Whisper's journey to a profitable pops campaign (32)
Firstly I just want to say thank you for this amazing resource and all the incredible information and guides that are available on this forum. I would never have been able to get this far already in my journey if I hadn't made the good decision of joining STM. Particularly mrbraun, caurmen, matuloo and vortex have written a lot that has inspired and guided me so far 
.
Yesterday I read wakeboarder's follow along and was inspired to do one as well since he is at a similar stage in his journey to me and seems to be having some good success.
Since discovering Amy's excellent 40-day newbie tutorial about a month ago I got really excited and took a lot of action. I have run about 20 direct linked campaigns, one of which I managed to get a positive ROI, but I have not been able to increase the volume to anything meaningful.
Direct linked campaign with positive ROI (this is seven days worth of traffic):
My totals so far (past thirty days):
Impressions: 947,510
Conversions: 162
Revenue: $48.08
Cost: $330.70
ROI: -85.46%
Note that the real cost is about $100 more than this since it took me a while to get all the tracking set up properly and reporting costs.
I became a bit frustrated with direct linked campaigns not working well. I got very excited by reading this post from mrbraun, and desktop pops/sweeps with a lander showed a lot more promise, so I switched my focus to this and have tried a couple of campaigns which seems to be getting more conversions. My goal now is to get one of these to positive ROI and profitability.
My approach so far has been less organised than I would like, so I am documenting my steps here so I can be a bit more scientific in my approach.
Here are the things I want to work on this week:
- Set a weekly budget for ad spend
- Set up a SYSTEM for testing geos/offers/landers
- Attempt to increase volume to the direct linked campaign I have running with positive ROI in order to make it worthwhile
- Find the best performing pops campaign and attempt to optimise it to profitability
- Re-read The 40-Day Newbie Tutorial - Intro and Index
Weekly budget for ad spend
Amy wrote a good post about not setting goals for $XX or $XXX per day, and instead setting a goal for weekly ad spend. I like this, since ad spend is something completely in my control and removes attachment to any one particular campaign.
So I will set my weekly spend at $100. This will allow me to test 6-10 camps per week.
Set up a SYSTEM for testing geos/offers/landers
I think for a system, wakeboarder's one for testing offers is pretty good:
1) Find offers by asking AMs for offers satisfying the following criteria: tier2/3; max payout $1.5; sweepstake; ... Win iPhone X or Samsung S9
2) Rip the the most popular lander from AdPlexity for particular GEO/offer
3) Spend $10-$15 to test each offer
4) Document each campaign along with total spend, ROI etc in a spreadsheet to force a more scientific approach
5) Look for positive signs and optimise the camps that show promise.
I have a question about this, how do I know when to stop launching multiple campaigns and instead focus on optimising one? In other words, what markers do I look for to know when a campaign shows enough promise to be worth optimising?
Attempt to increase volume to the direct linked campaign I have running with positive ROI in order to make it worthwhile
I'm not sure how to do this. Should I be upping my bid on PropellerAds? Trying different traffic sources?
Networks and traffic sources
1) I am currently signed up with the following affiliate networks:
Big Bang Ads
Addiliate
Gotzha
PeerFly
Affiliaxe
Max Bounty
ZinQMedia
Is this enough? Am I missing any good ones? Even though I am signed up to all of these, I still struggle to find decent offers.
2) I am relying on PropellerAds for all of the traffic. Should I be getting traffic from other sources as well? When is the right time to do this?
I have accounts with the following traffic sources:
PopCash
ZeroPark
PropellerAds
PopAds
Wiget media
RTX Platform
Exoclick
Plugrush
Am I missing any? Which one should be my next pick after PropellerAds?
I know it's a lot of questions, so even if you can just answer one I will be very grateful! Thank you again everybody.
First of all congratulations on getting your feet wet on campaigns and on the first green campaign!
Since things have changed quite a bit lately, going for direct linked campaigns makes campaign optimization somehow harder compared to using landers.
Depending on the offer type you are running, one thing I would suggest on trying out with your profitable direct linked campaign is to ask your AM if they can provide you the deep links of the offers you are pushing. That way you can start using similar pre-landers provided by the affiliate network as well as ad like 2-3 LP variations to see what lander drives most of the conversions.
Considering that you already have a profitable campaign, if you have room to allocate more budget on that campaign that would be positive. This way you will be able to milk out even more profits from it as well as go through the process of scaling a campaign. Keep present that pop campaigns don't last too long and because of that you have to scale them on time.
Every network is interested on pushing their offers as the best ones while on the other side you are interested on maximizing your revenue of every dollar you spend in order to cover operational and traffic costs + profits. When testing new offers, landers and geos in my opinion the best approach would be to work on a combined set of information you are able to get. Gather all the information you can get from your AMs about their offers (including volumes, conversion rates, network EPC and best performing landers - when they can provide you with such info, then based on that try to find out on how these offers are performing using spy tools. Don't just use spy tools to rip landers since the most valuable information you can get from them is the offer (where it is being run and how).
When it comes to testing new offers on new geos, 10x the offer payout on the big sources may not be enough. For instance you want to test a UK offer with a $1 payout on propeller or zeropark. In the first $10-$15 ad spent you will have to aggressively optimize to cut out crap placements then check how the offer works. As such, focusing on one geo/vertical at a time in the beginning will save you quite a lot of ad spent budget and frustration.
It seems like you already have all the traffic sources and affiliate networks you need to get good results. Some of the networks listed in your posts are advertisers as well which means that you are cutting out broker margins. Lastly if you see some potential on some of the offers you are pushing ask for a payout bump, odds are that you are hardly gonna get a profitable camp on street payout.
Thanks for the reply platinum. I have one question, what do you mean by deep link?
Originally Posted by whisperscuba
Hi whisperscuba!
Great to hear that you were inspired by reading my FA.
Originally Posted by whisperscuba
Does the offer convert?
When testing an offer with low payout (< $1.5) and I get less than 3 conversions I don't even bother optimizing further.
What's the ROI?
Of course, there is a correlation between the number of conversions and ROI. If I have ROI after the initial budget lower than -60% ... I move on.
Traffic volume
So far I have multiple offers with initial ROI around -50% but I decided to move on because of the combination of carrier traffic and GEO. If there is not enough volume on the network to spend $50 / day, I rather spend more budget on testing new offers. Eg. GEO Cameroon and carrier Orange... it can be -50% ROI after the initial test but there is not much inventory available to optimize and scale later down the road.
I am learning the ropes so please take my advice with a grain of salt. I believe more experienced STMers will give you more accurate instructions and we will both learn from them.
We just had to keep testing and testing... I believe when we hit the right offer, we will both know that immediately. Two days ago I had one, but it was capped.
@wakeboarder if you find a good converting capped offer like the one you mention, the best thing to do is to confirm the quality with your AM then ask for higher caps.
I’m afraid there are not any general rules based on what ROI to cut an offer after some testing. At least without knowing the performance of the offers any wrong suggestions can lead to giving up on some potential profits.
Of course it is silly to keep on sending traffic to a half dead campaign which never gets to go above the -50% ROI, but this is why split-testing is a must. It helps on finding good converting spots with the best converting offer.
Another thing worth considering is the amount of data you have for a specific geo/vertical. Create your lists and reuse the data you have already purchased, it’ll heavily cut down testing and optimization costs 
Hi everyone, so an update.
I have narrowed it down to one campaign that shows a lot of promise. It started out at -60% but I did some lander testing, cutting two major zones and finding a better offer I have it hovering close to breakeven. I am now starting to test a lot of things to see what I can do to get this campaign into stable, positive ROI. It's in a decent geo with plenty of volume available. This is the first campaign I've run that looks like it has the potential to go profitable AND scale so I am very excited.
I have used both of my competitive advantages so far, in both modifying the lander code and working in a non-native language that I happen to be fluent in.
Here is what I am currently testing:
- A spread of bids, 5 to be exact at different intervals around the original
- Mobile only vs. desktop only traffic
Here is what I want to test:
- 3 different variations on the winning lander theme (will knock these up over the weekend)
I have one question which seems very curious - here is a comparison between browsers:
Look at that difference between Firefox and Chrome! I am very surprised to see this because my campaign isn't even personalised to browsers. Does anyone else have an idea why this might be?
Here is another update.
I have continued to focus my efforts on this campaign, bearing in mind what Dr Ngo has said about polishing a rock until it's a diamond, rather than throwing shit at a wall until it sticks. I had a great result with a split test today. Adding just one small photo to the landing page increased it's conversion rate by 258% with a 99% statistical significance over the other one!
I have also expanded to another traffic source and written an automation tool that continually monitors the campaign and applies some rough optimisation rules to rule out bot traffic and poorly performing placements.
I have also rotated in two new offers to test. So now I am split testing three offers and two landers - I have another type of lander that I need to fix up and test as well. My stats for the past three days:
Here are some lander stats:
I wonder if there are any more in depth guides to campaign optimisation apart from vortex's excellent one in the 40 day tutorial?
So, here is another update.
For some reason which I cannot understand, for the past three days this campaign has made no conversions at all. It's not the offer because I am split testing three, and the domain has not been blacklisted by google.
The only thing I can think of is that I made a few small changes to the landing page copy due to a request from an affiliate manager. It's possible this small change has tanked the conversion rates this much, although it seems unlikely.
I am now not sure whether to keep sending traffic to this camp (is it beating a dead horse?) or instead to invest my energy into testing large numbers of offers with little investment in each one.
It's likely you are being scrubbed by the advertiser because your lead quality is bad. Another possibility is that a placement that was previously sending you conversions no longer is, because someone is out bidding you or something.
Hard to say exactly but 90% of the time if there is an issue it is the offer itself.
Andrew
So, this was super weird.
Today I spoke to my affiliate manager at one of the networks about this, and he said it was weird because 19 conversions came through in one day! I check the two other offers I was running and they all had the same thing.
So it must be the case that all networks are using the same upstream offer and the postback was delayed for three days (!). I checked in
So my real stats for the past three days are much closer to breakeven.
It makes me wonder though... if all three of these networks are rebrokering the same top-level offer... why can't I just go to the source and work with them direct? I'd probably get better margins...
Just in case, before approaching the advertiser directly make sure you have the required budgets to run the required volumes for them. In most similar situations, advertisers may choose to bypass the network in case they can get high volumes directly.
So, I talked to a few people from the forum and read a post from Mr Payne and came to a realisation which is that pops campaigns make or break by the offer. So I need to be testing a LOT more offers to find the golden nuggets.
I have got my one sweeps campaign in ES to profitability but the bid is low and the traffic is limited. If I increase the bid, it becomes unprofitable again. So I don't think I can scale this.
Instead I am shifting strategies to "throw shit at the wall til it sticks" . So starting today I will be launching 2-3 new camps every single day on a competitive bid with a wide variety of geos, verticals and offers. I will run $10-$20 of traffic through each camp and throw away the ones that don't show promise right away. Today I launched:
- A push notification campaign for my ES sweeps that was already working
- 1x adult dating offer using direct linking
- 1x sweeps offer in new geo via mobile POP
- 1x sweeps offer in new geo via mobile redirect
Do you guys think this new approach makes sense? And the important question... HOW can I tell if a campaign shows promise? Should I look for positive ROI right off the bat?
Following some advice from Platinum I implemented a bot detection technique and already weeded out a few bot placements. It has helped my ES campaign to do quite well, today I scaled up traffic massively and conversions have kept up. See stats:
So I ran $100 of traffic through it today and ROI is still in the green! Tomorrow I will increase my budget and run more - I am collecting more data all the time and this will put me in a strong position to ask my AM for a payout bump on the offer.
I want to do everything else I can to optimise the ROI further into the green. Anything I can think of to try to boost the conversion rate of the lander by any more percentage points since this will make my margin. Any advice on what strategies I can try? I also read about some JavaScript tricks to implement on the landers, is there a comprehensive guide to optimising landers somewhere?
NOTE: I already read these:
https://stmforum.com/forum/showthrea...actics)-Part-1
https://stmforum.com/forum/showthrea...actics)-Part-2
https://stmforum.com/forum/showthrea...-Tricks-PART-1
Some tests I can make:
- Add a sound
- Try different variations on copy
- Try different sizes/location of the image (I know that image converts better than no image)
- Try different targeting (e.g. city, brand, browser, referer etc)
- Add some arrows
- Try some animation (e.g. show a spinning iphone on page load)
Don't know why I've missed this thread until now! GREAT work whisper! You've tested a lot of stuff including some of the advice from veteran members and are getting good results - that's always amazing to see.
Regarding increasing lander performance - here is a list of threads ("On Landing Pages" section):
https://stmforum.com/forum/showthrea...-am-faq-thread
While improving lander performance can be a very worthwhile project, if I were you I would make it a priority to scale this campaign to as many traffic sources as I could. And not just the ones you already have accounts on either, but also new ones you've never tested before. It's always cheaper to test new traffic sources when you have a working lander+offer (not to mention the fact that without a converting lander+offer, there'd be no way to test the traffic on a new source).
Here's a post that contains a list of links that have traffic source recommendations:
https://stmforum.com/forum/showthrea...l=1#post327140
The more good traffic sources you have access to, the faster and bigger you can scale all your future campaigns. There's no better time to start branching out than now!
The second most important task would be to test more offers. It's usually easier to increase ROIs by larger degrees by testing offers than by testing lander variations. Of course that will also depend on your luck and skill, but the point is not to stop testing offers. Even when you have a profitable campaign already, testing new offers don't have to cannabalize on your profits - you can run a smaller portion of the traffic to the new offers to test them.
Have fun!
Amy
Thanks Amy, really useful advice. I will focus on scaling this camp to new traffic sources.
Just looking at Zeropark now, it's quite different from Propeller. I'm wondering which of the following options to use:
Optimised, Premium or Standard? (I assume Remnant is just garbage)
Haven’t run any campaign with Optimized inventory type but looks interesting. I’d suggest starting with Premium first then move on to Standard (for scaling) once your campaign shows some good potential.
On Premium inventory bids may be slightly higher compared ti the Standard one but traffic quality tends to be higher as well.
Sent from my iPhone using STM Forums mobile app
So, I implemented your advice Amy and spun up three new campaigns on new sources. Two of them have done very well, even though the optimiser still has not had time to run for very long. Here are my stats for today:
I am still on street payout and negotiating a bump with my AM. Street payout is €1.65 and she offered me €1.73 which I think is not fair. Now that I am sending $150 per day of good quality leads I think something more like €2 would be reasonable. This would put me strongly into the green and allow me to throw a lot more money at traffic, which would be a win for everyone.
My automation software hooks into
- Cut any placement sending > 95% bot traffic, using various bot detection techniques
- Cut any placement >= 0.5x payout in loss with no conversions
- Cut any placement > 2x payout in loss
- Cut any sources > 2x payout in loss (Zeropark only)
These rules are applied automatically every 10 minutes.
With a payout bump and the new sources I have spun up I hope I can make this campaign profitable, scale it even more and recoup a lot of the losses I spent in testing to find it . Maybe I can even come out at the end of the month in the green! That would be a huge success for me.
Great results Whisper! Keep up the good work
Sent from my iPhone using STM Forums mobile app
BAM! Good job!
Don't worry about negotiating a higher payout. I don't know which network you're working with and I'm certainly not referring to any particular affiliate networks when I tell you that many networks WILL shave leads. If you push for a bump too hard, chances are you'll get that bump but you'll get a corresponding shave in the backend that will leave you with the same or even lower profits at the end of the day. It would be better to test more offers instead - or even the same offer available on other networks. If you're not wanting the new offers to eat into your overall ROI, just send a smaller portion of the traffic to them.
And I would suggest that you focus on testing even MORE traffic sources while you have a profitable campaign on your hands!
I know you're wanting to make your current campaigns green with a higher ROI - but think about it - you're spending $150/day right now. Even if you can increase your ROI from 2% to 30%, you'd only be making $50 in profits. As for recouping your losses - unless you're running low on budget, your focus should be to maximize future earnings rather than recoup past losses, because the past is the past. By testing more traffic sources, your chances of hitting your next profitable campaign out of the park becomes exponentially bigger. And by using theoptimizer you'll be building blacklists so that any new offers you test in the future, you'd be testing on good placements so the testing would be much cheaper. These are investments that will give you an edge and better-equip you to win over competition that is starting to test from scratch, or that doesn't have access to as many sources as you do.
Plus, when you scale bigger, you can test so many more offers and reach statistical significance so much faster due to having a lot of traffic volume.
But I'd understand if you're running low on budget and want to make some profit right now. Listen to your heart and feel your way into things. If you're not ready to scale big, hold off for now. But remember that the sooner you make the investment, the sooner you'll reap the benefits.
Have fun! And thanks for all the updates!
Amy
Thank you Amy for your clear advice. The message I am getting is: forget everything else and just scale on new traffic sources. I'll shoot to get it running on at least two more today. How many is a good number? Five? Ten?
To clarify, by traffic sources you are talking about both different companies (PopAds, Exoclick etc) AND different types (Popup, Redirect)?
I also have one other quick question about day parting - is it worth it?
For example, here is a screenshot from my stats:
Should I bother blocking the unprofitable hours?
Originally Posted by whisperscuba
Don't know if I've dropped this link in your thread yet - but it contains a list of links to different pop traffic network recommendations - would be good places to start:
https://stmforum.com/forum/showthrea...l=1#post327140
When I say "traffic source" I mean all the different places you can get traffic from. Basically the more good-quality traffic you have access to, the bigger you can scale your camps - whether it be traffic from different traffic networks or different types of traffic within the same network. As long as the extra traffic can potentially increase your profits, then you'd want to buy it.
For example, here is a screenshot from my stats:
Should I bother blocking the unprofitable hours?
I would day part if I see a big chunk of time that's clearly in red, for example during sleeping hours - that kind of trend makes the decision easy. But if I see a random/sporadic "pattern" (or rather, a lack thereof) - such as an hour or two here and there that are in red - I usually won't try to day part. Usually, I'd sooner cut placements or OSs or browsers than times-of-day - unless I really need that extra bit of optimization to push into green.
Another way to make use of day parting is to apply that in the beginning when you're doing initial testing (of landers/offers) and want to get the most bang for your buck. Typically the best-converting hours (and this is true for many verticals and geos) are the evening hours between when the local people get off work and when they sleep - so around 5-6pm to 10pm-12am. If the geo you're targeting has a lot of traffic and you're trying to test landers and offers for cheaper, you can just target those few hours every day and still get the testing done in a reasonable number of days. It's not very significant savings, but when cash flow is an issue in the beginning, and when you apply that to multiple geos, you could save enough to launch one more campaign than you otherwise would be able to.
(You can actually apply the same reasoning to other traffic segments: After running in a geo for a while you'll know for example which placements are the best and the worst, so every time you want to test new offers/landers you can just either do that on a blacklist with the worst placements excluded, or a whitelist of the best placements. Then you can take the winners and test them on the rest of the placements.)
As for your stats: It DOES seem like there's a big chunk of time of many hours that are all in red, but it's kind of early for any action to be taken as there isn't a lot of data yet. I wouldn't bother checking for statistical significance for this type of thing - pop camps are volatile and performance can decline over time. I would suggest to just wait until you think to yourself "the trend is pretty clear that this big chunk of time just won't likely make money for me" before considering blocking it out. And before you do, try to drill down into other traffic segments first (placements, OSs, browsers, etc.) to make sure you cut segments that have worse ROI first - because after you do, the originally-red hours may turn green.
The same logic can apply to other traffic segments as well - many segments are inter-related and there's an overlap, so that when you cut a segment that has bad ROI, the overall ROI of the rest of the segments will increase. I try to start cutting from the smaller-volume segments that have the worst ROI FIRST - especially if there isn't a ton of traffic volume to begin with - and leave the bigger-volume segments for later (because cutting the smaller-volume segments may already make them green).
Having said all of this: Try not to fall into the trap of over-optimizing campaigns. Focusing on testing offers and scaling campaigns will increase your ROI and profits in leaps and bounds compared to spending all your time tweaking campaigns to increase the ROI by another 5-10%. Do larger tweaks and avoid smaller tweaks or else you'd go past the point of diminishing returns on your time and effort.
Will end my rambling here to avoid making things unnecessarily complicated!
Amy
Thanks Amy, this was really great advice as you can see here:
I discovered a new traffic source which is doing absolutely fantastic. So crazy that ROI can differ as much as 100% between sources!!
As for my other sources, I've paused Zeropark Pops and Propeller Ads SmartCPM because they weren't performing well even after a lot of optimisation. Propeller smartCPA is making me about $10-$20 per day, Zeropark redirect is about breakeven and this new source is just absolutely killing it. I have taken the daily limit off and will buy as much inventory as they can sell me at this bid since ROI is great.
I have a camps on exoclick and hilltopads for this flow ready to spin up (waiting approval) and will also try trafficstars based on a friend's recommendation.
While waiting for those, I am experimenting with a sweeps offer in Asia. Initial results seem promising, around -40% ROI on the first ripped lander I found with no optimisation, I'd like to try a few different offers/traffic sources here and see where I can get to. Payout is low ($0.10) which means testing is cheap so I can low daily testing budgets and very quickly get an idea of how it will perform.
In other news, over the past 30 days I have run 3M events through
I'm going to get hit with some pretty nasty overage charges this month so I have been looking into a self-hosted tracker solution. Lots of people have good things to say about
Very nice to see 2-digit green!
3 and 4-digit greens can be achieved similarly - basically it's about juggling more campaigns in more geos on more traffic networks.
Those camps you have on ZP and Propeller: It would be a good idea to test another batch of offers while you're at it, since you've already invested into cutting placements and finding a good lander. Failing that, you can try going them the "last ditch effort" by drilling down into different traffic segments and pausing the lowest-ROI segments to increase profits. If however you're left with just a trickle of traffic afterwards then it would be better to pronounce the camp dead, instead of leaving it running and having to spend time on monitoring it. The risks of it dipping into red would be great as it hovers near break-even and it just may not be worth the time to monitor it to make sure it's still green.
Amy
Thanks for the response Amy. When you say another batch of offers, do you mean in the same vertical? For example I am running iPhone X sweeps, so maybe I could try Samsung S9 or a gift voucher sweeps offer? Or could it even be a completely different vertical e.g. app installs.
I have had difficulty in finding good sweeps offers for ES geo, this was the best one I found so far.
Originally Posted by whisperscuba
However, to prevent lander/offer combo saturation on those good converting publishers, a good option may be to introduce a totally different lander/offer combo. Something like a Supermarket Voucher or even app install instead of a Win iPhone.
You may find a good list of sweeps offer source from spy tools. Make a shortlist of the networks and see if you can get access to their top offers for the targeted geo.
Originally Posted by whisperscuba
But obviously if you have trouble finding offers in the same geo and vertical, then you may want to try another geo or vertical.
Especially if you're confident you've already tested most of the good offers for the original geo. But do try platinum's suggestion of looking in spy tools to see which networks the offers that are receiving the highest volumes are at and apply to them to test those out.
Amy
Sent from my SM-G930W8 using STM Forums mobile app
Thanks for the reply Amy. Actually I haven't found any other offers that work better than the current one. It's past breakeven now so all further cash it generates is basically hands-free money.
I have chosen instead to leave this camp, let my automation tools run it on autopilot for now, while I take the winning lander over to other geos and use the profits to fund experimentation with new camps.
Not quite at $XXX profit per day yet (soon!) but check these numbers:
This camp really is the gift that just keeps on giving. Pretty pleased with this progress considering I only started my AM journey in May .
Congrats whisperscuba! You're doing an amazing progress.
Impressive progress for sure!
Scaling is exactly about what you're doing: Taking over additional geos and also testing more traffic sources to expand your reach.
So happy for you and eager to see you reach new heights!
Amy
Sent from my SM-G930W8 using STM Forums mobile app
Hey Whisperscuba,
How's your journey going on?
Would love to get an update from you.
Home > POP / PPV / Redirect > Follow-along Campaigns