My first campaign for mobile appetizer!! (20)
Hi all its been a week since I joined STM and have been reading up a couple of stuffs such as the mobile appetizer guide, mobile tracking and many others. I have decided to took a pause from passive activities and get some work going . Hence, here I am with my FIRST campaign!
Offer: DUSPEED Booster
Geo: Brazil ( Portuguese as chosen language)
Payout:$0.78
I am currently using DECISIVE for my traffic source, have only a network affiliate at the moment (Appflood). Currently still figuring out how to set-up
-I have came up with 4 different ANGLES with 3-4 different banners style for split testing.
-I am using PIXLR for my banners and a free gif making site http://gifmaker.me/
(going to learn photoshop in the meantime)
-I do not have a landing page currently.
-I am using google translator for my translations. I did read about google translator not being accurate , but i have split test those with some other online translators and they seem to give me the exact result. Do most of you still using google translator? Or do u recommend other paid methods such as fivver or one hour translation? I came across a case study http://stmforum.com/forum/showthread...gle+translator
regarding google translator and OHT, although it was a small sample but at least it gave me some assurance that google translator can still be used.
(WIFI-APP)
1)HELP! MY ANDROID IS TOO SLOW! INSTALLING… CLICK HERE TO FIX!!
Socorro! MEUS CARGAS ANDRÓIDE muito lento! INSTALAÇÃO ... Clique aqui para corrigir!
(MOBILE-APP)
2)MY ANDROID LOOKS DEAD... I THINK? REVIVE IT NOW... CLICK HERE!!!
MEU ANDROID ESTÁ INOPERANTE ... Eu acho que ? reanimá-lo agora ... Clique aqui!!!
(WIFI-SITE)
3)YOUR ANDROID IS FAST? SO WHAT! MAKE IT FASTER!!! INSTALL THIS!
Seu Android é rápido ? Então o que ! Faz mais rápido ele !!! INSTALAR ESTE!
(MOBILE-SITE)
4)BOOST YOUR ANDROID NOW! INSTALL NOW!! DELETE TOMORROW…101% FREE! 160% FASTER
Impulsionar o seu ANDROID AGORA !!! INSTALAR AGORA !!! APAGAR amanhã… 101% GRÁTIS ! 160 % mais rápido
(WIFI-APP)
http://img.crwd.io/8tDqUimHBdP34uEzxDEEhIYya7h.gif
http://img.crwd.io/rQ4UZ7e5K3qQapW8Ll1tK8s3frv.gif
http://img.crwd.io/xsFa1MVOjfsKU0jbkFirQWS0ohy.gif
http://img.crwd.io/93ha58jSdqovK3PPTy2BTLkaCrr.gif - got rejected due to misleading content. Had to take away the "!" picture on the right on the first slide.
(MOBILE-APP)
http://img.crwd.io/afjzc75hi8AiY4R8i92mSoBI7F2.gif
http://img.crwd.io/Aelen4tGGKrkqurQbuxefC7zZiZ.gif
http://img.crwd.io/iZpug3NbzHB3vuUgEQyFdCMGwPa.gif
http://img.crwd.io/p9Q95n3Y8OlVHYoYo6oauxORWJh.gif
(WIFI-SITE)
http://img.crwd.io/vD11l4YSeF6DxeGkCD9IRhUHKnQ.gif
http://img.crwd.io/b2NTUyJB49WIofACeBVnFfHj51b.gif
http://img.crwd.io/ocLwMDg507Rnm4szjWhAnqZL9XD.gif
http://img.crwd.io/556yHAHaXJaKZZmiITi2rP7cySZ.gif
(MOBILE-SITE)
http://img.crwd.io/rwewSjaHE61Aw4NdujNKkYlHPmN.gif
http://img.crwd.io/uHM2XDOc6Xjzb4ApnBjY0QarChH.gif
http://img.crwd.io/5ektsGOJ5UPsFekhjfps6GA0Sf4.gif
http://img.crwd.io/V2UtnlRJ1C6Mfq7bdkPnVa9bB5.gif
Alright, so i have ran my campaign for roughly a day (18-24hrs):
Overall view: 
WIFI-APP
MOBILE-APP
WIFI-SITE
MOBILE-SITE
Have about 1223 clicks in total, 1 conversion(not shown on Decisive, it is shown on my aff network instead), spent about $21.78. Am currently setting my campaigns to paused to see if i need to optimize any area, also wanted to double check if i am on the right track. I do know that we should allow each of our campaign to reach at least $15. Are these for data significance purposes?
Some QUESTIONS to be asked right at this point :
1) How do i track conversions from my Aff network( Appflood) to Decisive? I have read many guides and forums threads around but still unable to figure out . I tried using several links 
. My aff network use other form of variables such as aff_sub for s1 and aff_sub2 for s2 respectively and so on, 
. Are my links correct? If not , how should it be?
This would be a important step because i cannot track other data such as CPA. So much help is needed here!
2)Should i blacklist all Carriers accross my campaigns if they spent 3x-7x times my offer payout? Or should i blacklist placements such as apps and sites(3x offer payout) before considering "carriers".
I remember reading a follow-along thread where caurmen suggests that carrier should be given more time(maybe 7x) before considering culling those. He also suggest blacklisting only placements and not handsets or OS version according to the 3x payout rule.
This post i guess: http://stmforum.com/forum/showthread...l=1#post168758
3)Same goes for creatives, should i blacklist them after i have blacklist respective placements?
4)Are the over stats acceptable? CTR%, CPC, CPM, SPEND. Will it be considered as campaigns worth optimizing? Or should be classified as campaigns that should be cut? Or give it some time to run after some optimizing till it hits $15?
5)Are the banners style and Angles alright?
I feel that my CTA appeal on my banners seems low and could have affected the conversion rate. Could i be wrong here? Should i include a landing page?
6) If any of your banners (listed under "Creatives" in Decisive) have a CPA greater than 1.5 times your offer payout over the duration of the last $15 spend, blacklist them.
Do they mean that if my payout is $0.78, and my cpa is >1.5X$0.78= $1.17, i blacklist them. Is my calculation correct?
7) So i chose smart CPM bidding for all my campaigns , my payout is $0.78, hence my CPA GOAL should be $0.78X1.5= $1.17. So i put a range between $1.17-$1.20( in my case $1.18). Are the calculations correct?
Please pardon me for all my questions , if they are stupid or not. Its my first launch, and i want to be sure that i am in the right track before i proceed. I rather be a fool now than a fool for the rest of my life. Any form of help and guidance would be much appreciated thanks!
Why are you using completely different creatives for each of your campaigns?
It makes no sense to tailor custom creatives for WiFi app, WiFi site, Carrier app and Carrier site if you have no data to guide you. This destroys your split-testing as you are enforcing a different set of experimental conditions for each campaign.
Now, there is no way to say if the banners themselves are bad or if it's the connectivity/placement, or a mix of the two. Basically, you are trying to compare green apples, red pears, blue bananas and yellow passionfruit.
1) Your postback URLs are not correct as they are not using the tokens specific to AppFlood, which uses the HasOffers platform. As you can see from that list, if you want to send back the value that you passed as aff_sub2 in your affiliate link, you need to use {aff_sub2} - not #aff_sub2# and certainly not {{{bidhash}}}. These mean nothing to that platform. Always use the tokens as explicitly stated by that platform, anything else will not work.
In this case, make sure you are passing values to the correct entities.
Are you using
If so you should configure your offer in Voluum to send ....&aff_sub2={clickid}
Then, your postback URL in Appflood should be blahblahvoluum/postback?cid={aff_sub2}&payout={payout}
If you are using a tracker you send conversion information to it, not to Decisive.
Then, in Voluum, your Decisive traffic source should be configured to use an externalid with the placeholder {{{bidhash}}}. The postback URL for Decisive should then be there as well, with {externalid} to pass back the bidhash ID.
Explictly: hxxp://win.crwd.io/convert/{externalid}
2) Leave the carriers alone for now. Placements are more important at first.
3) Yes, though hard to say at the moment with the varied creatives across each campaign. Could banner A in Campaign A do well in campaign C? Possibly. We may never know.
4) Hard to say right now, if you have an average CPC of $0.02 and the payout is 40x that then you have plenty of room to find working placements/banners. Just set ground rules like in the Cookbook and cut the fat then review if the offer performance campaign-wide.
5) Banner variation is fine although, as I've mentioned, need to conserve this to being the same in each campaign when starting. Your angles are all about the phone being slow so there could be more variation and creativity here.
6) Well yes, that calculation is correct, but I would blacklist placements first otherwise all the banners will be cut by this metric before you even get off the ground.
7) If you offer payout is $0.78 then to be profitable you need a CPA of <$0.78. You can however start with a higher CPA goal in Smart CPM so that it is more aggressive with bidding. It's an 'intelligent' bidding mode so don't worry too much about the CPA right now.
Originally Posted by zeno
thanks for the reply zeno. I screwed up from the start of my split test.Some other questions before i proceed :
1) Should i continue running the campaigns until they hit $15 each? Assuming i am blacklisting placements and some creatives along the way while allowing carriers to run. Or should i make another 4 campaigns with my current angles and the standard 3-4 banner variations across all 4 campaigns?
2)Should i include a landing page and see if i can improve the conversation rate?
1) Might as well. Use the data you have collected to start better with the next campaign - e.g. placements that sucked across the board.
2) Landing pages are always a sensible idea to test.
Originally Posted by zeno
2) If i follow appetizer, how big should my landing page be?
1) I mean placements that were a long way from profitability in all the campaigns. I have no rule here, it's intuition. E.g. if it spend $2.75, $1.38, $3.41, $0.97 in each campaign with no conversions anywhere, then I would probably kill it.
Well, if you have apps and sites then I guess placements will only exist in 2 out of 4 campaigns, but you occasionally see universal placements where they have a site and app.
If the BBC News app and site appears to perform like crap in both WiFi/Site and carrier/app campaigns, then it probably sucks for your campaign. A duck dressed any way you like is still a duck.
2) How big your LP is has nothing to do with the appetiser. Just make it as slim as possible. That's all. There is no rule here, just use common sense and good mobile design. Avoid external CSS and JS if possible, don't use many images. If you do, make sure to optimise them through Kraken. Etc.
Originally Posted by zeno
Decided to pause my campaigns after 2 days, here is the overall view of the data 
. Only 2 conversion for Mobile-app.
Have yet to setup
Things to improve on my next campaigns
- Include some LPS for test
-Setup
- DO NOT create totally different banner variations like what zeno said, Stick to 3-4 variations across the campaigns. Felt really stupid here, as my split test didn't really make sense. Thank you zeno for knocking some sense into me.
- Use more creative angles
1) I noticed a similarity in both the mobile-site and wifi-site: I blacklisted "tubidy.mobi (tubidy.mobi - Mobile Web - 320/50 - RTB)" in both of these campaigns because they don't seem to convert after eating over 3x-4x of my payout. They both have a decent CTR% tho i feel. Assuming i am going to test the same offer again , should i be blacklisting them at the start to prevent it from eating my budget without conversions? Or should i be testing with a few more campaigns with the same offers before i conclude?
2) Should i be excluding WIFI-APP and WIFI-SITE for my next test? They seem to have the most clicks without a conversions. 517 clicks and 796 clicks respectively. And they spent about 12x my payout without any conversions. Will it be necessary to run a bot-test at this point?
3) The CPM varies across 4 campaigns , some were as high as $0.46 and some as low as $0.17. Is this normal? Anything i can do to improve? For eg. change to manual cpm?
4) For making LP, zeno u mentioned avoid external CSS and JS if possible, may i know what it means?
1) I'd keep tubidy in on your second angle, but if it doesn't convert there either, blacklist it from future campaigns with the same offer type in the same country. Tubidy's an odd one - its traffic can be pretty dodgy, but in some countries it seems to convert pretty darn well.
2) No, don't exclude them yet - different angles will result in different reactions. However, doing a bot test on that traffic is probably a good idea.
3) Yep, that's normal. You could change to manual CPM but at this point the varying CPMs are actually the system attempting to optimise.
4) Don't link out to stylesheets or Javascript - instead, keep all your Javascript and CSS on the same page in <style> and <script> blocks. Does that make sense? Google "inline styles" to get more info.
1) Yep, try switching to manual CPM. These days, to be honest, I'm getting a bit disenchanted with Smart CPM - it seems to choke your traffic pretty often.
2) High spend is a good sign - it means there's tons of traffic there. Top priority here is to get your tracking sorted out - without being able to track conversions accurately across
Wifi USUALLY performs worse, but not always.
3) Not a great start, but it's early days yet! Have you been killing placements on this? I'd look to cull any placements that aren't working (anything that's spent more than $1.80) but otherwise keep testing.
Good work on testing multiple offers. That's the way to do it!
Ask Decisive support if your conversions aren't tracking properly. They can usually hit the back end with hammers and fix it!
Bidding higher: if you've killed a load of placements and traffic has tailed off, then yes, bid a bit higher and see what you get.
Looking at your stats for this campaign, I'd recommend killing it and testing another, unless you have one or two placements or banners that are showing significantly more promise. -80% ROI is difficult to recover from.
1) Yes, you'll probably have that problem, but worry about that later. Right now job #1 is getting to profitability.
2) I'd say yes, personally. But you might want to try flat CPM rather than smart CPM and set your bids pretty low.
3) Do you mean that you're testing several offers in the same campaign, or different campaigns?
4) Hard to tell - how many bids are you winning? If you're winning most of your bids, you could probably drop it down. It may be best to optimise bids on individual placements differently if you have a high-quality placement in there.
If you want to get some more suggestions, try sharing your banners and landers here - we can probably suggest things that may improve your campaign's performance.
First thing I'd suggest: try some much less clever headlines.
Your headlines are creative, funny, and entertaining, but most of them don't clearly communicate what it is that the ad's offering. This is a fairly common problem for smart, creative people when they start AM - it's easy to be too clever.
Try much simpler headlines, spelling out what the benefit you're offering is. For example, for Quikr, try "Post A Classified Ad For Free!" or "Buy Or Sell Anything!".
Something that shocked me when I started AM - I was trying very clever, elegant headlines for dating ads. Then one day I got frustrated and tried "Do You Want A Girlfriend?" as a headline.
It achieved a CTR of something like 3x what my other ads had been managing!
Simple sells.
How do you check the CTR in decisive?
Originally Posted by adrianegerrard
Running some other offers at the moment , and had some questions that pop in my head:
1) Had a dilemma about cpm and payout recently.
For offers which payout is as low as $0.15-$0.50, do we bid higher than the average cpm for the geo we target? Doesn't it defeat the purpose, since bidding higher would mean higher spent and its hard for low payout to be profitable this way? However, if we bid lower than average or our competitors, wont we get too little traffic also?
2)Is mobile spying tool really that necessary to get to $xxx/day or $xxxx/day? I've heard of WRW and Hide My ass VPN, But WRW is really too costly for newbies, and not affordable. I am really not a fan of jacking people's landing page or creatives just to get it to profitable. I want to learn what kind of banners and lps are profitable by doing them myself and grow it from there. Am i being naive here? Should i be more realistic and be spying more?
3)I currently have mobile-app campaigns winnings only about 6%-7% of the bids, how do i get more traffic without getting too much spent, since i am working or low payout offers. I am using smart cpm , set it at $0.40, offer payout is $0.25.
bring up my post. Need some advice here!
My camps are currently getting only 4-6% wins from the bids. Is that common among other affiliates? Its almost the end of day 2 and i have not even spent $15.
Once you got a good feeling for the smartCPA set the daily budget to $100 this way it won't limit your bids, but watch the campaign constantly when doing this.
Originally Posted by ckzhou
If the placements aren't profitable then there's not much point moving them into separate campaigns... you would normally do this to bid higher and take a bigger % of the traffic that works.
Be aware that bidding higher campaign-wide will give you access to more competitive placements so the campaign profitability might actually go up - you just need to test it.
2) Test more angles. Yours are very generic and have been done before. Think outside the box.
3) 4 is fine to start.
4) Yes, that's fine. If you have conversions then run it a bit more. If you get to $10 spend on a $0.01 payout offer with only two conversions though... so yeah, use some intuition here. If you have 1-5 conversions but this still puts you at -90% ROI then it's not looking great.
5) Yes and no. Focus on killing placements. But check the other data sets and kill things if they are doing terrible relative to everything else. E.g. 3 banners doing 2% CTR and 4-5 conversions each. One banner doing 0.5% CTR and no conversions. Obviously, kill that one.
6) When you need PHP code.
Home > Mobile > Follow-along Campaigns