting devices i guess i need a lot more data than what i had: http://screencast.com/t/yf1hGiAffu http://screencast.com/t/4HU9UDrMpA
Here is my final data for OS versions too: http://screencast.com/t/3BX5EQI9s2m
And also some of my data on placments: http://i.imgur.com/qHLFaie.jpg
I have no idea when to cut Placements, OS version and Devices, any suggestions ? again my offer payout is $0.35
Ok thats it for now going to bed will see what data i have harvested for my new 3 best banners - wifi - apps only campagin.
Conclusions and mistakes that i did:
I am not sure if i am on the right track, but i think writing all these things helps me to make better descisions and understand the whole process better.
-I ran with 8 banners inside 1 campaign all at once at the start, which spreaded my traffic too much and confused me.Will run 2-3 at once in the future.
-Lowered my bids too much which gave me bad performance (lower CTR of ads and lower CR) Maybe its better to not play with the bids too much.In the future i am thinking about just duplicate the camp with higher/lower bids and see what happens.
-Dont run campaigns with too much variables in it minimize those, so i dont confuse myself with too many variables to look at.
+Did a good job matching the offer with the traffic source, there is volume and i think i can scale once i have good roi.
+Bought $130 worth of data, posted a follow allong and i am learning a lot from it
+I feel i have a little bit Better understanding on how to optimize now[/COLOR]
Looks good!
OS versions: have you checked the OS requirements for the app? It's definitely worth making sure you're not getting traffic from an OS version that can't install the app!
@caurmen Well they paused the offer.I found similar offers and will check their OS requirements, thanks. Once i get some data i will update my follow allong.
Little update for today: Atacked new country, using similar offer.
I used my wining banners from the previous offer that i tested which was paused on the first day.Translated the banners to the local language.
My banners were doing around 1% CTR. I guess while RUNING RON i cant get my CTR very high like 1.7-2% ? Do i need to make custom banners for each placement if i want to get super high CTRs ?
Got some data on the first day. Saw that most of my traffic is coming from one app and the traffic converts pretty good.
Site traffic converted pretty bad.
Analyzed my traffic and added more offers to the mix.Tested 3 lps.
Created 4 campaigns, as you guys recomended, thanks again.
1. WiFi and Apps only - $89 Spend $98 Revenue, 10% ROI
2. Direct app placement - WiFi only $115 Spend $292 Revenue, 154% ROI
3. Direct app placement - Carrier only $49 Spend $45 Revenue, -8% ROI
4. WiFi and Sites only - Nothing yet.
Still dont know what should be my priority when optimizing. So far it is Banners > Offers > Landers > WiFi/Carriers > Apps/Sites
My questions now. Runing 4 campaigns at once is overwheling because i am not very good at analyzing stuff, i am better at solving my problems creatively not with mathematically .
So far i havent cut much of the targeting just some VERY VERY obvious losers.
Should i pause all other ads and focus on my winning ad.Testing lots of Banners,Offers,LPs?Then after i cant squeeze any more ROI to start Scaling the same way to more apps and site placements directly with my wining LPs.?
How do i get most ROI on my traffic, and SCALE for $XXXX or more ? So far i am thinking to look at the placements which have the most traffic analyze the traffic cost and what the performance of my funnel should be in order to make it work.
Then if i think i can make it work i start testing.
How do you guys do it? Do you run many ads at once and testing banners and lps inside each ad all at once. Or once you find AD that shows promise i should pause all other ads if they dont have huge ROI like in my currenty situation, focus on my winning ad test the shit out of it with tons of landers and banners make the ROI insanaely high then, unpause the other ads and use the wining lp and banners for them and try to scale on other placments?
I think i answered most of my questions myself while writing them, but still i will really appreciate some feedback on my way of thinking about this.Thank you.
PS: I am not shaving my beard or cuting my hair until i hit $XXX,XXX Profit, So follow me allong 
Focus on what is working. Is there considerable volume available for your Wifi/App and Direct app placement/Wifi campaigns? If so I would pause the others or leave them on low budget and focus on tweaking these winners. Split test banners, LPs, offers and so on and find what works well.
Take the results over to other campaigns - basically let the winning campaigns do the testing legwork for you.
Focus most of your budget and time on these winning campaigns. In the green = happy, getting data, building leverage through volume that could later bring less promising campaigns into the green.
Spend your remaining resources on testing the other placements/targets that can offer reasonable volume (little point spending time on $5/day campaigns if your beasts are doing $500 a day)
Home > Mobile > Follow-along Campaigns