Pops + Afflow (15)
Great start, it's perfect that you've been able hit some profitable campaigns straight away.
Looking forward to your future results!
Originally Posted by matuloo
Just a case of cutting some sources/targets and expanding in to other geo+carrier combo's.
Very interesting, how did it go?
Missed this for some reason! Any updates?
Amy
Sorry for the lack of updates with this - it's been a bit 'stop start' over the past 3 weeks or so.
After the first week or so I decided to look into getting some kind of web app created to take the live feed data and have it visually displayed in a dashboard.
I am a complete and utter newb when it comes to progamming etc but managed to find a guy over at UpWork who was interested in taking the project on.
Just as we'd finished putting it all together Martin (@manamanamedia) came out with a similar dashboard.
In hindsight it would probably have served me better to simply throw that budget at more campaigns (close to $900 total cost for my dashboard) - although I think there are still ways to take the data it displays and make it more useful. I'll probably make it public (as Martin did with his) once I've tweaked it a little and looked to move it - it's a little slow in terms of page loading and it's looking like a hefty bill each month based on the server etc that the developer used.
Currently, the live feeds data doesn't include carrier info - just country / vertical / payout (plus other details such as browser, lead time, offer id, os).
What I've seen from testing is that everything is very 'liquid' - a carrier + vertical combo may be flying for a couple of hours and then die off completely, and vice versa - it's certainly not something you can set and forget, and just log in to
I haven't ventured into any adult traffic / offers or WiFi yet.
What I've seen from the carrier side is that most countries have 1 or 2 primary carriers that convert on a semi regular basis - although it is all quite sporadic in nature.
So far I've deposited $785 with ZP (although the actual credit in my account is quite a bit less once the PayPal fee and VAT (20%) has been taken off). Current balance at Afflow is approx $300.
Looking at the Afflow CVR values and my own I think it's clear people are getting some high quality traffic from somewhere other than ZP.
Despite my current position (-$1579) = 785 traffic + 900 app + 194
The method currently fits with what little spare time I do have and I can see pockets of profit here and there - I think you'd need to be pretty dedicated and use automation to make this your "bread and butter".
My only 'stumbling block' at the minute is FOMO. I'm torn between just hitting one country and focusing on that and that only (ie ZA, AU) or working on a small 'portfolio' of smaller countries.
If you want to take a more long-term approach (relatively speaking - longer than the couple of hours windows you're talking about), try this:
1)Identify segments that have been performing consistently well for at least the past month. Preferably longer.
2)Invest into cutting placements for that carrier and/or geo so that you're only running the best-converting ones.
To take advantage of the short-bursts, your choices are basically either:
1)Keeping track of the "trends" chart to identify large spikes in eCPM, activating the camp (you'd need to have a lot of camps set up ahead of time for this - e.g. one camp for each major carrier; it would be too late to set up a camp from scratch and wait for it to be approved), and continuously monitoring offer stats to make sure the traffic is still going to high-eCPM offers (or stopping the camp when it isn't).
2)Use automation like you're doing, to identify these trends 2 hours earlier than everyone else. Better: Allow the automation tool to activate a camp automatically when a good trend is found, and pausing the camp when the trend is over. I'm not a programmer, so will need to leave the details to you to figure out.
The second step will help you to optimize any camp and not just afflow camps.
A good start would be to automate the cutting of targets and even sources, according to the criteria you specify.
Another good thing to automate would be the tweaking of bids for each target. For example if a target is in loss, decrease the bid, and if the target is in profit by more than a certain ROI, increase the bid.
Of course it's not as easy as it sounds. The programming part is relatively easy - it's the algorithm that will need to be refined again and again until you have something that works.
But if you do make a tool like this, it can give you a big edge over competition that are optimizing manually.
While tweaking your tool and test method, it would be a lot cheaper if you could use a cheap geo with low payout offers. ID, BD, EG immediately come to mind. This way you can collect a lot of conversions to allow you to see quickly whether your latest algorithm tweak is on the right track or not, and give an idea on how to tweak it further.
Focusing on tier 1 geos can bring great rewards, but would also require larger investment in budget, because you'd need to test and cut a lot more placements, and spend a lot more to collect data because payouts are much higher. It may be good to use tier 3/4 geos or smaller and cheaper geos to recap some of your investment first, before attacking the big and expensive tier 1 geos.
Thanks so much for the update! Wishing you the best of luck and looking forward to hearing about further progress!
Amy
Thanks for the reply Amy!
I'm going to take your advice and focus on a small number of lower tier/lower payout countries and take it from there.
I've got a lot of camps all set up and ready which is good - I've just got to spend some time getting my head around the data my tool provides and trends etc.
I'll definitely be speaking to my developer re additional automation in the near future.
Thanks again for all of your help
Sorry for the lack of updates recently!
It's been a bit of a slog recently with no real progress to report 
Having spent a few hours putting data together etc to decide on the geo's I'm going to specifically target my latest ZP campaign was rejected with the following message:
Your campaign was rejected for the following reason:
It came to our attention that recently there were problems with fraud detection in Afflow campaigns due to unclear compliance mechanisms Afflow uses right now. We are waitinig on confirmation from Afflow about their methods of traffic detection and in the meantime would like to ask our users to abstain from running Afflow campaigns on Zeropark until the issue is resolved. If you have any questions about traffic detection in your Afflow platform with current campaigns, please speak to your Account Manager about their Traffic Detection System. Afflow campaigns that are ran on Zeropark will be rejected by our compliance team starting from today.
I'll be trying to get hold of my Afflow account manager and will update as and when I have more info.
I got the same message but my campaign was approved minutes later. Did you check the campaign on ZP?
Yeah, the campaign was listed as 'Rejected' in my ZP dashboard.
I've also paused all of my other Afflow camps with ZP until the issue is resolved, don't want my account there banned.
Have emailed my Afflow account manager and will report back any response.
One note on afflow trend detection: data filtering is very important in this sort of tool. You can use moving averages and other similar things (linear regression is probably appropriate here) to identify longer-term trends, which may be more valuable, but will probably be lower ROI.
As Amy says: counter-intuitively, the coding's the easy bit here. It's the data manipulation / data science end that's harder.
It appears that the Afflow/ZP issue is now resolved having spoken to ZP and my Afflow AM.
Will return with a more 'in depth' update of where I am with this project shortly - once I've started to resume my camps
Yes the issue should be resolved now.
ZP is a great network! But do keep in mind that there are others you can test and scale to.
Amy
Originally Posted by vortex
what other traffic network would u suggest is also working well. i.e propellerads?
Originally Posted by omar17
And if you want to avoid the competition, you should try smaller sources.
And don't forget that afflow also has adult. So try Exoclick, Plugrush, etc.
Amy
Home > Mobile > Follow-along Campaigns