OK lets have a bash at mobile then.. (33)
So my sitescout follow along lasted about 2 days. I have a background in $60+ payout health CPA (SEO) and was trying to do that on sitescout. Problem was my testing budget and lack of experience with the very tricky traffic source. So that failed miserably......
SO I am taking all the advice on that thread and trying out something completely different. ive been reading loads on mobile over the last few days. Angry Russians PDF was very helpful. Also read some of the Mobile Manifesto from Ruck IM Grind (is he Chet from Weird Science?), plus tonnes of threads on here. Anyway mobile looks a much better traffic source to get into (particularly considering my target GEO) and really excited to start.
SO heres my first campaign:
Apus Launcher App (Direct Linked)
Payout $0.22
GEO: Thailand
Traffic Source: Decisive
Banners: Photoshop and BannerSnack
I am targeting Thailand as that's where I live and I figure its pretty uncompetitive. My Thai is pretty good and I have a good understanding of local culture and slang which should be an advantage over people using Google Translate. Nearly all Thai traffic is mobile hence wanting to try mobile traffic out.
I was going to use a mobogenie type offer but it was a direct APK download which decisive doesn't allow. I dont want to make things harder then they need to be at first so I have shelved that idea for later. Shame as I made a load of banners for it but never mind Ill come back to it maybe.
So I came up with a few angles (some animated some not) and set up my campaigns according to Caurmen's awesome mobile cookbook.
One thing I found straight away with Banner snack is that its not compatible with Thai font. This turned out to be a good thing as my alternative saved alot of time. Banner snack is great but its pretty slow to save and download gifs. The way I got around the thai font issue is to make templates in bannersnack with no text and then add that in later with photoshop. As I am using the same gif template over again but changing the text this worked out much faster anyway.
Anyway the campaign is all set up and just waiting for approval. USA and Thai time are completely different which might work out ok as when I wake up the campaigns should be approved. Then I can unpause them and see what is what.
Good start bud.
Don't expect Thailand to be a uncompetitive - it certainly has it's fair share of offers and affiliates running them.
It's kind of like Brazil... It was uncompetitive until everyone said that, and jumped on board.
I recently ran a mobile campaign on Facebook and was shocked when NF clicks in Thailand were costing me $0.05-$0.10 straight out of the gate. Yikes!
Also word of caution, ruck's mobile manifesto is good but its ultra biased towards their own products (revived media, imobitrax, ex imgrind forum), it doesn't tell you - you need to split multiple cpa networks with the same offer to determine which converts the best in the geo.
I would HIGHLY recommend trying Turkey, Brazil & Pakistan too 
.
Nice one thanks a lot for all the encouragement/advice.
20 hours later and campaigns still not approved on Decisive...... hmm
OK so my campaigns were approved and started getting traffic straight away.
I have let them run all last night and today.
Some negatives and positives.
Positives:
My CTR overall seems pretty decent. In the 4 campaigns I launched, overall CTR ranges from 3.8% (WiFi APP) to 0.6% (Mobile Site) The CTR for all my campaigns averages at 2.5% (I think - see below).
Wifi CPC was below $0.01 and mobile CPC was $0.02 - $0.04.
This offer is only for Android 4.0.3 and I had been targeting ALL Android devices, so I will call my stupidity a positive as I was wasting quite alot of traffic so the stats should possibly be better.
Negatives:
Only 5 Conversions at $0.22 with a total spend of $26. That's something like a -99% ROI methinks.
Decisive is a nightmare! The Ads overview shows 3 conversions, 'reports' shows 1 conversion, live reports (if they actually load) shows 0 conversions and Neverblue is showing 5 conversions. I manually triggered the pixels in case they hadn't fired and they said they were duplicates (eg already in the system) so it must be that decisive is slow to catch up. Seems weird though that different parts of the same system show different data.... WTF is that all about. Not only that but the whole system is slooow. Feel like I am on 56k dial up. I want to punch someone. The live reports dont load at all and everything else takes forever.....
So anyway will let these ads run a while longer and hope I can get the proper data from Decisive.
I am wondering if maybe a landing page would be a good idea, the ads are in Thai but it looks like the play store page is not translated so that could be what is causing the miniscule conversion rate: 4500 clicks to 5 conversions = 0.01%!!!!!!!!!!! My banners are not random click bait images, they actually do sell the benefit of the app. So I am thinking a nice lander with bullet points as the the benefits could work wonders. If I also cut out all the traffic from non eligible OS plus probably cut out mobile traffic then it could improve alot.
Also thinking whether I should launch this in another cheap but English speaking country such as Phillipines or Nigeria as I can translate the banners myself and they would not require a LP.
Or maybe I should clone the campaign and try a different phone optimisation app?
All in all a good first day. Happy to have had some conversions and a decent CTR.
I'm hooked.
So I decided to duplicate the campaign over to India. My AM says this is one of the top GEOS for this offer.
Campaign is approved and live on decisive now....
Spent about $12 on Indian campaign with much lower CTR of between 0.43% and 0.8%. Zero conversions from over 500 clicks.
So paused all my Indian campaigns too (paused Thai earlier today).
There is something seriously up here. Getting plenty of clicks but a painfully low conversion rate. My AM says this should convert at 2 - 3%.
I had a mess around with the ROI tool and with a $0.22 payout and $0.01 clicks I still need a 4.5% conversion rate just to break even.
Tempted to try a landing page but can that really change conversion rate from less than 0.1% to 4.5%? And that would just be to break even!
Maybe my angle is shit? I am using the 'speed up your phone' angle. It accurately portrays what the app is all about plus it is getting clicks. But not sure why no one is installing. Perhaps I will brainstorm for a new angle. I have messaged my AM's on networks with plenty of mobile offers asking for what is hot at the moment.
Decisive is still showing just 3 conversions whereas Neverblue shows 6. Thought it would have caught up by now.
OK so going to brainstorm for new angles or failing that try a new offer.....
As my ads were getting a good CTR in Thailand and I was getting clicks at less than $0.01, I thought I may as well tweak the ads and try them for a new offer.
SO... just set up a new campaign for Baidu Browser using the same ads but changed slightly.
Furthermore, I don't make a distiction between APP and SITE traffic. I both see them as placements. If a placement is working I keep it, if it is not than I ditch it.
In general this campaign doesn't look very promising considering the total spent and total revenue
You might have trouble scaling if you are just focusing on the mobile app campaign - you are winning 30% of the bids. So, there is ~$15/day of traffic there with your current targeting.
I would try new angles or offers first -- trying to optimise that very low volume campaign from $1.20 CPA to <$0.20 CPA is not worth the time.
Yep you are probably right, in fact I was winning alot of the traffic from all sources.
So now I am confused. There is supposed to be 28M impressions per day in THailand available for bidding. I was only bidding on around 500k when I was bidding on all 4 campaigns (wifi app wifi site mobile app mobile site).
O I don't really understand why I was not bidding on all the available impressions for my campaign. I was bidding on Android 4.0.3 and above size 320 x 50. According to the planner this should be at least 15m impressions. I dont understand why I was only bidding on around 500k.
I think this is a basic thing I need to understand in order to choose a decent campaign.
Oh actually, I think Decisive throttles your impression auction attempts based on budget - I forgot about this.
There may indeed be a lot more available.
We should really confirm this with Decisive - on it!
If you're not seeing many impressions on Decisive, the first thing to check is your daily budget. The second thing to check is your bid.
If you push both budget and bid sky-high and you're still not seeing volume (keep a CLOSE eye on your live stats whilst you're doing this, obviously), get in touch with Decisive support. Sometimes their system is a bit too clever for its own good and needs a kick to start delivering impressions appropriately.
Also, be aware that not all their advertisers accept all verticals. In particular, they just added a bunch of new traffic sources that don't accept antivirus, sweeps or "unbranded dating" (not 100% sure what that is ). Again, worth checking with support to see if that's the reason you're being throttled.
Right so I have been trying to get the next stage of this set up over the last couple of days. Not easy when its school holidays and I have a 3 1/2 year old driving me up the wall all day.
Anyway I managed to get all set up on CPVLab (I think) was a bit of a mission as I needed to upgrade to new version plus I had lost my log in but I think I got there in the end. I hope.
The landing page issue was causing me ballache as I am a lazy wordpress type affiliate who can just about do an h1 tag if I need to. I ripped and edited a few LPs but its too time consuming and gives me a headache. So I looked through some of the shit ive bought over the years to see if I had something suitable and found Funnelkit. Its designed for desktop LPS mainly but I found that I can make pretty okish mobile LP's on it. I checked the page size loading speed with pingdom and its 188kb and was 'in top 97% of websites' which seemed ok. My server is in USA but I am using cloudflare on the domain so I think should be ok for Thailand. ((( ####EDIT Hang on a minute after I typed that I rechecked and according to pingdom its gone slooooow and loaded in 6 seconds WTF#####))) ((EDIT 2: checked it again and its loading in 564ms now. phew))
Anyway I have produced a lander which ties in with the angle I already tried previously and with the new angle I have been working on.
So in the next round I am trying all the banners from the first Baidu Browser campaign plus the 7 new ones from the new angle. They are all going to be going to the same LP with no DL.
Submitted 4 campaigns : Wifi app, Wifi Site etc and am waiting for approval.
189kb is way too heavy for mobile. My LPs are 10% of that
3 tips
1. Thailand is very compeititve
2. Animated banners!
3. Try a LOT of angles, the Apus app can be seen as a wide variety of things with the right marketing
Originally Posted by keepthecar
Originally Posted by stackman
Yep using animation and getting ok CTR just not converting well at all.
I ditched Apus for the moment and am focusing on Baidu Browser.
OK I added a new LP and got the filesize down to 50kb.
Campaign is live now. Lets see what happens.
dont focus on optimizing banners at first, optimize placements and devices. Get your conversions tracking prpoerly to decisive and you'll be able to see which banners are total shit much more easily.
Can you show us how you've formatted your offer URL? That'll help track down the problem.
In fact, the more info you can give the better. Obviously blur out your domain, though!
Yep here is my campaign URL:
http://xxxxxxx.com/base.php?c=13&key=eced656cfc1f3e8d9216d8ca090f9d99 &bidhash={{{bidhash}}}&creid={{{creative_id}}}&osv ={{{os_version}}}&dos={{{os}}}&dev={{{device}}}&pl a={{{placement}}}&site={{{site}}}&app={{{app}}}&ca rrier={{{carrier}}}&aid={{{ad_id}}}
Here is my offer URL:
http://hwoxt.com/?a=254237&c=1275747&m=31&s1=nerd&s2=
Had to pause the campaign. Only one conversion and that was from Talking Tom App so must have been an accidental install.......
Looks like my LP's were doing more harm then good.
That setup looks entirely fine - not quite sure why postbacks aren't working.
In your click report network-side can you see unique click IDs in the s2 column?
Originally Posted by zeno
Going to take Stackman's advice and ditch Thailand.
I am going to sniff out a more uncompetitive GEO.
What were the stats on the campaign other than 1 conversion? Spend, CTR on the lander, clicks, etc?
My initial suspicion is that your lander just had too much text on it, but I could be wrong!
Right so I am now running just one placement with 3 OS and one Carrier. Luckily the OS and Carrier have a large share of the volume in the GEO.
Using the planner recon there is about 400k impressions available for this combination.
The 'good' news is that I have managed to get to break even point. So far today my CPA is $0.36 which is exactly my payout. Feel like its a milestone as I have never managed that on a paid campaign. The problem is I am not sure where I can optimise further. There are a few times showing in day parting which look promising but if I do that my volume is gonna be tiny.
I think what I need to do is what I said I was going to do but didn't, which is to test some new creatives. If I can get a better converting lander or banners with a higher CTR I can probably start adding back in some of the things I have cut. I think I would rather do this than keep cutting down my campaign until there is no volume left.
Todays Stats
Yesterday before I cut down to just one carrier I was -18% ROI.
I have a decent amount of data now for this offer in this GEO as I have spent around $150 on it at a payout of $0.36. I think I have a fair chance of getting this profitable and keeping a worthwhile amount of volume.
One thing I have noticed is how careful you have to be when cutting stuff from campaigns. For example one of my banners was the best performing one day with 15 conversions and a CPA of $0.30 and on another day it was the worst performing with one conversion and a CPA of $1.50. So even though I have a low payout offer and quite alot of data I am a bit worried that I may have cut stuff too early. Tell you what I am glad I took the advice at the beginning of this follow along and went with a low payout offer - my original plan was a $60 payout offer!
There's no easy way to get your actual spend into
yesterday ROI 0%
Today -80%
Exact same campaign except for I added some new creatives. The problem isnt the creatives as the new ones are perfoming as well (as badly) as the the old ones.
WTF
Just when you think you are getting somewhere
The one hope I have is that the offer converts on app install and opening sometimes I seem to get a flurry of conversions, so I think maybe people open the app a long time after installing it.
**Update** Its a few hours later and still terrible conversion rate. CTR is same as yesterday but conversion rate has gone from over 1.5% to 0.6%.
No idea what to do next. Keeping the campaign running for a little while to see what happens.
How much data do you have i.e. clicks/conversions between yesterday and today?
Fluctuations are normal especially if you don't have much data.
And of course day-to-day variation is normal.
Yesterday was a $10 spend (1200 clicks) and today $17 (2200 clicks) so far.
The campaigns were run at completely different times of day so maybe day parting is more important than I thought.
I will let it run a while longer and see what happens.
Originally Posted by weebay
As an example, comparing $10 spend between 10 AM - 2 PM on a Monday and $17 spend between 4 PM and 9 PM on a Tuesday is just nonsensical.
If your test isn't sound, the data comparisons won't be either.
So Turkey seemed much better than TH but the only way I could probably get that profitable/break even with the current creatives is to do day parting and that will kill my volume.
I like this offer so instead of changing offers I am going to change GEO. So I had a look for a new GEO and found what looks like a great one with cheap CPM and lots of the type of traffic that converted in TH and TU. So getting my stuff translated and going to try that out now.
Home > Mobile > Follow-along Campaigns