Making Afflow/Monetizer Rain - Using 1 Optimization Tool (18)
Making Afflow/Monetizer Rain - Using 1 Optimization Tool
For a while now, I've been wanting to do a case study on how to use commercial automation tools to optimize pop campaigns.
A couple of weeks back, I hit up Ervin and Losid from theOptimizer to gather some insight in preparation for the case study. During our conversation they mentioned a small test they had been doing with Afflow/Monetizer - that by setting a few simple rules on theOptimizer, they were able to optimize an Afflow campaign to green.
That got me excited. If you remember, around a year ago, I posted a tutorial on how to run Afflow on Zeropark, which can be found here:
https://stmforum.com/forum/showthrea...utorial-Part-1
Since the tutorial was first published, the game had gotten more competitive. So adding automation would definitely be a desired edge to anyone that wants to give the method a go today!
So I asked the gentlemen from theOptimizer whether I could share the results they were getting with all of you here on the forum, and they kindly said yes! This post is a result of that conversation.
All the data and tips below were provided by theOptimizer. Aside from minor editing to make things flow better, I haven't made any significant changes. Enjoy the rest of the post!
************************************************** **********
This is a 7-week long case study, where the goal was to show that it is possible to optimize pop campaigns to green with 90% of the optimization being done using an automatic tool (we still adjusted bids & budgets manually).
Optimization platform: TheOptimizer Mobile
Offer network: Afflow / Monetizer
Traffic Source: Zeropark
Geo: Italy (top geo on monetizer for the past few weeks now)
Targeting: 3G Mobile Carriers - Tim & Tre Italia
Final Results After 7 Weeks (Feb 23, 2018 - April 7, 2018):
Total Cost: $3594
Total Rev: $4123
Profit: $528
ROI: 14.7%
This campaign is currently still running and generating profit. At the end of this post we will share all the information you'll need to clone this camp and give it a go - including zeropark blacklists/whitelists, targeting, bidding details and tips, as well as the exact offer IDs from Afflow.
Why Monetizer?
It is the quickest way to start a campaign without having to constantly deal with problems such as the offer reaching cap or going on pause. You simply grab a link and run it, knowing the traffic will always go to the best performing offers.
Plus, we thought that if we could make it work with Afflow, then it would be much easier to do the same with specific offers from affiliate networks, including exclusive offers.
After a quick look at the Monetizer dashboard we decided to pick Italy. It was the top geo by volume and we also noticed that it had some nice EPC spots which we could work on making profitable.
Why Zeropark?
2 reasons here:
Their API is great - makes for smooth integration with TheOptimizer
They have all the targeting and bidding options we needed
Action Plan:
-Start with a generic monetizer link for Italy.
-Start with a RON campaign on Zeropark, then after a couple of days pick only the best-performing traffic segments (e.g. WiFi vs 3G, OSs, carriers).
-Create automatic optimization rules on TheOptimizer and let it optimize our campaign.
Process and Results:
See the chart below for a breakdown of campaign stats and what actions were taken when.
For the first week, daily records were logged. For subsequent weeks only a weekly record was logged as most of the optimization tasks occurred in the first week.
Here is an animation that shows the actual numbers from TheOptimizer's dashboard, for each period recorded in the table above:
Basically the only thing we had to do manually during the entire period, was play around with bids to see how the campaign reacted, and blacklisted non-performing offer IDs from our monetizer link.
The rest of the optimization - such as limiting placement spend, and pausing the campaign on bad days - was done in the background by theOptimizer, via the 3 rules we had set up.
The 3 rules were as follows:
RULE 1: Block publisher if spent more than $0.5 with no conversion. This is to keep bad targets from draining our budget.
RULE 2: Pause campaign if spent more than $20 and had no conversions In the event that conversions were to stop happening for any reason, this rule ensures we'll only lose a maximum of $20.
RULE 3: Resume profitable publishers If a target is blocked by error, for example if late conversions change a target from red to green, this rule would re-activate those targets.
Benefits of Automatic Optimization
Some easily-measurable benefits of this automatic optimization are:
-A total of around 1000 targets with $0.50 spent and no conversions were blocked, resulting in an estimated $500 saved. In reality if we were to pause them manually, we wouldn't have caught every one of them at the $0.50 spend, meaning a lot of these targets could have overspent past the $0.50 mark. Thus actual savings would have been more than the $500 calculated.
-Campaigns were automatically paused more than once when performance went down due to offers getting pause or other technical problems. This also saved us a lot of money.
-A lot of time was saved.
CLONE THIS CAMPAIGN
Now, as promised, I’m going to share all the data collected during this case study so that everyone interested in giving this a go can easily do so.
Step 1)Generate a link in Afflow/Monetizer
-Login to Afflow/Monetizer > Links
-Create a new link for Italy and Tim carrier, whitelisting these offer IDs: #90fdee #0507f6 #945963 #f3e144 #5d3cfe #4c0415 #6dd5ad
As shown on the screenshot below:
Step 2)Create the campaign on Zeropark
-Create a whitelist campaign with only the top-performing targets from these files (choose based on performance as you please):
https://www.dropbox.com/s/rdsv5rk7ih...mance.csv?dl=0
https://www.dropbox.com/s/9poskkrkjb...mance.csv?dl=0
Targetting: Italy, 3G Only, Tim, Android
Bid = $0.027 (it may look high but we noticed that only high bids perform well)
Budget = $20 (increase if performance is good)
Create the rules on TheOptimizer as shown in the screenshots above, so that they will start executing as soon as the campaign starts getting traffic.
Lessons-Learned & Tips
-If we were able to have profitable campaigns simply with a Monetizer smartlink, it should be much easier and stable with direct offers from networks.
-Automation is a must if you are looking to scale your Pop campaigns. It is almost impossible to do manually even with one campaign
-Exploring high bids is a must. Many times affiliates do the mistake of wasting all their testing budget with low bids and then at the end giving up. Low bids never worked for us, we only started seeing some good results once we started being on the top 3 positions.
-On zeropark don’t try to be 1st; The estimations zeropark shows regarding positions or top bids are not real time so you should not base your decision on those. If you found a winning bid that allow you to spent a decent amount daily and have good results just stick with it.
-Keep an eye on the targeting and make the necessary changes if a given device, OS or carrier is not performing good.
************************************************** **********
So that's it! The entire method laid out in detail. Many thanks to theOptimizer for allowing us to share this case study with the forum!
Note that the above case study only involved a single geo. If you don't want to be competing with a bunch of other members (there will inevitably be people that will try to rip this campaign in its entirety without changing anything), pick some other geo+carrier to target - my Afflow tutorial goes into detail on how to choose highest-performing geo+carriers:
https://stmforum.com/forum/showthrea...utorial-Part-1
Also note that Zeropark isn't the only pop traffic source you can use theOptimizer on. PopCash and PopAds will also work - however, you may need to tweak the method a bit to cater to the different interfaces and campaign options.
Amy
Thank you so much for this beautiful case study, AMY!!!!
Great case study!
It shows that pops + afflow can still work on certain geos and pockets of traffic.
Curious to see how many people will be able to replicate this.
Massive value, as usual Amy, thanks!
I have tested those offers in the past few weeks and had some good success with some of them but unfortunately the best converting ones just got paused today! Maybe it's time to look for a new Geo/Carrier at Monetizer but also not sure if I should continue with direct linking.
The Monetizer + Optimizer combination is powerful indeed. So many little gold pockets for those starting out.
Amazing case study Amy!
Nice use case of both The Optimizer and Monetizer = Automation beasts :-)
Hi Amy! Great tutorial!
Curious... how did you define your cutting threshold of $0.50? Doesn't seem payout related here...
Originally Posted by silvertongue
Basically you can never be 100% sure whether your cut-off rule is too high or too low.
Too high, you waste money on dud placements. Too low, you cut a placement that would have turned out profitable.
So, pros and cons. And the trick is find that sweet spot that will give you maximum ROI.
Which is not an easy task, as Monetizer often sends traffic to multiple offers, so you can't really set the cut-off to a multiple of the payout.
The guys at theOptimizer, though, must have extensive experience on data analysis. And that cut-off of 0.50 must have worked well for them.
Also: For a tier 1 geo especially, that has a ton of placements, you would much rather "err" on the side of spend, by cutting placements earlier rather than later - because you can stand to lose substantial amounts of money otherwise.
Furthermore: Once the camp is in profit, you can always "revive" blocked placements by running more spend to them. Especially the larger ones that can give you traffic volume - you may want to test them more thoroughly.
Amy
Sent from my SM-G930W8 using STM Forums mobile app
Originally Posted by silvertongue
A lot of times you just can not afford to spend 2x, 3x payout for each placement, it would result in big testing budgets.
Instead we decided to go for a more aggressive way. What we did was, after gathering some data, do the math of what was the average CPA (cost per action) for the placements that already converted (how much a placement was spending before seeing a conversion). The value that came out was $0.5, so we decided to use this as an optimization metric.
Could we have cut off some good placements because we did not give them the necessary time? Sure we could've. But overall I think the aggressiveness paid out.
This way we focussed on the placements that showed some potential immediately, instead of waiting and hoping that the rest would be good.
Originally Posted by vortex
Yeah, I find that playing it "safe" often ends up costing me, because I end up seeing 2-4 sources sending me 80% of the traffic in testing phase... not enough to judge the potential of a campaign in my opinion. That's why the $0.50 peaked my curiosity here...
I like the idea of an aggressive start and re-introducing "grey" sources later on, i'll give that a bash.
BTW, absolutely loved reading the report you and Matuloo released yesterday! Incredible work!
Originally Posted by ervin
Hi Amy,
Would like to ask something about those settings on LP Entry Pop, etc. Do you leave them to be on default or you have everything enabled including LP Exit Pop? Is the LP Exit Pop making any difference if it's enabled/disabled?
Thanks.
Hi Amy!
The offer id's you whitelisted in Monetizer, were those Italy specific? How did you choose them? When creating my Monetizer link I entered the country and choose "Rotate All Offers". Its day three using the Optimizer tool with Monetizer and Zeropark and I have only had 2 conversions. Does that seem realistic in today's climate or should I use the offer ID's you mentioned or should I just try and different Geo?
Thanks.
Patrick
Originally Posted by patrickpeak
Please read the tutorial again to get a better idea of the different approaches. Basically you can either:
1)Target a specific offer from the start if you see good stats in monetizer for that offer.
OR
2)Not specify offers - just specify geo and carrier for example and let all available offers run. Then you analyze your stats to see which offers are converting the best and then target them.
Hope that helps!
Amy
Hi @vortex; or @ervin;, whaty type of traffic at ZP would you recommend, standard or premium. Any experience with that?
Originally Posted by janm82
@janm82 - Just for quick reference with he way our traffic works tiers work.
Standard = All available inventory.
Premium = Inventory only from sources with proven traffic quality.
Standard traffic is cheap and higher volume so will take more optimization, premium will convert more but will be more expensive.
Home > General > Affiliate Marketing Forum