Low-Budget on Mobile - Small Steps to Freedom (14)
Hey guys,
The time has come to introduce ourselves. We are 3 good friends; Bob, Erik and Ruben from Amsterdam. We have been working together as a team for the last couple of years now, in our free time. Thanks to a tip from a friend (who’s also here and some of you might know) we found this place and everything changed from that moment on.
Background
Our journey started 3 years ago when one of us had some crazy idea to build an online clothing generator. Because each of us had a different background (IT, Online Marketing, Development) and skillset (ofcours somewhat overlapping) it was a good match. We spent almost 2 years building the product including a fully operational affiliate datafeed back-end to power the shops front-end. There’s much more to this story but it was getting way to big for us to finish in time and with just the 3 of us. Decided to move on.
Our second project we recently quit 3 months ago (for about the same reasons), was an online app for influential bloggers (high traffic) enabling them to create custom lookbooks or whatever type of widgets from affiliate product links. We build an automated scraper which was capable of getting all product images with all related information from just any website. It is as simple as filling in an affiliate product link, picking a template, choosing the best pictures and make something beautiful to show your visitors while getting paid for every sale. After about a year of development it was beta time. We really underestimated the time that it took to market this product. It could be a real profitable business if we had found the right partners. Sometimes u have to give up something in order to advance and find success elsewhere. In the end it’s the result that matters, not what you have done.
Bothe products still don’t exist on the market today in such a fashion and are in what we call the ‘freezer’ atm. If ur interested, let me know. I tried to keep it short but we have so much happened but also learned a whole lot about what it takes to be successful. Now back to our current goal, running profitable mobile campaigns to give us the freedom everyone is looking for.
Preparations
- We spent the last 3 months reading and digesting every mobile guide, follow-along and loads of other threads concerning mobile and affiliate marketing. Talking to people etc.
- We spent an entire weekend spying on WRW trial (and copying GB’s over to Mega)
- We split tasks and responsibilities and put them in an comprehensive planning with weekly deliverables.
- We applied for about 10+ traffic sources and 10+ networks and got approved.
- We started researching our vertical, offer, geo and eventually ended up going with 360.
- We started making loads of ad-copies for all the angles we came up with including a buffer for fast banner swap-in so we save time during the week.
- Erik created a massive campaign forecasting which we will improve/optimize with raw data from the campaigns.
- Ruben got crazy with designing loads of banners for the ad-copies, had to redo a lot because of not getting approved with icons (especially them sweepers/brooms), ad-copies or designs. (yeah we already knew it was strict these days but everything varies per network even with the same advertiser, damned AM’s 
)
- Setting up tracking on
- We had endless discussions about just anything, from angle and banner creation to campaign management and so on. The more you read, the more there is to talk about, which helped in the end but slowed us down at the beginning.
Some Random Facts
- Bob will be updating the follow-alongs.
- We have alot of experience with affiliate marketing on a very technical level, now starting with mobile.
- Ruben has a beard, we call him “The Epic Beard Man” or simply “The Guru”. He is always cookin up something. Our alias here refers to him.
- Erik has been stalking us with mails/calls of all sorts to keep us sharp and moving.
- Bob and Erik will be coming to STM London for the full 3 days and would like to do a bunch of meetings/networking with you guys.
Traffic source: Decisive
Vertical: Utility Apps
Offer(s): 360 Security
Payout: $0.78
Geos: CA
Networks: Appflood & Clickdealer (offer got paused on Avazu yesterday)
Budget: $3k
Bid: CPA bid based on average CPM. $15 daily spend per campaign, going to up this to $90 total daily spend for the next 2 days.
Angles: Categorized and sub-categorized all angles into the 3 main features of this app; Security, Boosting and Cleaning. Made lots of different creative ad-copies for each one. Couple of more categories are being filled with new angles to test coming days. We know angles make all the difference so we’re testing everything within the limits of our budget.
Campaign structure: Starting out with 12 campaigns (WIFI/3G for the 3 main features) with 12 banners per campaign (6 different ad-copies in 2 designs).
I’ll be posting updates and more detailed information soon, we just hit the start button on our campaigns yesterday.
Hope you enjoyed this little intro and we are looking forward to learn everything about mobile and running profitable campaigns in the future.
Some interesting projects you guys have worked on there!
I'll be looking forward to seeing how you get on in the AM world. If you've got any questions don't hesitate to ask!
First update (2Feb)
Spend for the day: $2.91
Bids: 321,937
Wins: 26,050
Clicks: 212
CPM: $0.11
CTR: 0.814%
CPC: $0.01
Second update (3Feb)
Spend for the day: $10.91
Bids: 663,485
Wins: 63,358
Clicks: 353
Conversions: 1
CPM: $0.17
CTR: 0.557%
CR: 0.283%
CPC: $0.03
We got out first conversion so that's pretty early in the process. ROI is of course absolutly red for now 
I'll try and give a brief overview of how everything is progressing.
What we did:
- We learned that CPA is better than CPM (avoiding the bid war) for low offer payouts. So we set CPA 1.5x payout but data is coming in slowly. Quite some clicks for awsomely low costs but we are winning not enough bids yet (about 20% now) to go up to 5k per banner faster.
- Killed 1 placement
So not much yet besides playing with the bids so we can hopefully start cutting banners and more placements etc soon.
What we did today:
- Just upped the CPA bid to 2.3x payout.
- Killed another placement
- Added same offer from F5 in the mix (so currently split-testing the offer between Appflood/Clickdealer/F5)
Another conversion fresh off the bat (4Feb), will post update tomorrow 
Some issues we are facing:
A) About 30% traffic from Ashburn takes up our total traffic, steady for every hour so far. This is being redirected now to another offer to have some clear stats but it still shows up in general stats.
B) About 20% traffic seems to be iOS, which is weird. Targetting for every campaign is set correctly in Decisive with the supported Android versions.
C) We are not sure about the correct bid, tried a creative CPM bid just above country average but that didn't change much. Maybe we can get some feedback on this.
Nice work, and a great initial write-up!
On your issues: definitely get in touch with Decisive about the iOS traffic issue. They're pretty responsive, and that's the sort of thing that they should really sort out for you.
Bids-wise, it sounds like you're on the right track. You could potentially push your CPA up a bit whilst you're in the testing stages - in the long run you'll almost certainly switch to flat CPM for some bids anyway.
Looking forward to your next update!
Congrats on the 1st conversion!
When just running RON I would go smart CPA on Decisive. the manual CPM on Decisive only worked for me when targeting one or a few placements. That's because the smart CPM can adjust the bids - for example one placement will be way above your manual fixed CPM bid - but it converts so well that it's still profitable. Without smart CPA you'll never discover it / you would need to bid super high fixed CPM.
The smart CPA on Decisive is not very sensitive so you can experiment with higher smart CPA bids as well - just get a feeling for it, maybe go x2 or x3 payout and watch what happens / switch back if it spends too fast. What worked for me is just pushing it up step by step until I got the amount of traffic I wanted. Then isolate profitable placements into a seperate campaign where you can bid manual CPM based on the data you got through the initial campaign.
Idk what Ashburn is - but if it's some wrong GEO or something similar - what often worked for me is doing a drilldown report in
You've got a massive advantage by working as a team! Keep going and you'll hit success. See you in London!
@Caurmen thanks for the support!
@Kepe95 thanks for your eleborate post on CPA/CPM! Hit me with ur Skype and we can setup a meeting
Unfortunately we had a setback, 360 got paused on Clickdealer (uncompliant promotion), Appflood and Avazu earlier. We got the advice of pausing our campaigns because it's appearent that the offer is being stopped. Obviously we want to split-test between different networks so it's not very useful to continu promoting this offer.
Now checking for good CPI offers with AM's and perhaps moving to our home country (NL). Which means we only have to translate from english and replace the ad-copy on our existing banners.
Added: got feedback from Decisive about the traffic from Ashburn, they couldn't see it in their stats because it's only visible in
Next update will follow soon.
Well guys, we went all out last couple of days and picked another offer (ya'll know this one) and moved to our home country. This previous 360 story isn't going to let us down anyway, we made it to 5k banner views and couple conversions. So some data to analyze and moving forward. Luckily we can recycle almost all copies and banner desings.
What we did:
- Translated every ad-copy to Dutch from our previous 360 campaign (got pulled from the networks but is back now)
- Created new ad-copy to fill-up our buffer (basically same amount of copy) for fast banner creation
- Used the same 3 categories (main-features) for categorizing all angles
- Improved one ad design and looks even more killer as before
- Pumped out 96 different banners (2 ad designs) in one day using Photoshop handles (Ruben cooked this up)
- We decided to create seperate campaings for both ad designs so it's easier to drill down in
- Means creating 32 campaigns on Decisive (hail the clone function, although it still caused frozen hands syndrome with Bob)
- Had to replace ALL banners in every campaign with new icon colour (yellow couldn't be used, meh), everything compliant now
- Improving team workflow and trying to systemize everything we do (key to long-term success)
- Currently working/testing on a sweet bot detection tool (again Ruben cooked this up) but there's another thread on that
Traffic source: Decisive
Vertical: Utility Apps
Offer(s): DU speed booster (Global)
Payout: $0.23-0.25
CPA Bid: Couple times payout (need data much faster)
Geos: NL
Networks: Clickdealer, Appflood & Mundo (great AM). Avazu got paused as of yet.
Feels good to be pushing traffic again and looking forward to the next rounds of spending, so we can start cutting ads and placements really soon.
Still need to go through the whole campaign management process from start to end. Will update todays stats in tomorrow update.
Couple of issues/questions:
A) Still receiving alot of iPhone visits, only shows up in
B) About 10-15% of traffic is tablet, our ads are 320x50 only. Wouldn't it be better to just disable tablet? I can't imagine people converting with such small ads. Any advice on this?
Stats so far
Decisive reports not working well atm but data below is pretty much the result of 2 days.
Bids: 1004783
Wins: 293081
Win%: 29%
Clicks: 2524
Conversions: 2 (+1)
CPM: 0.11
CTR: 0.9%
CR: 0.08%
CPC: 0.013
CPA: 16.28
Spend: 32.56
We basically played around with the CPA bids and let the traffic come in, but it is still going way too slow. CTR is much higher than expected and 3 conversions isn't that bad but we could use some feedback!
Questions we have:
1) Is it normal to have a 0.08% conversion rate when you run the campaign over the network? If not, any practical tips to increase conversion rate when direct linking? (we add LP's at -50% ROI, or should we do this earlier?).
Our 2 ad designs below (click to enlarge)
2) Is 11 conversions out of 10.000 clicks enough data to choose a winning angle out of 3? (Cleaning/Boosting/Battery as main cat. with each 12 ad-copies in 2 ad designs). At the moment this is our target for the first run to get data. Should we edit this target?
3) Should we increase CPA bid? Our avg. CPC is 0.016 and we win around 40% of the bids but we don't get data fast enough (maybe it will take 5 days!). Target CPC for 10.000 clicks is 0.018.
4) How important is it for a banner to explicitly state that it's about an app install? (we use lots of different CTA's)
5) Does it help to focus on filtering out bot traffic now immediately, or is it statistically insignificant enough to delay until a later stage of the campaign?
6) Maybe someone can answer point B) in the previous post.
What are your thoughts?
* Bump *
- Those offers have been hit HARD on the DSPs. If you want to succeed you need a fresh approach/angle
- Improve your banners. Try some plain text banners and look at some Admob banner layouts. They are super simple and a great starting point.
- Direct linking utility apps can work but landers tend to help a LOT. The problem is that networks caught on and are now very strict about landers.
- You are winning 40% of your bids. That's quite high so you could bid even lower (initially)
- Forget about ''bot traffic'' - yes there's a lot of that going on but focus on weeding out non performing placements instead.
- Don't worry about tablets. They often conevert even better than phones for utility apps
Yep, those certainly look worth testing. Looking forward to hearing your results!
Hey guys,
I want to give you a quick tip about the previous banners although you don't use them anymore. (Talking about this ones: http://prntscr.com/66gfpp )
It's better to present a problem + fix then to talk about something vague.
You now say : Frustrated from garbage on your phone? clean it up. (users don't know if they have garbage or what you mean with it. It's to vague.)
For example: Slow phone? Clean it // phone not running as fast as it should? We have a fix! (this are the simplest and most overused examples but now you show them a problem that they (probably) have and that you have a fix for it. Much more user orientated)
I hope to meet you guys on STM meetup I'm from Belgium btw!
@Karika, thanks for the feedback. We're making a short-list of all people we are going to meet at STM Londen. Drop your skype in a pm, lets set this up
Updating our Follow-Along to let ya'll know we still alive It has been a couple days and our 360 campaigns pretty much died out because we didn't replace the banners. Got some conversions but not enough to make any viable conclusions and this is exactly the problem we need to solve.
@Caurmen, adding LP's (from last post) didn't make much difference. Everything seems right, we constantly cut placements and steady banner ctrs of 1/2+% and lp ctr's of 30-50% but conversions are rare.
We have been struggling with the same stuff this thread is about: http://stmforum.com/forum/showthread...-strict-terms)
All spying shows more or less uncompliant banners (Adstrenght, last 14 days). They are probably cloacking? Seems you really need to push such banner designs and ad copy before people are going to install like Attila mentioned. Feels like we are throwing a money against the wall running utility app offers like 360 Security (besides the fact we're learning alot). Should we move another category like gaming or just pick another offer that has not been hit so hard?
Now we are testing some completely new designs and created a high variation banner set to properly test if an angle is working or not (we started only with 2 ad designs). Just for testing purposes.
Also we joined a Mastermind where we decided to start running sweepstakes/pin-submit offers. Bunch of good and committed guys so hopefully we will smash this vertical together.
How many offers have you tested? Normally you have to kiss a LOT of frogs to find your princess, so to speak.
Also, have you been split-testing offers across a few networks? You will not believe the difference that can make if you haven't done it.
Home > Mobile > Follow-along Campaigns