Home > > Newbie Follow-Alongs

---

Daring to Live the Dream: A Non-Techie Newbie Comes Out of the Dark (28)

---

Follow Along Entry #1

Greetings, fellow STMers! I’ve only been a lurker since joining STM a couple of weeks ago, but this community has been so supportive and amazing that I decided to jump in full throttle and do my own follow-along.

Some background: Something clicked in me last December when I came home from work one day just beyond exhausted and emotionally drained. I worked in the medical field and basically thought to myself that I did not get paid nearly enough for what I do, and it was time for a change. I wanted the freedom that we all on this forum seek, and I was ready to do whatever it takes to have that freedom. So here I am!

I signed up with my very first affiliate network back in March of this year, and have since been “dabbling” in AM. I signed up for Voluum and Adplexity, figured out how to run a few failing campaigns and felt like I was going in circles with no real direction. So I bit the bullet and joined STM, which I had read about when I first started this journey through Charles Ngo’s tutorials, but at the time I didn’t want to add the extra cost that early in the game (mistake #1!!!).

I would say I’m computer savvy, but definitely not technical as in I don’t know how to code – AT ALL. I have a pretty solid background in graphics design, some HTML and a little CSS from WAY back in the day and have slowly been relearning these things by going through Amy’s awesome Newbie Tutorial! I do want to learn enough coding to get by though.

I’ve spent the last couple of weeks reading and rereading Amy’s Newbie Tutorials and am now up to the point where I started running a campaign to narrow down my ripped landers. So here’s some info about this camp:

6 landers for same sweeps item x 3 offers in same GEO but different networks.

I had a hard time finding sweeps offers of the same item in the same GEO, even after applying to a few more networks. So I ended up choosing offers that I found myself that I thought might convert, since the offers my AMs suggested were either not in the same GEOs or the same sweeps items.

I first ran an initial bot test campaign on Propeller Ads with 1 lander and 1 offer in my selected GEO using Caurmen’s method, and blacklisted 11 Zone IDs that had more than 70% bots. I also set the bid to about 10 cents more than the recommended bid. I ran this under “All Mobile” and did not separate 3g and Wifi. Not sure why I didn’t, should I have?

After the bot test I removed the bot scripts by commenting them out on each LP, then reuploaded (is that a word?!) them. Not sure if that was the best way to do it, but I figure if I wanted to run the bot test again in a few days I could just remove the comment tags. I then set up one camp that rotated my 3 offers and 6 LPs with equal weight. In Propeller Ads I excluded my previously identified bot zones, and set the bid to about 30 cents more than the recommended, though I probably should have bid the same as the bot test, 10 cents over. Should I have? I set the traffic distribution to be over 24 hours to throttle my traffic, mostly because I needed to go to bed and most of my PA camps have spent my $ in <1 hour. Oh, I set my daily and campaign total amounts to $30, which was about 10x my payout of $1 for each offer. Then I realized I forgot that I also had to take my 6 LPs into account, and if I did that then I should have set the total budget to be $180 (6 LPs x 3 Offers x $1payout x10). Is that how I’m supposed to calculate it? It seemed like a lot of money to test unproven offers, so I just left it at $30 for now…also, at which point should I start split testing different bids?

So my 24 hours is almost up and I have only ONE conversion. The conversion didn’t show up in Voluum, but it was on the new network I had just joined so I fixed the postback and manually added the conversion to the tracker. Does anyone know how to manually add the revenue from that one conversion?

Here are some screenshots:

Total campaign stats:

Drilling down to the 3 offers:


Drilling down to the 6 LPs:


I’m really tempted to say that the lander with the highest CTR is the winner even if it had zero conversions!


Next steps: Even though, I only spent $30 and not $180, I should really try to find proven sweeps offers in the same GEO, and give up this camp. Should I just keep applying to more networks if I’m having a hard time finding offers?

Sorry that was so long, but thank you for reading and following along on this wonderful journey!

Wan

---

I am only on my phone online so I don't want to write too much for now but there is one thing: Switch from fixed CPM to Smart CPM bids.

---

Congrats on taking the leap man, its a big step done!

Regarding looking for some hot sweeps I'd try out anyone of the following networks - Gotzha, Mundo Media, Clickdealer, Response Concepts, Advidi, Wewe - all of these have pretty up to date sweeps doing decent volume.

Remember as you approach campaigns, the guides here are only just that... guides. You will develop your own standards for testing as you see what works best for you, its just a template to help people break the anxiety of knowing what 'can work' to get a campaign rolling

TBH I would probably pick one of the first 3 networks as they don't just carry a lot of sweeps but many other offers too which you can begin to grow with your affiliate manager there if need to branch out into different verticals.

Still early to tell, do what Twinaxe mentioned above and maybe let is spend some more to see what happens, remember its attempt 1 it might take a few... but hold onto the wheel its going to be a ride haha

---

Thanks for the tip!

Haha, I'm SO looking forward to a wild ride! Thanks for the network tips - I am on 2 of those you mentioned and will apply to the others. Doing that and looking for offers will be my focus today.

I also found this coding thread by Caurmen so will go through it as well:
https://stmforum.com/forum/showthrea...ing-Parameters

---

Following Along Entry #2

Ok, I found some proven offers recommended by my AM in my chosen GEO. Last night I started running a campaign rotating through 4 offers and 7 landers total, with a limit of 10x payout for each offer. I woke up to two conversions, both from one offer, and the same lander actually...WoooooHOOOOOO!

So I paused all other offers and am continuing to run the campaign. This offer was set up in a Voluum path with 4 landers, so that is what is running now. My question is, do I run until I reach statistical significance? Or until the I've reached the 10x payout? My payout for each offer is $2, so with 4 offers total, I set my budget to $80. I've run through about $60 when I hit the 2 conversions, so I then paused the nonconverting 3 offers, and restarted it. Do I reset the amount to spend to 10x payout again for the one offer? Run the rest of the $20 of the original camp? I guess I'm not sure what I should be spending once I cut things down to one offer.

Here are what my stats look like:

Entire campaign of 4 offers and 7 landers:

Drilled down to the 4 offers:

I should note that my paths with weights were like this:
Path 1: 1 Offer with 4 LPs (Weight 50%)
Path 2: 1 Offer with 1 LP (Weight 12.5%)
Path 3: 1 Offer with 2 LP (Weight 25%)
Path 4: 1 Offer with 1 LP (Weight 12.5%)
(I'm not sure if I should have set all the weights to 100% but it made more sense for me to distribute them depending on the # of landers each offer had...)


Drilled down to 4 LPs for the winning Offer:


I feel like I should take the converting LP and use that as the winning one at this point, but maybe it's too early?

---

Follow Along Entry #4

So I let the rest of the original funds run through (~$20) and finish, and I received 2 more conversions for a total of 4!

Here's the result from one offer and 4 landers:

It seems clear what the winning lander is, but should I keep running to see if that 4th lander with 562 clicks will ever convert?

---

Ahoy captain!

Did you swap to smartCPM like twinaxe suggested?

Lander 2 and 3 definitely suck man, the CTR is non existent so cut those two they're eating ya $$'s. TBH lander 1 is standing out a lot more then the others, but with sweeps on the low payout you can often see the epc rise day 2 and 3 of testing so you may have cut the others a little early but you're close to breaking even now.

I would just run lander 1 with the offer and see if it picks up to above break even, but remember its your first campaign and it can take a few to get a winner just to keep that in mind.

---

Greetings, Maynzie!

Yes, I switched to smartCPM, thanks to Twinaxe.

I am going to try to run the 1 lander and 1 offer, but as you said, I'm not expecting much at this point. I'm thinking of cloning the campaign and running the 1 lander with 4 other offers now too and see what those offers look like. My AM said the other offers that I cut are converting like crazy, even better than the one that I actually got conversions on, so I want to try them out again. Gonna get that set up tonight, and let them run while I'm in dreamland....dreaming of conversions swirling about in my head!

---

That sounds like a good idea to me man, haha yeah the life of an affiliate :P

Let it run for a day with similar budget just across lander 1 x 4 offers and post your results, looking forward to it mate

---

Hi Wan! It sounds like you're doing a LOT of things right - I hope you're as pleased with the progress as I am!

Sorry for arriving late - but being guided by the legendary Maynzie is the best thing that can happen to anybody! You're in the best of hands!

Now that I've posted in your thread, any new updates will appear on my feed of subscribed threads, so I won't be so slow to answer.


Below are my thoughts...

If depends on what types of traffic the offer accepts, AND more important, whether the offer has different conversion flows for 3g vs. wifi.

It's also important to know that 3g traffic is considerably more expensive than wifi traffic.

What you want is to adequately gauge potential of the offer without paying too much for traffic if possible.

If the conversion flow is the same for both 3g and wifi, then there's no reason to test the offer using 3g traffic, because wifi traffic is cheaper. You can target 3g and wifi together anyway - and most of the traffic you get would be wifi, and whatever little 3g traffic you get would be a bonus. Over time, you'll get enough 3g traffic to give you an idea on how well the offer can convert on 3g. Often, offers DO convert better on 3g even for the same conversion flow. If and when you see indication of that, you can duplicate the campaign to ONLY target 3g and set a higher bid to get more 3g traffic.

But no need to target 3g in the beginning and bid higher to get enough traffic - because if the conversion flow is the same, and if the offer sucks on wifi, chances it would suck on 3g too.

So that's for when the conversion flow is the same for 3g and wifi.

So what if the conversion flow is different? Let's say for 3g traffic, visitors are presented with the 1-click conversion flow of the offer, whereas visitors on wifi would get a more-complicated flow.

In that case, I would suggest the opposite: Target 3g from the start and not wifi. Reasoning? If the offer can't convert at the simple conversion flow, then you bet it would suck even worse with the more-complicated flow on wifi.

Only if the offer converts well on 3g, would I test it on wifi.


Shouldn't make too much of a difference - although at different bids you'd be triggering traffic from different placements.

0.30CPM more probably wouldn't make any considerable difference though.

And if you're using smart cpm, then you'd be getting traffic from a wide range of placements (I didn't recommend smart CPM in the tutorial because when I wrote it, PropellerAds did not have that option yet).


I wouldn't start testing bids until I have a offer+lander combination that looks promising (i.e. winners of split-tests).

You don't need to spend your entire budget before checking for statistical significance.

Basically, the budget is just for gauging whether you need to ditch a campaign or not. For example, if you've spent 10x payout on one offer and not getting a single conversion, that's a pretty good sign to not waste money on it. (Of course, the assumption here is that the landers you're testing are 1)popular in spy tools so you know they must work at least semi-well or better, and 2)you've optimized load time and made sure they function correctly and display well on various devices.)

But unless the conversion is so bad that it's costing you too much to do the split-testing, you would just keep running traffic and keep cutting until you have a winning offer+lander combo. Here's the strategy:

-Run landers and offers until an offer has made 2 conversions, then pause the rest. (Again, if you've spent what you feel is lots of money without getting conversions, then your split-test is too expensive - you'd either need to swap out the offers for another batch, or examine your landers again to make sure they load fast etc., or just target another geo altogether.)

-Continue running that offer with all the landers, and checking regularly for statistical significance and cutting landers - until you have a winner.

I would test bids after that point.


You can add the revenue in Voluum settings > conversion upload, in the form of "clickid,payout". e.g. c384EFV6JHQODRN70575OK6U,10.0


Never do that.

OK maybe not never - but in most cases it would be a bad idea.

High CTR does NOT mean high CR!

For example, I can artificially inflate CTR by setting the lander to automatically redirect the visitor to the offer after x seconds (sometimes this will get good results - but only for specific landers). But if the visitor hasn't been presold adequately before getting redirected, then the CR will be shit.

Or, I could promise the moon on the lander to effect a high CTR - but then when the visitor gets to the offer page and sees a contradiction, they probably won't convert.

I've seen too many cases where the winning lander had a lower CTR than some of the other landers in the same split-test. It happens way too often for us to even use high CTR as a good guess.

So when would you take action based on CTR?

You would do that only when the CTR is too low that the CR would have to be impossibly high for the lander to even break-even.

For example, if the payout is $2, you're paying $1cpm for traffic, and the lander CTR is 0.10%...

That means only one visitor per every 1000 is clicking through to the offer! So you're paying $1 per visitor!

That also means one of every two visitors to your offer need to convert in order for you to even break even. And we all know this is nearly impossible - unless you were selling water in the middle of a desert.


Maynzie has given you lots of good recommendations. And all you need to do is search this forum for others:

https://stmforum.com/forum/showthrea...ike-affluentco

https://stmforum.com/forum/showthrea...s-aff-networks

https://stmforum.com/forum/showthrea...-sweeps-offers


Having access to more networks can help. There are affiliates that will test extensively by getting a list of recommended offers from every AM on all networks they've signed up to, and testing all of them. If you're wanting to "take over" a vertical, that's what you would do. (After the initial heavy testing, you'd only need to ask AMs regularly for any new good offers in that vertical and test those.)

Try to find out which geos are hot at the moment for sweeps - and test those geos first. You have access to Adplexity - so start there. Browsing through hot offers lists from various networks is also a good way, as is asking your AMs for the hottest geos.

(Also: On Adplexity, check to see if sweeps is even a strong vertical in the geo you're planning to run in. If the geo is dominated by another vertical that has good offers with high profit margins, you may not even want to try to compete.)

Use "proven" offers in your initial testing for each geo - ones that either look like winners on Adplexity (e.g. uptrends over longer periods, large volumes, multiple affiliates are running), or ones that are recommended by an AM that multiple affiliates are running with good results.

If after testing at most 10+ offers from 3-4 networks that are known to be strong for sweeps, you're not seeing promising results, then I'd say test another geo. Of course there will always be a chance that you can find better offers on a network you haven't signed up with yet, but if you're looking for low-hanging fruits, then it would be easier to just test another geo.


Thanks for providing all that information and data!

Otherwise we wouldn't be able to give you as much feedback or guidance.

This is what you're paying your STM subscription for - so no need to apologize. You're entitled to this service!


If you're using landers that you're confident in (popular + load fast + function right + display well) and an offer doesn't convert after 10x payout in traffic, I'd say it's pretty hopeless - no need to restart your budget count.

You would want to continue running and checking for stat-sig on the landers, and cutting them until you're down to a winner. The budget won't really be relevant here - you cut until you have a winner.

The only reason for you to stop would be if your offer stops converting for some reason - or maybe your first 2 conversions were just by luck, or something fishy's going on with the aff network or the advertiser (e.g. trying to get you "hooked" with some conversions in the beginning, but then started scrubbing/shaving heavily), or your traffic quality suddenly changed for any of a plethora of reasons.

In other words, only stop if your lander split-test becomes too expensive to justify keeping it running. You're trying to find a promising offer+lander combination here. If it doesn't look like the offer has promise with any of the landers, then keep running to identify a "winning" lander would be a waste of money - because the "winning" offer+lander offer won't be making you money.


Wait! Why would you be running different offers with different landers? Are the offers not of the same type?

If that's the case, then you really should make sure to choose the same type of offers!

You technically can set up different paths to run different sweeps offers (e.g. different prizes like MacDonald's vouchers vs. iPhone X), but in that case, every path would be a separate test, and should EACH have multiple offers and landers. Running 1 offer with 1 lander would be a waste of money IMO, unless you're using an offer and a lander you somehow KNOW for sure are winners.


Hope that helps! Looking forward to updates!



Amy

---

Hi Amy!

WOW! Thank you SOOOOOO much for taking the time to give me such thorough feedback! I truly appreciate it!

As far as running a test solely for bots prior to testing landers/offers, should I be doing both 3G and Wifi at the same time if the offer was for both and the conversion flow was the same? Then after removing bot-infested placements, run a separate Wifi camp, and a 3G camp.

The offers I ran were all sweeps but of different prizes in the same GEO, as I had a hard time finding proven offers of all the same prizes and in the same GEO. I asked about this in another thread and this is what I learned:

Now I'm confused. It sounds like it's important to stick to offers of the exact same prize, ie. all $100 H&M giftcards. And this is only important for testing to find winning landers?

Thanks a bunch, Amy!

Wan

---

Following Along Entry #5

My 2 camps are done running and over the past 24 hours - they didn't do so hot. Actually they ran ICE cold.

One camp was using my "winning" lander to split test 3 proven offers(different sweeps prizes though). After a $50 spend, I got a big fat ZERO conversions, although my CTR was not horrific, which I learned that this does not matter!

Stats:

My other camp was continuing to run my winning lander and offer, while cutting unprofitable placements. I basically cut out large placements that did not convert at all, which didn't worry me much as far as running out of traffic as the GEO is huge. I couldn't use Amy's non-aggressive method because each of my conversions came from different placements, arggggghhhhh. Anyway, I ended up with 4 more conversions than the previous 24 hours so I doubled my numbers to a WHOPPING 8 conversions! Here are what those stats look like:






My ROI from this campaign yesterday was in the -20's%, and today it's -81%

I spent a total of $160($80 yesterday, $80 today)
I had a revenue of $16.
I think it's safe to say goodbye to this one, but I'm keeping the lander to split test with other landers and offers.

Tomorrow: Look for more offers to run, spy some more, work on more landers.

---

You're welcome! Just doing my job.

I would also like to thank you for being a member and for sharing your experience!


Since conversion flow is the same for wifi and 3g, I would suggest to bid average and target both 3g and wifi. That way you'd be getting mostly wifi traffic, and any 3g traffic you get would just be a bonus.

This applies both to the bot test and to the "actual" offer test. If it converts OK on wifi, it would be worth the time and money to test 3g traffic, which is more scarce and more expensive.

If and when you decide to do that, you may want to do another bot test, this time focusing on 3g traffic - or just look at the data from the bits of 3g traffic you got in the first bot test (of wifi-plus-a-bit-of-carrier) and see if you had enough to tell you whether bot % is high enough OVERALL to justify doing a bot test specifically for 3g.


All twinaxe is saying is that you can run multiple prizes in the same campaign, as long as you run them in different paths.

Whether you choose to run different prizes in different paths of the same campaign, or in different campaigns simultaneously - in order to maximize chances of finding success, you should try to test multiple landers and offers for each prize.

It's not always possible to find multiple offers for a given prize, especially if it's a voucher offer for a specific product or company.

However: Even for those offers, you can always test multiple landers! If you can't find landers specifically for that prize, simply take other popular landers and customize them for that particular prize - by changing the image and ad text.

But that's just a suggestion. You may decide to just avoid the trouble of customizing landers, and instead test more offers in the same amount of time. That would be a good strategy as well.

However, by customizing landers you'll have an edge over people that are JUST ripping landers and using them as-is.


Uh - OK I got 2 questions here:

1)To come up with that winning lander, were you using conversions from multiple offers? Even offers with different prizes? If so, that would be mistake #1.

2)For this recent campaign where you were using the winning lander to test 3 offers - you said the offers are for different prizes. Then which prize is your lander for? That may just be mistake #2.

I'm MUCH less concerned about the success of an individual campaign - because everybody will encounter bad campaigns. But we must make sure your test methodology is solid. That way, overall, you'll end up with enough profitable ones to offset the unprofitable ones.

So let's focus on locking down a good test strategy!



Amy

---

Hi Amy!

To answer your questions:

No, my winning lander came from one single offer that received all 8 conversions.

Yes, I took my winning lander from question #1, and applied it to test 3 more offers of different prizes, but I did adjust the winning lander accordingly so it matched the offer(i.e. swapped out Iphone image for Samsung, replaced the word Iphone with Samsung), but all other elements of the winning lander remained unchanged.

---

Following Along Entry #6

Advice needed please:

About 30 hours ago, I started a brand new campaign to split test two different PIN submit offers(same prize category but different prizes) with a much higher payout ($15), again using my winning lander in which I made 2 versions of, accordingly. The first 24 hours received no conversions, but today I got 2! I'm still shocked! My original plan was to only test 2x payout, and I honestly did not think I'd get any conversions! Now I'm actually in the green!(for now)

However, I don't think I've reached any sort of stat significance to do any cutting yet. Both conversions are from 2 different zones, from 2 different offers. The only thing I can see to optimize right now is the OS, but I still think it's too early. How much more should I continue to run this?

---

Congratz mate on seeing some green!!

I would leave for another day or 2 how it is or even increase the budget to $40-50/daily and see if the trend continues. To early to cut android tbh only $6 spend and the offer is a $15 would let it spend $15 minimum first

Keep coming mate, you're on the right track!

---

Thanks for the replies to my questions! Sounds like the right approach!

Something to keep in mind though: The same lander theme may work better/worse for different prizes. But what you did should be good enough for initial testing. Once you find a good offer, you can choose to test more landers and/or lander variations if you need that extra boost.


The green is always easy on the eyes! Congrats!

My advice is: Either don't pick such a high-payout offer, or be prepared to spend the budget needed to test the offer(s).

Like Maynzie pointed out, it's way too early to decide whether Android should be cut. (Oh and please remember to never make the classic newbie mistake of comparing OSs using the peakconversion calculator to cut down to a winner - because that's not the goal! We want to keep as many profitable traffic segments as possible to maximize total profits!) Just 1 conversion can push Android into the green.

Having said THAT - if you're wanting to spend less money on testing, you can stop targeting android for now. You can also do that with any other traffic segments (browsers, placements, hours of day, whatever). However, a bit of gambling would be involved: You'd basically be betting on IOS being the better converter (over Android). If in reality that isn't the case, then you may end up spending more money than if you had kept both running. But you have that choice.

And if you want to try that, remember to RETEST everything you've cut out haphazardly (i.e. Android in this case) once you find a good offer or verify that one of the offers you're running is good.

I would DEFINITELY keep running traffic to these offers! Don't be afraid to spend money when you see something promising. Even though you've only made 2 conversions, it's a good sign when you see green before even starting to cut anything.

Please run $100 worth of traffic first to collect more conversions, then post some stats again (placements stats sorted by increasing spend and by decreasing conversions; and any other stats you find interesting) and we'll take another look.

Exciting times ahead!



Amy

---

Boy this really is a rollercoaster ride!

Unfortunately I reached a roadblock with the campaign the very next day after I reached green briefly. I logged into my affiliate network the next morning and the offers were no longer active(good thing I checked because I got no notification!). It took me a while to figure out what happened because my AM was on vacation, but I eventually found out that they stopped the offer because the advertiser did not like my LP flow. I was confused because my LP was just a simple survey lander page that’s all over Adplexity…everyone was running it! Anyway, in the end they said I could start running it again if I removed the survey questions and changed the CTA from “Enter Your Phone Number Now” to “Subscribe Now”. IMO the changes they wanted were way too drastic and I would have no chance of getting any conversions and with the payout of $15 a pop(no pun intended, lol), I wasn’t willing to risk that much money testing it on a lander that I didn’t think would work. I don’t know if it’s worth trying to come up with a completely different LP for this offer, but I think I’m going to go back to trying the lower payout offers again. Maybe I do need to learn how to cloak, but I didn’t really want to go down that route, nor do I have the technical skills. Maybe I just need to submit ALL of prelanders for approval before running. I feel like that would be so time consuming, and would take forever, especially when working with networks from opposite time zones. Do most affiliates actually do this???

I’ve been using this tool to cut ONLY offers and landers:

http://www.peakconversion.com/2012/0...al-calculator/

Is this correct?

Speaking of which, I’ve been a bit confused as to which calculators are to be used for what. From the 40 Day Tutorial, I believe some of the calculators referred to in there were for direct campaigns only, which is no longer recommended. For example, the Kill List Calculator, which I believe is more useful for cutting LARGER PLACEMENTS, so wouldn’t necessarily be good for pops because that has hundreds of smaller placements. Am I right in this thinking?

Searching on the forum for further info led me to this link:

https://stmforum.com/forum/showthrea...rules-of-thumb

Is this what I should be using to cut placements for pops? What do you recommend setting the CPM to if your using SmartCPM?

I think where I’m getting confused during the optimization process is how much of the info that’s in the 40 Day Tutorial I should be using to apply to NOT direct campaigns but LP-->Offer campaigns. How much of the awesome flow chart that Amy originally made applies to LP-->Offer campaigns?

What about rules of thumb for cutting traffic segments? I have just been going by what makes sense to be, which isn’t exactly scientific!

I really would like to get my process down before trying my next load of campaigns.

Oh, I found this gem from Amy today, and thought it might help other newbies reading this:

“Especially when running pop traffic - with the typical shorter life-span of pop camps, being able to cut placements faster can be more beneficial than being able to cut with more accuracy.”

I’ve also been working on time/life management during all of this learning, so while on the treadmill every morning I started watching past videos from AWC 2016 and 2017, trying to gleam some “gems” that still apply today. And today I saw Amy on one of the panels! It’s so great to put a face to a name!

Thanks for reading, everyone!

---

I worked on figuring out and writing down my process of running a POP campaign, pulling info from Amy's 40-day guide, and various other threads. Here's what I came up with:

1. Fix up 5-10 landers from Adplexity
- Load time should be no more than 2sec max
2. Find 3-5 top performing offers in a vertical and one GEO
- Ask AMs, search Adplexity
3. Run bot test (Caurmen’s script)
- Cut placements that are 70%-80% bots
4. Split test all landers and all offers
- If offer doesn’t convert at least once per 10x payout, cut it from rotation
- If offer converts twice, pause all other offers and proceed to step 5 below:
5. Use best offer to split test all landers
6. Find best lander and continue to run with best offer, if there are positive signs (are there enough profitable placements?) (It is also during this step that I would test different bidding tiers for the lander+offer)
A. -50% or better ROI prior to any cutting
B. Placements with 2 or more conversions
C. Major traffic segments that are green
7. If there are positive signs, use Rules of Thumb calculator to cut placements:
https://stmforum.com/forum/showthread.php?32267-how-to-cut-placements-using-rules-of-thumb
8. If no positive signs, stop campaign and start over with new batch of offers.
9. After cutting placements, cut out unprofitable traffic segments and/or target only profitable traffic segments.

---

I'll tell you what I understand - not saying it's 100% accurate, but it's based on a mix of personal experience plus what I've gathered from talking to people.

There ARE a lot of people that are cloaking. Won't lie to you.

But it doesn't mean there aren't profits to be made by people that don't cloak.

And there's a difference between cloaking to get approved on traffic sources, and cloaking to get past the advertiser / aff network. Either can jeopardize your relationship with either party, but I'd suggest to avoid the latter especially. If you want to get the best payouts (or even just to get paid, period) and get notified about good offers, you need to have a good relationship with your AM.

Many advertisers nowadays are smart, and know how to spy and track to find the real landers affiliates are running. And if they have a solid offer that they're planning on running for a decent amount of time, then they can't let affiliates run too aggressively - both due to legal reasons and to ensure lead quality. Because often, monetization tends to be more difficult with leads that result from aggressive promotion.

Your offer's owner may just be this type of advertiser. And if you're seeing promise with the offer, it may be worth it to try to work with the advertiser to come up with a lander that both converts and satisfy their requirements at the same time. Perhaps show them completely different landers to see which one they feel is the closest to meeting their requirements, and modify from there. Remember, chances are if they're so strict, they're weeding out a lot of other affiliates also. So if you can make them happy, you'd have an edge.

MacKeeper for example, when it first came out, I started running it but the advertiser didn't like my lander. I ended up designing a custom lander from scratch that was different from anything anyone was running at the time in that vertical, and then edited it multiple times until the advertiser was satisfied with it. Was making good profits for a while from many geos (it was an international offer). Too bad I was such a newbie at the time - if I had known how to scale, that would have been an easy high 5-figures/day for a while.

Of course that was before all the spy tools came out. But you can still make tweaks to one of the popular landers, so that it would 1)blend-in in the spy tool results, and 2)satisfy the advertiser, and 3)still be effective enough to convert profitably for you.

It's a lot of trouble - so you'd just have to weigh that against the profits potential and decide whether or not to pursue this offer.

As for submitting all your landers ahead of time: I wouldn't do that unless the advertiser was really, really strict about seeing every single lander. Your AM would be able to tell you if this is the case, but that type is rare. Usually even when an AM asks me to submit a lander, I would just submit one of the less-aggressive ones and not all of them and that'd be enough to keep the advertiser happy. Don't keep secrets from your AM - they can tell you what you can/cannot get away with for an advertiser. Your AM has the same goal as you, which is to maximize profits, so they WILL let you know if you can get away with something. And if they tell you not to do something, it's usually because risks of getting caught by the advertiser would be too high, or the advertiser has high expectations on lead quality - in that case just comply rather than go behind your AM's back and get caught by the advertiser who will then refuse to pay. A lose, lose, lose situation for all parties.

On a separate note: It was really unprofessional of your affiliate network to kick you from the offer without having notified you first. Try to look for an email from them (check your junk folder too). If they haven't sent you ANY notification, I would think twice about even working with them again in the future. Ideally they should have notified you on chat (e.g. skype) so you'd know to cease traffic. Not nice at all what they did to you. And your AM having been on vacation wasn't a valid excuse either.


The peakconversion calculator is good for cutting landers.

For cutting offers - if they have the same payouts, you can use the peakconversion calculator still. But if they have different payouts, this would be the method to use:

https://stmforum.com/forum/showthrea...211#post289211


You are absolutely correct!

Pop camps performance is so volatile - most of the time efficiency will trump accuracy when making decisions.

For landers and offers it's usually recommended to use statistical significance because they are what can make/break a campaign. For other campaign optimization efforts such as cutting placements, quick and efficient is the way to go. The bigger placements and big traffic segments such as major OSs/browsers/carriers/etc. justify spending a bit more time and money to get a better indication of profitability. Other than that I wouldn't waste too much time optimizing to squeeze out that extra bit of profit - you'd be better off focusing on scaling or testing new stuff.


Yup! Rules of thumb will do.

And if you're targeting tier 1 geos that have tons of placements, and don't want to lose your shirt, then you need to set a very stringent cutting criterion - at 1x payout or even 0.5x payout without conversions.

Once you're making profits you can always retest the bigger placements. But especially in the testing phase, you'd want to test for cheaper without losing too much money.


I'm pretty sure I explicitly stated that that flow chart is only to be used for direct-linking to carrier-billing offers - sort of like training wheels. You've progressed way past that point.

You can refine the nitty-gritty of your optimization process as you go, but the overall general idea I've pretty much summarized above, i.e. focus on optimizing the big traffic segments for efficiency, and focus on scaling and testing.


LOL I LOVE it every time this happens - when I provide feedback and then upon reading further down the post, discover that the OP has realized the exact same thing I've just written.

We've discussed placements. For stuff like major OSs/browsers/carriers/etc., where large chunks of traffic are involved, try to either cut out the low-performers very early on and then retest later, or leave them until last and cut placements first. Here's when to use each approach:

-When testing landers and offers, because you're not using your best offer+lander combo yet, you need to TRY to cut out as much non/bad-converting traffic as possible. (In fact, if you have a whitelist of placements from a previous campaign, that would be a good place to start!) This is to minimize the money you lose while testing. Therefore, cut as aggressively as you can afford to without leaving so little traffic that your testing slows to a crawl - this will of course depend on how much traffic is available in the first place, and this is why ideally you'd want to stay away from tiny geos.

So for example, if you drill down to OSs and see that Android is converting twice as well as IOS after a few conversions, then you may want to stop targeting IOS for now, and retest it later on only when you have a good offer+lander combo. (Note that some offers will perform better/worse for different OSs, so by cutting out an OS early on when you're testing offers to find a winning one may not be a good idea. But if you're still testing landers for example, then it's probably OK to exclude an OS - unless you have reason to believe your lander will convert a lot better for one OS over another, such as the display is different or the ad copy is aimed at owners of android vs. apply devices.)

-When you already have a profitable offer+lander, you should try to use them to make as much of the total traffic profitable as possible to maximize profits. So try to cut the smaller chunks first, i.e. placements, rather than larger chunks. After cutting more placements, an OS/browser/etc. that was previously in loss, can become green. But before taking this approach, be realistic: Look to see how many placements you've already cut, and how bad the big traffic segment is performing, and ask yourself: "Would it be possible to cut enough placements to push this traffic segment from red to green?" If the answer is "likely", then keep cutting placements. If the answer is "probably not", then just cut the big traffic segment. Personally I wouldn't waste too much time trying to make that green unless it was near break-even to start with, AND I haven't had a chance to cut a lot of the bigger placements yet.

But you get the idea - tweak the bigger chunks and not the smaller chunks. Also: Don't get caught up with optimizing by cutting stuff out (as it's time-consuming and expensive and won't give you major results). Instead, focus on testing offers and landers (with special emphasis on offers) until you can make a big part of the total traffic profitable, THEN proceed to cut out the unprofitable parts.

It all sounds so complicated when I explain it, because there are many scenarios. But after you've run a few more pop camps you'll begin to find a balance.


That's a good start! But you'll refine that over time. Plus you'll see many different situations which you'll know how to handle differently after a while.



Hope that helps!



Amy

---

Oh my gosh, I'm so sorry for the delayed reply! I did not see your post until now, when I was getting ready to post an update. Thank you for all of your feedback. I've read through them twice and I'm sure I will continue to refer to them as I am testing and running campaigns.

I love this story...thank you for sharing it. It inspires me to start thinking out of the box, and imagine what more I can do once I learn the basic process. And it always is humbling to remember that everyone was a newbie once!

Isn't that awesome how this happens? I think that's one of the reasons why this type of "journaling" is so helpful - you keep writing down your thoughts and you might eventually come to the answer on your own!

Ok, here's what I came on to post originally today:

Hi Everyone! I haven’t posted in a little while, but I’ve been working behind the scenes! I’ve been working on my process and realize the one thing that was taking up most of my time was the darn landing pages! I’ve gone through the simple coding tutorials, but it feels like some of the landers I’ve downloaded from Adplexity that I really want to use are too complex and take too long to figure out how to change/fix to make it work for me. So I made the executive decision to hire some help from Upworks. I haven’t selected anyone yet, but I received a good response from my job post so I think I’m going to select 3-5 people that I think could do a good job and test their work with a batch of landers. I found some helpful tips on STM on this so I’m going to make a template of what needs to be done to all the landers and send them out to the individuals.

I’m concurrently testing a batch of offers/landers in a tier 2 GEO right now. I tested 4 offers and 4 landers and used 1 winning offer to find 1 winning lander using statistical significance. Here are my stats from the 4 landers:



I think my “winning” lander is crap as the CTR is atrocious, but I think the offer looks pretty good! Am I interpreting this correctly? I’m thinking tomorrow I may try to work on the lander to try to get that CTR up because 3 conversions from 11 clicks sounds promising unless something fishy is going on. So far I’ve spent $40 on this campaign for a $1.20 payout. This includes bot test, testing 4 offers with similar payouts, etc.

---

Well, I logged into my traffic source this morning to get ready to run some more traffic to my campaign but low and behold, there was a big red "rejected" X on it! It seems like I've earned my ROP (Right of Passage) in AM today because I just got my first Google flagged domain! It looks like they only flagged my subdomain where one of my LPs live, but I'm thinking I need to always keep a handful of domains for backup. I'm hoping this isn't going to become a persistent problem though. Off to do some research!

---

Haha no apologies needed - forums are really good for this purpose: We can reply at our convenience, and the replies will be here waiting for the next time you check in. They aren't going anywhere.

This is definitely a big advantage of writing things down!

I would do this when I'm stuck on a problem - just write it down to get those brain juices flowing, then I'd leave it for a bit, come back to it, pretend it's NOT my problem but someone else's - and "give advice" to myself on how to solve it. Works quite well!


I would rip and fix up and test a few more landers rather than try to tweak the current landers.

Also: Have you optimized those landers for speed, made sure they displayed properly on all devices and functioned properly when clicked on?

Moreover: The low CTR may be due to the google flag which would have drastically lowered the average CTR for the day or whatever time range. After you switch the domain, run a bit of traffic to the original landers first, to see if CTR improves.


There are lots of posts on this topic - you'll no doubt find some.

Frustrating problem - but I trust you'll get it resolved! You have the tenacity to work through problems - that's the most useful quality IMO.



Amy

---

Great to hear from you again!

Wow - although switching trackers may or may not be the reason for the end of the domain flagging issues, it's certainly great news!

And yes, binom is very similar to Voluum and they even have a Voluum exporter/importer for transferring all the data from Voluum to Binom.

And congrats on the xx/day milestone! The difference between xx/day and xxx/day is just testing more widely and scaling more quickly. Wish you all the best in hitting that next milestone very soon!

X/day on autopilot is good, but only if you don't currently have other campaigns to focus on, or else it wouldn't be worth your time to monitor it to make sure it doesn't dip back into the red. How many traffic networks have you scaled it to?



Amy

---

Hi Amy!

I totally jinxed myself because after my last post about not getting flagged for a while, I woke up to a bright red Google flagged page so my theory about switching trackers went out the window. Oh well, at least I was able to switch it out quickly.

Thank you for your encouragement and feedback! Knowing that more testing/scaling is the only thing between me and my next goal of XXX/day is reassuring and motivating! I can't wait to post back here when it happens

As for my X/day autopilot campaign, I haven't tried scaling because the offer has a cap unfortunately.

I've also been thinking about the idea of moving on from pops(to FB perhaps) at some point, but I don't know when would be the right time. Do you have any thoughts on that?

Also, I saw that I can view the videos from AWE 2018 for 399 Euros, but about half of it is related to Ecommerce, which I don't think I'm interested in doing. Do you know if there be access to portions of this for less cost at some point in the near future? For example, if I wanted to only watch the videos related to affiliate marketing, I would only pay for those that I watched. Just wondering if you knew if something like that might be in the works.

---

AW!

Still, if the frequency of the flagging has decreased since the switch, then it's still an improvement, no?


You can start running on FB even right now if you wish.

Whether this would be a good time or not will depend entirely on how you feel. Let me tell you though that FB isn't half as simple as pop is. If you want to give FB a try, make sure you have the tenacity to give it a REALLY good try. The learning curve may be pretty steep compared to pop.

This is the only reason why I still recommend for newbies to start with pop - it's so easy to get started. Basically ask AMs to recommend some offers and rip some landers from Adplexity and off you go. It's not hard to make x-xx/day in profits. Veterans can do xxxx-xxxxx/day - although fewer people are able to do 5-figure/day now compared to say 2 years ago (mainly due to increase in competition, spy tools leveling the playing field, and decrase in the number of good offers available in general).

Facebook on the other hand, even if you're running 100% whitehat, you'd still need to figure out how to create and test ads to get low-enough CTRs in order to keep costs low, how to target a suitable audience which is a big topic in itself, etc. And if you're running greyhat or blackhat, then you'll need to figure out where to get accounts from (buy/rent/JV/farm), how to fund them, and how to cloak.

And because FB gives you the ability to target unlimited subsets of audiences, there are numerous verticals you can promote - unlike for pop where there are mainly only a few big verticals.

I'm not trying to discourage you from jumping to FB now, but make sure it's not because of shiny-object-syndrome, or because the grass looks greener on the FB side. Otherwise, you'd grow tired of FB even faster - because it's harder.

I would suggest for you to get more experience running pop first, before transitioning to FB. But if you REALLY have your heart set on starting FB, don't let me stop you - I would suggest that you try running the same thing - sweeps - on FB. You'll need to get accounts though, because sweeps are greyhat, so account bans are a real possibility. There are recent follow-alongs on FB+sweeps that you could read through, plus multiple threads by stickupkid on how to run sweeps on FB. Please let me know if you can't find them, and I'll make you a list.


I haven't heard of such plans, but you can always check with the conferences team directly, by emailing support@awconf.com. To be honest though, I doubt this would happen.




Amy

---

Testing more is the way to go! But it's such a difficult mental block to get past having to spend money on multiple campaigns at the same time - so moving past this is a breakthrough in itself! Congratulations!

Having said that: Make sure you're spending that money wisely. Optimize your landers to make sure they load fast and display and function correctly. There are things we can ignore for a first test, and then there are things we can't. The basic necessities have to be there.

And please do remember to set a reasonable daily and overall budget for each campaign. As long as you control costs in this way, no other accident would be too disastrous.

Also important: There's no need to do initial testing on so many traffic networks! Stuff like split-testing offers and landers should need to spend 5x the money on 5 different traffic networks. I would suggest to pick either PropellerAds or PopAds for testing - find a promising lander+offer first, then scale them to the other sources and focus on cutting placements and testing bids.


How many landers have you tested? When I first ran in AU, I tested over 20 landers, many of which I developed custom landers for. It's not an easy geo, but definitely easier than the other tier 1 English speaking geos (US, UK, CA).

Lack of profitable segments: That's OK as long as you have enough profitable placements. Also, it's too early to tell if you'll have profitable segments or not - too few conversions.


I would suggest to start by doing research on Adplexity and also talking to all your AMs, to find out which geo+verticals are hot right now, then plan things out from there.


1)Yes! Are you using the {cost} token?



If in Binom you're using the PropellerAds template when setting up the traffic source, you should have this token in the right place. Please verify this in your traffic source settings in Binom.

2)Most definitely!! Next: Focus on getting more experience so you can get your testing and optimization process down.




Amy

---

Doing some nice greens! And a 3-figure profit day! Congratulations!

After all that red from before, the green must be such a soothing balm on the wounds.

I know the temptation to recuperate previous losses must be strong, and it's understandable that you would want to wait a while before spending money again.

However, I'd like to remind you that it's easier to spend money while you still have money coming in.

So, try to start some testing - even if only small campaigns to ease your way back into the spending mode again.

So happy for you on your progress, and looking forward to following your progress this year!



Amy

---

Home > > Newbie Follow-Alongs