Quitting my day job by the end of the year - [Prize Possibly Included] (25)
Quitting my day job by the end of the year - [Prize Possibly Included]
I decided to ditch my previous follow along as it was a bit fragmented. However, it's a great resource with a lot of valuable information on campaign creation and the mistakes you should avoid when setting up a campaign.
I'd like to thank everyone who contributed to that thread especially Amy! She was patient in helping others succeed and I feel like she's one of the reasons why STM is so awesome! This follow along won't be as thorough as my previous follow along but would be most likely used for displaying day to day stats and new techniques on improving campaign execution.
With this new follow-along, I'll be working on pop traffic only; verticals are open to app installs, but right now mainly focusing on sweeps. No display, no nutra, no facebook. I want to be really good in this particular traffic method. Now you maybe thinking what's the prize you're talking about Simon. Are you going to rob a bank and make all of us rich? Unfortunately, I don't possess the physical strength nor the emotional intelligence to rob a bank. 
The prize is one random member by the end of the year will get $500 paypal'd to them if I don't quit my day job by December 31st 2016. This would be a good way for me to be accountable for my success.
Next Goals:
Reach $100 profit a day by the end of April: [ ]
Reach $200 profit a day by the end of June: [ ]
Reach $400 profit a day by the end of August: [ ]
Reach $500 profit a day by the end of October: [ ]
Quit Job by December 31, 2016: [ ]
Current progress:
I'm consisting reaching $50~ a day profit since Wednesday(3/16).
- There's this one iphone6s sweeps offer that is available for 3 countries.
- Created 6 campaigns on two traffic sources. 3 campaigns on one traffic source and another 3 on the other traffic source.
- Working with AM to develop relationship with merchant and check on lead quality
- I feel as if I haven't optimized my funnel yet. I used the best lander/offer based off past data that I had. I didn't add anything with the winning landers, I just entered more CVR boosting techniques(back button, delayed meta refresh, popups). Yesterday, I created a new lander using Adobe Muse(thank you to Zeno for his awesome tutorial) to test against the winning lander. Depending on which lander wins, I'll start optimizing it through changing headlines, cta, colors, etc....
How I actually reached to profit:
In the past, I looked at the data in
1) Created a campaign solely on targeting the profitable OS-Browser.
2) After a week of collecting data, I will create a separate campaign targeting only the profitable placements while running on all browsers. While the old campaign will be used to scrape up new placments, test offers/landers.
Optimization Phase:
1) Haven't scaled it to many traffic sources yet due to some capital, but getting paid out soon from affiliate network. Scaling to other traffic sources is optimal for me, but I still need to find out whether or not I can still improve the overall funnel. One of my goals is to launch a new campaign onto a new traffic source every week
2) Take time to research about how to block my pages on spy tools or at least make them harder to find for intermediate and newbie affiliates.
I just started on sweeps and pops also. Seeing figures like yours is really motivating, as it's much more relatable than $x,xxx / day at this early stage.
All the best on your journey and quitting your job!
PS I've been thinking about spy tool blocking as well. Haven't done much research yet but I imagine you'd need to tie the page scrapped by the spy tool back to an IP then add rules to redirect. Could add an AJAX call after page load that sends a unique ID back to your server, then match it to one in the scraped lander. I tried to get IP data out of
Originally Posted by simon_89
Going from 100$/day to 500$/day in 5 months is not a goal, its procrastination.
Test a new traffic source every day. Test a new geo every 2-3 days.
Scale the shit out of the traffic sources which works. Multiple camps. Multiple accounts. Different camps with high bids for placements which are on fire. Camps for low performer with low bids.
Don't get caught up with ROI and overoptimizing funnel, work the absolute numbers instead. Optimize on the fly by.
Awesome Simon. But honestly I've realized going from 50 to 200+ is sometimes as easy as increasing the CPM/Bids additional 23-40%, of course given the right offer and traffic source.
Originally Posted by sebastian_r
Good luck!
Originally Posted by sebastian_r
Originally Posted by webdev
Originally Posted by simon_89
Originally Posted by simon_89
Originally Posted by simon_89
Originally Posted by sebastian_r
1) Setting up multiple campaigns at different bid points to drive out the competition. If you're doing this I'm assuming that you're not using the same lander at different bid points or are you?
In terms of camps with low performers, are you referring to those campaigns that are generating only $5-15 profit everyday?
When it comes to bidding, I always try to take the number one spot after narrowing down my targeting.
Nice thread! updates?
Yea! Updates! Unfortunately (actually fortunately!!!) Success in affiliate mktg is not linear like you are aiming for... it is exponential.
It is common to go from $40/day profit to $600+ per day profit....your working on leveraged platforms.
So what does this mean?
Basically, dont be discouraged if you dont make money for months. One campaign could hit and make you 100k the next month.
Here's an update on my campaigns:
Taking Sebastian's advice, I was trying to work with one new traffic source every 2-3 days and every 2 days I was working with a new GEO. Between 3/21 and now, I've been on scattering on my campaigns trying to find something that works. Here are some of the things I've learned so far:
1) Don't start a new campaign while you're asleep. I set a new campaign with a budget of $50 dollars and because I targeted USA, the flow of traffic was so massive that the system couldn't stop my spending cap in time. I spent about $60-70 extra on top of the $50. This was a RON campaign, with 1-2 placements draining my budget. Lesson Learned.........
2) You know yourself better than anyone else. I was trying to implement Sebastian's advice, but I was just all over the place when I took his advice. I'm not saying his advice is bad, but for me I like to nail down one traffic source. I am currently running under one traffic source that is spending more than my own good. I'm targeting 4 countries in that traffic source. Since, the inception of running these new campaigns(4/6) I've been solely focused on making this traffic source work for me. This traffic source(hint: Not PopAds: Although I will revisit that source later) has given me good lead quality(according to AM) and has seen some good green. I have to optimize my campaigns within this traffic source meaning narrowing it down to the best placements and targeting only those placements to be your top revenue earners. Once that's done, then I can put that traffic source on the backburner and try other sources.
Understanding what to cut:
Working with this traffic source has made me developed a much more strategic way of blacklisting bad targeting/placement options.
1) Blacklist placements that spend 1.5x payout without any conversions. In this initial test when I started my campaigns, I did three rounds of just blacklisting placements first. Each round was focused on blacklisting placements only, however if I saw any OS or Browsers doing EXTREMELY poor then I will blacklist those as well. But, take in account whether or not a specific placement is making you drain that much money first.
2) After blacklisting placements, you can start to blacklist Browsers which I'm in that stage now. That will only be done once. But, the test will last a lot longer than blacklisting placements.
3) Start blacklisting OS after blacklisting Browsers, again this is even much longer test. You don't want to delete options that are generating most of the traffic for you. If you find yourself having to blacklist OS/Browsers that bring you a lot of traffic but no conversions, it's time to move on to another traffic source or GEO.
4) RESULTS: After all of this, you should have a OS that performs well along with a few profitable browsers along with a bunch of profitable placements. Then you could start a new campaign with those placements/targeting options and bid higher while leaving the original campaign in tact to seek out new placements that perform well.
By no means is my strategy the best, but it makes sense to me and I developed this thinking based off of many campaigns I've ran plus with the advice that people at some STM members gave to me.
So cutting should be in the heirarchy of this:
Placements ----- then browsers ------ then OS -------= PROFIT.......
Here's this traffic source campaign progress:
4/6: The Phase of Blacklisting Placements
4/7: Still Blacklisting Placements
4/8: Done Blacklisting Placements, Started Blacklisting Browsers. Got rid of the UK campaign because there wasn't significant amount of conversions for me to keep going with this campaign. Spending too much without much conversions.
Stay tuned for the next update.........
Originally Posted by johnaff
It's much more relatable to me more so right now plus this image as well
Originally Posted by simon_89
Taking Sebastian's advice, I was trying to work with one new traffic source every 2-3 days and every 2 days I was working with a new GEO. Between 3/21 and now, I've been on scattering on my campaigns trying to find something that works. Here are some of the things I've learned so far:
1) Don't start a new campaign while you're asleep. I set a new campaign with a budget of $50 dollars and because I targeted USA, the flow of traffic was so massive that the system couldn't stop my spending cap in time. I spent about $60-70 extra on top of the $50. This was a RON campaign, with 1-2 placements draining my budget. Lesson Learned.........
2) You know yourself better than anyone else. I was trying to implement Sebastian's advice, but I was just all over the place when I took his advice. I'm not saying his advice is bad, but for me I like to nail down one traffic source. I am currently running under one traffic source that is spending more than my own good. I'm targeting 4 countries in that traffic source. Since, the inception of running these new campaigns(4/6) I've been solely focused on making this traffic source work for me. This traffic source(hint: Not PopAds: Although I will revisit that source later) has given me good lead quality(according to AM) and has seen some good green. I have to optimize my campaigns within this traffic source meaning narrowing it down to the best placements and targeting only those placements to be your top revenue earners. Once that's done, then I can put that traffic source on the backburner and try other sources.
Understanding what to cut:
Working with this traffic source has made me developed a much more strategic way of blacklisting bad targeting/placement options.
1) Blacklist placements that spend 1.5x payout without any conversions. In this initial test when I started my campaigns, I did three rounds of just blacklisting placements first. Each round was focused on blacklisting placements only, however if I saw any OS or Browsers doing EXTREMELY poor then I will blacklist those as well. But, take in account whether or not a specific placement is making you drain that much money first.
2) After blacklisting placements, you can start to blacklist Browsers which I'm in that stage now. That will only be done once. But, the test will last a lot longer than blacklisting placements.
3) Start blacklisting OS after blacklisting Browsers, again this is even much longer test. You don't want to delete options that are generating most of the traffic for you. If you find yourself having to blacklist OS/Browsers that bring you a lot of traffic but no conversions, it's time to move on to another traffic source or GEO.
4) RESULTS: After all of this, you should have a OS that performs well along with a few profitable browsers along with a bunch of profitable placements. Then you could start a new campaign with those placements/targeting options and bid higher while leaving the original campaign in tact to seek out new placements that perform well.
By no means is my strategy the best, but it makes sense to me and I developed this thinking based off of many campaigns I've ran plus with the advice that people at some STM members gave to me.
So cutting should be in the heirarchy of this:
Placements ----- then browsers ------ then OS -------= PROFIT.......
Here's this traffic source campaign progress:
4/6: The Phase of Blacklisting Placements
4/7: Still Blacklisting Placements
4/8: Done Blacklisting Placements, Started Blacklisting Browsers. Got rid of the UK campaign because there wasn't significant amount of conversions for me to keep going with this campaign. Spending too much without much conversions.
Stay tuned for the next update.........
Anywyas am curious to know why you cut browser than OS? I have an optimization process too and i do OS then browser but honestly have no good reason (except it is easier to drill down on OS then analyze browser ) so just seeing your thoughts.
Originally Posted by sushiparlour
Here's a way of thinking about it:
You can have one OS using different browsers. For example:
Mac Operating System could use:
Safari Browser
Chrome Browser
Firefox Browser
Internet Explorer Browser
Windows 7 Operating System could also use the same kind of browsers as well.
Now notice if you take off a bunch of OS systems that nest several browsers, then you're taking a huge chunk of your traffic away. However, if you just blacklist browsers instead of OS, you'll reduce the amount of traffic you'll take away. But, then again if you started to test the OS/Browser after you BLACKLISTED THE PLACEMENTS, and if the OS/Browser combination is still losing you money then it should be cut or funnel could be improved. The thing about the funnel part is usually it should be set in stone before cutting any placements/os/browsers.
Hope my explanation clears up your question.
Originally Posted by simon_89
Nice! Another follow-along by Simon! (How did I ever miss this thread??)
I really like your "cutting hierarchy" strategy! However, before cutting each particular "thing" - whatever it may be - be sure to drill down on this "thing" to see whether it has profitable segments that can be salvaged. Example: Say you're wanting to cut a browser, you'd want to drill down to browser -> OS, browser -> landers etc. One of the landers may be so perfectly suited for that browser, that it has converted really well, although the overall stats for the browser may be red. In other words, don't throw the baby out with the bathwater!
Also, I would add another element to your cutting hierarchy if you're running mobile: ISPs/carriers. Carrier traffic can convert extremely well with the right offers. A placement that is in loss overall, may be converting very well on carrier traffic. If you just cut that placement based on overall stats, you'd be throwing out the baby with the water again.
Sebastian knows his stuff - good for you on the massive testing! Yes you lost some money, but the experience you got is worth a lot more. You're a highly analytical person which can be good, but that quality will make you susceptible to over-optimizing, a problem sebastian has mentioned. Optimizing your funnel IS very important, but be aware that it's a process of increasingly diminishing returns - the increase in ROI tends to drop after the initial few tests/tweaks/cuts. So if your budget allows for it, try to focus more on scaling instead of optimizing once you hit consistent profits at high-enough ROI. That's what will increase your overall revenue in leaps and bounds.
Thanks for the exciting journey!
Amy
Originally Posted by vortex
Aw thank you for the validation that I'm doing my job right. You're a massive action-taker and and an analytical thinker, and most of all, you were persistent in testing until you saw green. These are the characteristics that have led you to your first green, and will do so time and time again.
I really like your "cutting hierarchy" strategy! However, before cutting each particular "thing" - whatever it may be - be sure to drill down on this "thing" to see whether it has profitable segments that can be salvaged. Example: Say you're wanting to cut a browser, you'd want to drill down to browser -> OS, browser -> landers etc. One of the landers may be so perfectly suited for that browser, that it has converted really well, although the overall stats for the browser may be red. In other words, don't throw the baby out with the bathwater!
Also, I would add another element to your cutting hierarchy if you're running mobile: ISPs/carriers. Carrier traffic can convert extremely well with the right offers. A placement that is in loss overall, may be converting very well on carrier traffic. If you just cut that placement based on overall stats, you'd be throwing out the baby with the water again.
Sebastian knows his stuff - good for you on the massive testing! Yes you lost some money, but the experience you got is worth a lot more. You're a highly analytical person which can be good, but that quality will make you susceptible to over-optimizing, a problem sebastian has mentioned. Optimizing your funnel IS very important, but be aware that it's a process of increasingly diminishing returns - the increase in ROI tends to drop after the initial few tests/tweaks/cuts. So if your budget allows for it, try to focus more on scaling instead of optimizing once you hit consistent profits at high-enough ROI. That's what will increase your overall revenue in leaps and bounds.
Thanks for the exciting journey!
Amy
I find that it makes much more sense to go by that hierarchy. I read constantly that pop networks are inflated with bot traffic and bad placements in general. So it's important for me to cut out the bad stuff in the beginning.
In terms of cutting placements, it really depends on the GEO I'm working. For example, I'm working with a campaign right now in the USA that has more than 10+ pages of placements, and I had to spend quite a time blacklisting placements that have reached 1.5x payout without conversions or very minimal conversions. But out of this test, I spent around $1,000 to find placements that could be good. I take the "good" placements and then target those individually.... Usually those placements range in 4+ conversions and then I whitelist them, but this also have to depend on the costs as well. If the costs come very near to the revenue, I'll let it run.
At this point, once I reached a campaign that I feel like I have taken all the good placements from I pause that campaign and only run whitelisted campaigns, Again, this is to implement Sebastian's thinking. It's not concentrating on the blacklisted campaign too much since I have found the placements that seem to be already doing well.
As of progress right now: Almost done blacklisting placements, then starting to work on whitelisted campaigns and then move them over to a new traffic source.
I'm still losing a bunch of money, but every day it seems like I have the mindset that I'm so close to my goal.
4/13
Revenue: $370.45
Costs: $842.90
Profit: ($472.45)
4/14
Revenue: $435.30
Costs: $632.27
Profit: ($196.97)
4/15
Revenue: $394.81
Costs: $892.44
Profit: ($497.63)
These colors aren't pretty, but it is what it is. On these dates, I was testing more traffic sources. I have 4 GEOS that I'm working with in my main traffic source. I took one of the GEOS that I'm working with and creating different campaigns across various ad networks. I'm basically throwing shit at the wall to see if there's any traction and if there is I'll proceed to work with that traffic source.
The main traffic source I'm working. I took all the whitelisted placements and just started campaigns on those placements and left the backburner campaign in the dark. The backburner campaign is the RON campaign that I initially use to test. I looked at all the placements that were profitable to me and didn't bother working with placements in the backburner campaign only generating 1-3 conversions over the span of 4-6 days. I rather target placements that have generated 7+ conversions over the span of 4-6 days.
I like those spends.
You make progress quick with high daily spends. Even when its red.
Its only a matter of surviving long enough, testing lots of geo - offer - lander - traffic source - combos, and interpreting your data correctly.
Here's what you can try to get your camps into the green:
Don't overspend, its easy to waste money with high bids. Map your network and try to hit the marked point (those are of course examples, but most of the time the bidding curve looks similar to this).
Black vs. White
On all my camps on mobile / desktop I had not one whitelist working really well, even the tightest blacklist always outperformed my whitelist camps. The networks often reward you when you don't have the finger on your spend. So when I say run several camps, I mean several blacklists.
For example:
Blacklist 1: EPV > 0.0015 CPM 1
Blacklist 2: EPV > 0.004 CPM 2.5
If a placement gets good traffic at Blacklist 1 and is between EPV 0.0015 and 0.004 it stays there. If a placement has an EPV > 0.004 and does not get good amount of traffic it goes to blacklist 2. All placements which get good traffic on Blacklist 1 get blacklisted on Blacklist 2 to avoid overpaying.
BUT it depends on the traffic source if this approach works. Another good one is just running 2 camps with the same targeting and different bids to get more traffic but pay less for the second impression (works well on adcash, propeller).
All of this makes only sense if your bread and butter camp is green.
Offer > Traffic Source > Placements > Creatives
It seems your spending an decent amount of time on figuring out your traffic source and your placements / targeting. Thats good.
The traffic source has an huge leverage.
But your offer has an even higher leverage. It all starts and ends with the offer. I bet you can make your camp profitable just by running an better offer. Spend more time testing different offers on different aff networks.
To clarify my first post: This was advice for scaling your camp when it was green. By now when I hit a winner, within one day it runs on 10 traffic sources. Then new geos day by day. BUT if you don't have an winner, that makes of course no sense. As said, focus on finding an winning offer that works well with your go to source. Then go back to scaling. You will be surprised what an difference the offer can make.
Holy shit Simon! That's a lot of spend and a lot of testing! And sebastian - if that isn't the longest post I've seen from you! Simon you're really blessed. Sebastian knows a lot about pop traffic but most of his posts are short and sweet. He basically came and dropped a gold bar on your thread lol!
Spending and testing a lot is great, but when you get into mass-testing mode you need to be efficient or you'll spend all your savings before you hit on your next profitable camp. Some of what I suggest below will overlap with sebastian's but that's OK - I feel that some points can't be emphasized enough, especially when you're spending so much money on testing!
~~~~~~~~~~~~~~~~~~~~~~
-TIP regarding vertical+geo selection: Look at afflow/monetizer's stats to see which vertical+geo combos are hot right now and start with those.
Alternatively, pick a vertical you know is working for a ton of geos (e.g. sweeps, antivirus) and just mass-test geos. This way you can really get to know about the vertical (what types of landers work etc.)
~~~~~~~~~~~~~~~~~~~~~~
-Yes, it IS true that every offer+lander combination will perform differently on different TSs. But as Sebastian has pointed out, it's not necessary to test a not-yet-profitable campaign on all sources.
The way I see it: Although each traffic source has different placements that belong to different categories, pop traffic is broad traffic so we're only wanting to push broad-appeal offers anyway, so we can sort of assume the nature of the traffic as being similar across pop sources. Traffic quality and bidding landscapes will differ as well. But none of these differences will matter much when your offer+lander don't convert - cause they'd just suck across the board.
What I usually do is set up initial tests on 1-2 go-to sources I've had the most success with in the past, and go from there. Once you know which offer + lander + targeting converts well, there's a good chance that the combo will convert well on other sources as well. Not always - because other TSs may have more bot traffic and/or more competition so the ROI may not work out like on your original TS. But if you know Offer1 converts well with CarrierX for Android phones, there's a good chance you'll see the same trend on other TSs also.
When choosing a TS (or 2) for initial testing, other considerations are traffic volume and targeting options. I'd pick a TS that has enough traffic volume for the given geo. And if I'm testing carrier-billing offers, I would stay away from sources that don't have carrier targeting.
~~~~~~~~~~~~~~~~~~~~~~
-Agree with sebastian on the "Offer > Traffic Source > Placements > Creatives" hierarchy. When testing a specific geo+vertical, I usually start with a few recommended offers (by AMs etc.), rip 5+ landers, test 3 bids at 10-20x payout each, then pause all but the best-performing camp to continue the testing. To avoid over-spending, I would add $10-20 to the camp every time I check stats. As soon as I feel that the camp isn't promising enough, I'd cut my losses. When you're testing a lot of stuff at the same time, you can abandon a camp without a lot of justification - it's more like you're identifying the low-hanging fruit and only optimizing and scaling those.
To gauge whether a camp has promise or not - I like to see at LEAST one offer+lander combo in the green with 2+ leads. This doesn't have to be the "overall" stats - just needs to be true for at least one major traffic segment (e.g. major OS, wifi, a mobile carrier that has good traffic volume). (I'd also take a quick look at placement stats just to make sure there's not a couple of non-converting placements responsible for a large portion of the ad spend).
Regarding landers, my reasoning is that if everyone's using the same landers, then at least some of them are making money for some people. So if I rip the most popular landers and use them to test several proven offers for the geo+vertical, on my go-to traffic source, and STILL don't see potential - then I'd rather test something else.
For a campaign that shows promise: Once I'm down to a winning lander, I'd mass-test offers next - and if there's enough traffic volume, I'll do granular testing as documented here. Of course granular testing doesn't just apply to carrier traffic. The point is to find the best offer for each major traffic segment (OS/ISP/Carrier).
The exception to the "low-hanging fruit" approach, is if you KNOW there's significant potential. For example if your friend told you they're making a killing with a certain offer, or if you find out there's a ton of carrier/cellular traffic for a certain geo on a certain TS. For cases like these, I would spend a good amount of money and effort to optimize them, and test on more TSs.
~~~~~~~~~~~~~~~~~~~~~~
-The more I work with pop traffic, the more I agree with what sebastian said about whitelists not working very well in most cases.
Another reason I don't mess with placements too much anymore, is because I find it a major time suck that isn't worth the time. I'm looking into automating the process: have a script/software created, that could grab all the placement stats, use the placement-cutting calculator spreadsheet to identify the bad ones, and automatically pause them on the TS via API. Until that happens, I'm spending less time checking and cutting placements, and more time on testing new stuff and scaling profitable stuff.
~~~~~~~~~~~~~~~~~~~~~~
-Something I've started playing with is caurmen's bot traffic script:
http://stmforum.com/forum/showthread...367#post180367
http://stmforum.com/forum/showthread...n-Any-Campaign
It takes SO MUCH LESS money to kill bot-infested placements with this script rather than using the calculator spreadsheet! It only takes 30-50 clicks from a placement to tell whether it's mostly-human or mostly-bots.
~~~~~~~~~~~~~~~~~~~~~~
Exciting times ahead!
Amy
Damnit Simon, you squeezed some golden eggs out of these geese.
If I can't make some progress based off their suggestions, I'm packing my bags lol jk.... Still going at it with their tips embedded in my mind.
Hey! Looking forward for you finding a winner offer! Good luck!
Home > POP / PPV / Redirect > Follow-along Campaigns