First Conversion, Let’s Get More!! (My Pop Follow Along) (22)
First Conversion, Let’s Get More!! (My Pop Follow Along)
Hi STM,
Paid affiliate marketing is new for me. I joined the forum in 2016 for a month after reading a blog who said the only 2 ways for a young person to build great transferable skills is to learn copywriting and affiliate marketing.
At the time I was in college and didn’t have the money to devote to learning AM, so I left the forum and picked copywriting to get my feet wet, and then got into building authority sites via SEO.
Now, I can actually give AM the time it deserves and buy data.
I’m only saying this because if you’re new and following along I hope this FA will help you the most. I’m going through a lot of newbie issues now and I’m going to do my best to provide details of my exact problems and how I’m solving them.
Think of this as a campaign journal, but without exact details of offers and what not.
I know I have blind spots so please comment and give feedback if you’re seeing something which isn’t adding up.
Let’s get started...
Quick background so everyone is caught up to speed:
1) I went through Amy’s 40-day tutorial and executed on all the action items she added. I still need to learn how to prune campaigns and get them ROI positive, which comes via blacklisting.
2) I bought a domain which was already burned by Google. Lesson learned to check before purchasing lol.
3) I sent direct traffic to a couple mobile games which did nothing for me besides give me data.
4) I bought adplexity and ripped landers for a random phone offer I found on
5) I used 1 very bad and kinda jacked up lander to send traffic to the phone offer which got me conversions, but was direct linking instead of using my click URL. ROOKIE MISTAKE - that was frustrating seeing no conversions on
6) I hired a guy on Upwork to fix the landers I’m ripping via adplexity and comparing the code between the different landers to understand better what he’s doing. (My shortcut of learning code instead of watching tutorials - I learn better this way)
Current state:
1) Laptop offer SOI in a tier 1 country (not U.S.) with a $2.45 payout on info submit.
2) Testing 5 fixed landing pages (Spinner, box, confetti, and 2 static pages) for this 1 offer.
3) 2 different traffic sources. Propeller and Popads.
4) $2.4 CPM for Propeller
5) $2.1 CPM for Popads
6) Campaign started at 4am local time
Future State:
1) Let the laptop campaigns run until 10x the payout spend on each platform to evaluate and learn.
2) Fix up the landers for the above phone offer.
3) Get translations because the phone offer is in a different Geo.
I’ll add my results here!
Hey there.
When then a LP domain gets flagged your tracking domain won´t be affected.
For landing pages I just buy several cheap domains at once and prepare them for use.
When a domain gets flagged I have the next one ready to use.
When you don´t receive any conversions the data isn´t really worth it because then you don´t if placements are bad or if it´s the offer.
With a good working funnel you can get good data because then you know that the funnel itself is converting and can work much better with the placement stats.
There are so many offers out there, it´s better to run offers that already have shown good performance or offers that show potential.
When you can´t get help from a network or can´t reach them move on to the next one.
Regarding Clickdealer, also try to send them a PM here on STM.
They are basically ready to use when you enter your links there.
Maybe this can help you already and then you don´t need to hire someone from Upwork.
They are very competitive and have mostly high bids.
It´s better to learn in a more beginner friendly environment
You can still move to such geos when you know how to run campaigns.
That way you can reduce the risk to lose too much money in the beginning.
The technical progress how to run campaigns and the steps needed are more or less the same in all geos but learning in low tier geos is cheaper.
Hi Twinaxe - thanks for the reply. Appreciate the guidance on this. Hopefully this helps not only myself but others!
When then a LP domain gets flagged your tracking domain won´t be affected.
For landing pages I just buy several cheap domains at once and prepare them for use.
When a domain gets flagged I have the next one ready to use
When you don´t receive any conversions the data isn´t really worth it because then you don´t if placements are bad or if it´s the offer.
With a good working funnel you can get good data because then you know that the funnel itself is converting and can work much better with the placement stats.
There are so many offers out there, it´s better to run offers that already have shown good performance or offers that show potential.
When you can´t get help from a network or can´t reach them move on to the next one.
Regarding Clickdealer, also try to send them a PM here on STM.
They are basically ready to use when you enter your links there.
Maybe this can help you already and then you don´t need to hire someone from Upwork.
They are very competitive and have mostly high bids.
It´s better to learn in a more beginner friendly environment
Update on campaigns and current state:
So I learned today the laptop campaign was a bit of a dud. I got only 1 conversion from an $20 spend based on the CPM I've set. With Propeller I'm noticing the traffic is coming in such a low amount that I'm assuming my CPM bid is not enough to get traffic. Popads has double the traffic it's pulling in, so I'm going to turn off the spend because I think the offer is bad. First to Twinaxe's point I didn't get it from my affiliate manager, so I was flying blind, but second I pulled the landing pages from Adplexity which means they've ben vetted.
Onto the next offer!
I got the landing pages for the phone offer fixed up and ready to go with the correct click URL link. I know the offer has potential because I got conversions with a bad lander. I'm going to test it was 5 different landers I pulled which have potential. We'll see how they go.
Future State:
1) Pull the campaign for the laptop offer.
2) Get into more affiliate networks so I can get tier 3&4 geos.
3) Analyze the phone campaign tomorrow.
4) Start looking for a new offer to expand!
If clickdealer won't take you, just submit application to other networks, there is a ton of them out there.
Check our contacts directory : https://stmforum.com/affiliate-directory/
Use our network application tool: https://stmforum.com/central-vendor-app/
Try to enroll to a few more that are popular here on the forum: gotzha, advidi, glize, mobipium, maxbounty, leadbit... some will take you for sure.
Don't let a rejection from clickdealer to slow you down
Back for an update!
If clickdealer won't take you, just submit application to other networks, there is a ton of them out there.
Check our contacts directory : https://stmforum.com/affiliate-directory/
Use our network application tool: https://stmforum.com/central-vendor-app/
Try to enroll to a few more that are popular here on the forum: gotzha, advidi, glize, mobipium, maxbounty, leadbit... some will take you for sure.
Don't let a rejection from clickdealer to slow you down
Having said that here's what happened yesterday:
1) Once I got approved by CD I spoke with my AM about SOI offers for sweep which has shown success in tier 3&4 geos. He gave me a list of offers and I got approved for a phone offer in TH at $.24 CPA.
2) I updated 3 landing pages to their language.
3) Launched the offer last night while going to bed at $1.6 CPM. (Pro Tip - with Propeller it's a $30 minimum daily campaign budget. I made the mistake of letting the campaign overrun and I woke up to $29 spent. I thought I would wake up to it only spending $8-$12 in budget. Lesson learned is don't run a campaign at night unless you can keep budget tight for testing.)
4) Results are below:
As you can see - a little disheartening because my AM said this offer was doing well. I guess good news is I saw one of my landers performed way better than the others.
Future State:
1) Kill the TH offer since it doesn't fit the 10x conversion rule.
2) I'm testing another SOI offer in MY with the winning lander from the previous campaign.
2) Using Popads instead of Propeller because I can set a reasonable budget of 2x payout to see if the offer works.
Side note...
I can see what Amy means when talking about catch-22 issues. I definitely feel a little lost because the offer was supposedly solid, and I pulled the landing page from Adplexity. I think the only course of action is to continue testing offers but the only other ones given by my AM are credit card submit offers with payouts of $24 instead of $.24 which means budget goes up to get an appropriate test.
Guess I'll need to ask for more working offers in those tier 3/4 geos.
1. Just because a LP is in adplexity doesn't mean it is performing well. There are MANY failed landers in rotation too and aplexity pulls all of them, not just the good ones, the scripts don't have access to campaign data. It can also be a cloaked lander that fooled adplexity, so it grabbed the "Default" one, but in reality, the affiliate might be pushing a more aggressive/non-compliant one, hence the cloaking.
Adplexity, or any other spytool, is a great source of landers,creatives, ads... but it's a mix of both good and poor ones.
On top of that, you know nothing about the targeting an affilaite is using. The creatives can work well with specific audience, but might fail with a broad one.
To put it simple, do not assume that everything listed in the spytools is solid, you need to test it to see whether you grabbed a winner or a loser.
2. AMs see "mixed" data, they don't know everything about the traffic. So it might be a similar situation to the LPs I described above. The offer might work well with certain traffic type, targeting, ads... but unless you're able to recreate the same setup, it might fail on you. So again, you need to test the offers with your current setup and decide whether it's worth it or not. That's why you need access to a LARGE amount of offers, you need to find those that work the best with the traffic you have access to.
Obviously, chances are that an offer that's generally performing well for other affiliates in the network should be a solid one, so it's always a great idea to start testing offer that the AMs see good performance for.
It's just not a rule, what works for others, might not work for you, from whatever reason.
Hi Matuloo - thank you for the advice. The more I get into AM the more I learn there's no silver bullet. It's all about testing and data.
Obviously, chances are that an offer that's generally performing well for other affiliates in the network should be a solid one, so it's always a great idea to start testing offer that the AMs see good performance for.
It's just not a rule, what works for others, might not work for you, from whatever reason.
Current State:
Like I said I'm running 2 unique offers given by my AM. I've been running them for the past couple days and I'm at a -60% ROI. Granted, this isn't profitable but I found a lander which is doing really well in those countries and I'm going to start split testing offers. Below is the data for the past couple days:
I'm in the optimization phase so I've started blacklisting zones 2x in spend to get the good placements. Working to get it into the green. Once I get it into the green on this traffic source I'll scale into another one.
Here's another campaign:
Future State:
1) Now I have a list of 50 offers working in the tier 3/4 geos. Time to start testing and getting those offers into the mix.
2) Goal is to launch 5 new offers in the next 48 hours. I'm on a vacation so connection is limited but I can devote the time.
Just make sure to change images and text depending on the offer.
Otherwise you risk to cut good placements because the funnel isn´t strong enough.
When you start to optimize on too weak funnels you often have to cut too many placements to get the campaign into green so that you are left with too low volume to be worth it.
Of course try to get that campaign into green but when you then see that traffic volume drops too much better go back in test mode, find a better offer and then test the excluded placements again on a stronger funnel.
There you can test other bids, wider targeting, other connections or devices.
Update:
It took much longer than anticipated with getting new landers and offers uploaded. Here's the deal, I re-read through @jaybot's FA thread and got help from @twinaxe who advised I'm failing because I'm not testing enough offers. My spend so far has only been $200 and I've tested a total of 5 offers. Major mistake!!
So... I went back to the drawing board and got SOI offers on the Clickdealer network for testing. I got them uploaded last night after pulling an all nighter to start the campaigns. It took a couple days to do the lander build and translations. I got 10 offers up last night with a $10/daily budget. I'll be getting another 10 up within the next day or 2.
Here's the results from the initial test:
Now it's back to the grind of getting more offers up and continued testing!
After a day of running here are the results from the offer testing I'm performing. 10 offers in different geo's with the translated landing pages.
As you can see only 1 of them is producing any green, but the good news is this is my first green campaign after 24 hours of running. So my first win!
The hard part and where I'm getting stuck is in traffic volume. I set the budget for each campaign to $10/day but there's not much traffic. I'm assuming my bid price is too low causing me to not place for quality spots, but I'll defer to the experts on this. Maybe the volume is doing fine?
5 of the campaigns haven't gotten a single conversion yet, with only 2 of those 5 reaching the 10x point. I'm going to pause those campaigns completely since they're not meeting the 10x rule.
Next Step
1) Add more offers into the rotation for continued testing.
Sometimes I feel like Moses when it comes to AM. Making the sea RED with blood... lol
Lessons learned from the day above.
1) My bids were far too low. The traffic was trickling in and I wasn't getting any quality placements.
2) After split testing a couple different landers there's 1 which seems to be working for the Asia geo's which wins every time. Jury is still out for LATAM regions.
3) Offers, Offers, Offers. Never enough offers lol. I have 12 running now and doing another 3 today for a total of 15 running.
I'll post my results from the bid increase tomorrow morning. It's already working though, I'm getting a good deal of traffic. I should hit my daily budgets finally.
Update from yesterday's performance
Couple of quick thoughts
- Spend went up today and I'm getting quality placements now. Unsurprisingly so did the conversion rate from the previous days. I also added another offer into the mix but I got caught with my day job.
- I'm at 13 total offers and now I need to cut 2 of them which aren't performing at all. I haven't gotten a single conversion on 10x spend, which puts me back to 11. I'll need to add another 4-5 new offers to the mix.
- Questions around statistical significance. I had a test of 2 landers where LP1 is doing well for the Asia regions and I tested another lander against it. How many conversions do I need to wait for before deciding to kill a lander and move on. Also, what happens if there's a 50-50 split?
- Questions around optimization of the campaign. I'm assuming this is the secret sauce of affiliate marketing so people won't be too willing to reveal best practices, but during the test phase of offers do you blacklist placements only and wait and see or is there another method?
Originally Posted by painnoworregretlater
Couple of quick thoughts
- Spend went up today and I'm getting quality placements now. Unsurprisingly so did the conversion rate from the previous days. I also added another offer into the mix but I got caught with my day job.
- I'm at 13 total offers and now I need to cut 2 of them which aren't performing at all. I haven't gotten a single conversion on 10x spend, which puts me back to 11. I'll need to add another 4-5 new offers to the mix.
- Questions around statistical significance. I had a test of 2 landers where LP1 is doing well for the Asia regions and I tested another lander against it. How many conversions do I need to wait for before deciding to kill a lander and move on. Also, what happens if there's a 50-50 split?
- Questions around optimization of the campaign. I'm assuming this is the secret sauce of affiliate marketing so people won't be too willing to reveal best practices, but during the test phase of offers do you blacklist placements only and wait and see or is there another method?
And every affiliate does it very differently.
For lander testing,
stat sig calculator here: https://marketing.dynamicyield.com/bayesian-calculator/
But. If you wait for statistical significance on everything, you will go broke. Especially if you do it in Tier 1/2 geos.
In the beginning, it’s sometimes best to cut aggressively. Instead of waiting for 2x payout for every placement... If the geo you’re testing in has enough volume, feel free to blacklist placements at 1/2 or 1/3 payout. If the geo doesn’t have a lot of volume, cut it at 1x.
That’s what I do.
Wait for twinaxe to chime in with what he does. It may be totally different
Originally Posted by jaybot
Originally Posted by painnoworregretlater
If you cut so many that you run out of volume, drop the offer and try again.
In some geos (IN, US, MY, DE, ID, etc) you can cut thousands of placements (over days, weeks, months) and still have thousands left over.
In others (NZ, CY, AU, etc) you’ll quickly run out of placements and notice that you ain’t got no traffic no more.
Originally Posted by jaybot
Well, the first thing is that you need to really understand what statistical significance is and when it makes sense to use it.
It basically calculates based on the number of visits and number of conversions how likely it is for the different elements in the test to win against the other elements.
The calculator will show you then by what % probability the differents elements in the test will win.
You need to collect some data first to reach statistical significance, this means that you have to reach a point where the calculation has enough data to be confident enough that the results won´t change anymore so that it can show you a clear winner.
Maybe it sounds a bit complicated but you don´t have to mess around with it yourself, that´s what the calculators are there for.
We just feed them with few numbers and they spit out a result that is easy to read.
But now we come to a very important point.
It doesn´t make sense to wait for statistical significance for all your campaigns, it only makes sense to wait for it when the campaign is worth it.
When all elements (offers or landers for example) in the test are running on loss and you keep them running until you reach statistical significance you risk to lose more and more money just to find out what´s the least bad offer/lander.
Even when an element reaches 100% probability to be the best in the calculator it doesn´t mean that it´s a real winner.
It only says that this element is better than the others.
As an example, when you run 4 offers with $0.30 payout it and they all convert here and there but they all are running at -80% ROI then it doesn´t make sense to spend $200 on these offers just to find out that one of them is not that bad as the others.
Set a test budget and try to reach statistical significance with this budget.
When nothing is converting good enough to continue move on and don´t waste your money waiting for statistical significance.
Long story short, only wait for statisctical significance when it´s worth it.
For landers and offers statistical significance makes sense, for placements I also work with "x number of payouts".
As jaybot says, when the geo is big enough and has enough traffic you can cut more aggressive, when volume is low anyway it´s better to keep it running a bit longer.
You should take care to don´t make the mistake to "optimize" your campaign to death.
When you see that specific placements send high volume but they don´t convert or don´t convert good enough block them for the moment.
There you don´t need a statistical significance calculator.
The calculator can be used however with the landers or offers in the campaign and this is completely independent of how many placements you cut.
It really depends on where you run and so on.
When you run a huge geo with lots of traffic you can cut 100 placements and still have enough traffic available.
In small geos you would have no traffic left when you cut 100 placements.
Update:
Yesterday I think I found 1 campaign worth pursuing. It was sitting at around -50% ROI (obviously this isn't great, but it's the best campaign I can find so far in the testing batch.)
Goals for today:
-Begin to cut placements and get everything out of unprofitable.
@twinaxe - thank you for all the great quality information I really appreciate the time you're taking on these.
Couple quick questions:
Set a test budget and try to reach statistical significance with this budget.
When nothing is converting good enough to continue move on and don´t waste your money waiting for statistical significance.
Originally Posted by painnoworregretlater
For example, your promising camp above looks like a $0.06 payout. So the 10x payout rule for that offer would be, what, $0.60?
It would be tough to make a call if a camp is promising or not at $0.60 , especially in Tier 3/4 geos with loads of traffic.
On the flip side, you have CPS for $50 and crypto offers with $500... I don't care who you are, I'm not spending $5000 to check for 1 conversion.
But to answer your question: I would rarely bother with anything less than -70% ROI. You have a steep hill to climb and it's often more efficient to just test another offer.
Something else to consider is what the traffic distribution is like across your placements. For example if you have one or more big placements that are zapping up a big percentage of your budget without converting, you may be able to increase your ROI massively just by blacklisting them.
And if you still have healthy traffic volume after that, then you can continue cutting placements to reach green.
Rules of thumb are like crutches for beginners - they can be a good guide when you don't yet have experience to go on. After you've run a few campaigns, I would encourage you to let go of those crutches and start to question and analyze everything yourself - as rules of thumb are seldom an efficient solution for all scenarios.
Amy
Hey there
Originally Posted by mantas
Home > POP / PPV / Redirect > Follow-along Campaigns