Home > > Newbie Follow-Alongs

---

40 days starting with Binom (45)

---

I'm following behind https://stmforum.com/forum/showthrea...g-(with-Binom) closely

day 3 is really giving me a hard time!

sharing some screenshots to help future learners, and also some questions I'd really appreciate answers to

Created new affiliate network in binom:


Question: what is the offer URL template used for?

Found my postback URL in binom (it’s in settings):
Voluum.png  Views: 918  Size: 41.8 KB  ID: 19825" class="size_medium" />

Set that for my global postback in Mobidea making necessary replacements for REPLACE fields
Went to offer spreadsheet and made necessary changes for final offer URL:

Question: why do data1 and data2 now become data4 and data5?

Created new offer in binom and filled out details


question: this offer is valid for multiple countries, is there a way to select multiple countries in my tracker?

Setup new traffic source for PropellerAds in binom:

question: I just clicked load from template and selected PropellerAds. updated aid and tid from propellerads' tracking page. is that right? not sure what all the tokens are about...

Created new campaign in binom

question: what do i select for cpc/cpm/cpa? I guess that might become clear in day 4


question: why is the name of the campaign reflected off of one offer? won't campaigns usually run more than just one offer at once? or is that only in early stages when you're split testing offers?

---

It automatically attaches parameters and tokens to the offer URL when you add new offers so that you don't have to do it manually.

You can't specify multiple countries but when the offer is available for several countries you can choose global.

Postback for trafficsource is only needed when you want to track conversions on the trafficsource.
For normal CPM/ SmartCPM campaigns its not needed.
From all the other tracking tokens basically only the zone ID is needed.

As you use the cost token you select CPC and check Auto.
You also set currency for traffic to €, I would better keep it in the currency from the trafficsource.
Most offers are also in dollar so I only use the currency option for offer payouts in other currencies than dollar.
I am not sure what exactly you mean.
It's up to you how you name the campaigns and the offers although the offer name you usually copy from the CPA network.
So when you give the campaign a similar name then it is somewhat reflected in the offer name.
But yes, in a campaign you can test multiple offers and landers.
One more thing, I see that you set a payout for the offer.
It's better to use auto payout when you add an offer.
That was you will always receive the correct payoutfor your conversions through the postback.
That helps in case the payout changes.

---

Just a heads up that when I added the affiliate network to Binom, the only thing I filled in was the name field "Mobidea"...left offer URL template and postback URL blank for that. Got Vortex to confirm it for me as well/

---

Day 3 Followup

Not sure if can tag people on this forum, but thank you to twinaxe and GregMorrison for looking thru my day 3.

Changes I made as result:

1. changed one of the offers to "global" since it applies to 5 latin american countries
2. change currency in campaign to dollar (to reflect propellerads)
3. changed offer payouts to auto
4. removed url template and postback url from 'affiliate network' on Binom - though I think this won't matter?

Questions:
1. still confused why data1 and data2 in the original offer URL from Mobidea becomes data4 and data5 in the final URL we paste into Binom?
2. shouldn't I select CPM on the Binom campaign, not CPC? since propellerads is CPM?

---

Day 4 Part 1


2 days after I signed up I got this code from propellerAds for free $50 after $100 spend. that's neat.


not sure if I need all those tokens, but figure too much info can't be bad.


I bid 1.25 as recommended above, lower than what Vortex recommends in the tutorial but looks OK from the graph above


hmm so might hit $30 in one day??


offer restrictions: only phone devices, claro, all OS -- hopefully that's reflected above


already have an impression after <20 minutes, neat

Questions:
1. The offer also accepts wifi, but I couldn't figure out how to set that in 'Targeting' and also restrict to only Claro. Would I need to create another campaign with the same offer to run to wifi leads?
(answer is yes)

Note:
read ahead and saw others running into $30 spent instead of $10, so will try to change to testing multiple offers ASAP

---

Day 4 Part 2

read thru all that info, answered my question above (yes to create another campaign for wifi leads)

todo:
-read thru days 5-6, make sure setup is correct
-read thru GregMorrison's follow through again, make sure I didn't miss any helpful bits
-look into creating campaign with multiple offers
-skip optimization parts and move ahead to LPs (hopefully my coding background helps )

---

I'm 15 hrs in and only got 66 impressions???



I'm going to increase cpm bid by 0.25 every few hrs until I see more impressions, I guess. at the day job but will look at this more after I get off..

---

Change your bid to SmartCPM

---

thanks twinaxe, did that. will post an update in a few hours

---

Day 5 Update

Question: should cost in Binom auto populate, or do I need to fill that in manually like in the Voluum tutorial via 'update cost'??

- changing to smartCPM definitely led to more impressions + for cheaper
- I tried to include only 'phone' for device types, but I'm seeing tablets in Binom. so I switched to exclude and selected everything except phone. hopefully that works
- skimmed thru days 5-18
- read thru more of GregMorrison's follow along, have better understanding of postback + tokens now and updated timezone in Mobidea to be same as Binom/PropellerAds

Click loss #s:
Propeller: 987
Binom: 707
Mobidea: 529

looks normal I guess..

will run more campaigns and update with final #s. no conversions yet so want to get one and test that tracker is working properly.

TODO:
- run more single offer campaigns
- follow tutorial to mass test offers
- once get conversion, make sure it's showing up in Binom properly (reference day 5 in tutorial for that)
- start landing pages

---

Apologies for arriving late on scene! Making good progress!

Simple: Mobidea asked me to change the instructions. I don't remember exactly why - I believe I explained it in that lesson somewhere in the subsequent posts.

Normally, if you see other networks use generic tracking tokens like "s2 s3 s4 s5" or "sub2 sub3 sub4 sub5", you can use any of them to store any values (although I would usually NOT use s1 or sub1, because on some networks that variable can only be used for static parameters and not dynamic parameters, i.e. variables that have changing values). However, the affiliate network platform Mobidea is using now, they've developed in-house, AND they use different tracking tokens and logic, where some of that seems unconventional compared to commercial platforms like CAKE or HASOFFERS. So basically just follow what they advise when running their offers, and focus less on trying to understand or justify the "why".


I have yet to explore binom, but if it's anything like Voluum, it should already "sense" all that data from each visitor when the campaign url is triggered (i.e. when the traffic network sends visitors to the campaign). The tracker uses different scripts and looks up databases like IP databases to identify all that info from each visitor and store that as tracker data.


Good call! As long as you're not scraping the bottom of the barrel and getting shitty traffic that doesn't do your offer/lander justice.


Yes and that's the only "beef" I have with Propeller - they're great in all respects except for the tendency to run over budget. And you're completely correct in saying that the best way to combat this is just to test more offers at the same time if possible.


In "Connection Type", the options used to be "All Types", "3G" and "Wifi". Now they've been renamed to "All", "Mobile" and "Other".

So, if you want to target both 3g and wifi, select "All". If 3g only select "Mobile". If wifi only select "Other".

As for which one to test first: Usually if the offer has a more-convenient conversion flow for 3g than for wifi, for example 1/2-click flow, then I would target 3g first, and only test wifi if the offer converts reasonably well on 3g traffic. My logic is that if not even the easy conversion flow converts, then the wifi version would only convert worse.

However, if the offer has the same conversion flow for both 3g and wifi traffic, then I would test wifi first, because no reason to pay more for the more-expensive 3g traffic. HOWEVER, I would target BOTH wifi and 3g - I would just bid WIFI PRICES i.e. lower prices. I would then get mostly wifi traffic, but may also get a bit of 3g traffic as well, and that may give me enough of an idea of how the offer converts on 3g. But any 3g traffic would only be a "bonus" then - if the offer converts alright on the wifi traffic, I would definitely set up a separate campaign and bid higher to get more 3g traffic to more-accurately gauge performance.


I'm about to learn how binom works. But in the meantime, please ask binom support for their answer - I've heard VERY good things about their support team!


Great suggestion by twinaxe! I'm including this in my rewrite of the tutorial - when I was writing it, fixed CPM was the only option. SmartCPM is better to use, especially in the beginning when you want a sample of traffic from the widest range of placements possible, in order to identify good and bad placements. Fixed CPM can give you finer control over exactly how much to bid for each placement, but then you'd receive traffic from fewer placements (for example, placements that other advertisers are bidding higher for, you may not see any traffic from any of those when bidding fixed CPM).

All in all, starting with smartCPM is good practice when you're dealing with a new geo on the specific traffic source.


Weird! May be a glitch at Propeller - don't recall having encountered that while I was running there.

That's a smart work-around! Did it work?


Definitely larger than ideal, both from traffic network to tracker, and from tracker to aff network.

But it would be hard to say whether there's anything "wrong" per se. I intend to run some tests on binom to establish some benchmark stats to see how I compare with what the rest of you are getting. Perhaps it's due to the slight delay in redirection speed caused by the distance between your tracker server and target location you're getting the traffic from. I didn't get a lot of clickloss with Voluum because it was cloud-based so redirection speed was not an issue for most/all target countries. It's not a binom issue, but rather an issue that's intrinsic to all self-hosted trackers.

And then I'm not saying the clickloss is due to tracker redirection speed FOR SURE. Just listing that as a possibility. Some testing between trackers would be necessary to get more insight.

Some portions of clickloss are due to the different ways of accounting for impressions, from traffic network to tracker to aff network.

Another possibility: Some offers have slow loading speed, such that visitors are bailing before the offer fully loads - that can account for clickloss from tracker to aff network as well.


No need to mass-test direct-linking to direct carrier billing offers - that will be a waste of time. It's so much harder to find profitable 1/2-click offers these days, that chances are you wouldn't even breakeven unless using blackhat tactics, and/or have good intel from someone reliable, or be able to get super-cheap traffic.

Do feel free to still run a couple of direct-linked campaigns as practice, and to verify your postback is working (if you get a conversion - which shouldn't be hard - all the best!)

But after that, please do proceed to the lessons on landing pages ASAP. No need to waste time doing mass testing of 1/2-click offers.



Amy

---

Yes, Binom will auto populate the costs when the trafficsource has a cost token and when you use it.
So for Propeller Ads just use the {cost} token in your campaign URL and set it up properly in Binom, then the costs for Propeller Ads campaigns will be auto populated

---

few notes while I get back into things:

- Binom seems buggy, had trouble loading templates and editing traffic source. only solution that worked was going to 'Monitor' clicking 'full exit' and logging back in
- @Vortex: clicking 'Exclude' on PropellerAds and selecting everything except Phone worked mostly, but some tablet traffic still got thru
- a campaign actually stopped after hitting $10, so maybe PropellerAds fixed the $30 thing
- no conversions yet, been running campaigns with few offers that Mobidea support recommended with no luck
- changing campaign setting to CPC from CPM, on Binom, fixed the auto populating of 'cost' issue. thanks again @twinaxe

---

recent campaign had bad clickloss, but got 1 conversion so I think tracking setup is correct.

PropellerAds 4025
Binom 3335
Mobidea 753

TODO:

- skim thru these at day job:

https://stmforum.com/forum/showthrea...e-Your-Answers
https://stmforum.com/forum/showthrea...ing-Parameters
https://stmforum.com/forum/showthrea...Make-Decisions
https://stmforum.com/forum/showthrea...-Input-Or-Time
https://stmforum.com/forum/showthrea...ts-for-landers

- Do Day 21 (Hosting + CDN)

---

finished day 21, propogation took under 30 minutes. was expecting a few hours so that's nice.

applied to Advidi

---

more time to work today, will be updating as I go along:

- applied to a bunch of networks using STM's aff network application tool
- asking on Mobidea for offers that fit these criteria: 1) sweeps offer 2) tier 3/4 geos 3) SOI leadgen preferred, maybe DOI 4) just asking question: what is best prize for offers to promote in your opinion?

- I probably should have applied to more networks earlier, that might be bottleneck today.

---

Since you only redirect go to campaign => advanced settings => change detect to nothing

Then you can find out if the click loss is lesser.

My clickloss with detect all and 10Mio. daily is from TS => Fraudscoring => Tracker => ~ 5%-10% thats normal and i always set 15% as default on the traffic source creation to track more cost then to less costs ;-)

but it always depends on the location of your tracker and how far away the GEO is you buy traffic from AND the power of your server but this you can ignore with your little clicks ;-)

---

Making good headway sheet731!

Like codeflame

I would suggest to check with your Mobidea AM on the clickloss. Like codeflame mentioned, the location of your binom server and how far away from the target geo it is, can have an impact on redirection speed. But unless you're using a very slow server, I don't see how you can lose such a large percentage of clicks.

Quite a few people have reported huge clickloss with Mobidea offers lately - I'll need to get in touch with a Mobidea rep to find out more details (the rep I was in touch with before had left, plus they no longer have presence here on STM - I'll provide more options in the updated version of the tutorial, which I'm in the middle of writing).

I wouldn't worry about that right now - but once you start testing landers with offers from another affiliate network, DO keep an eye on clickloss. We'll compare numbers and will need to do some troubleshooting if percentage is too high.


Codeflame could you please explain this in more detail? Is this on Propeller or Binom or Mobidea?




Amy

---

I jinxed myself by saying I would have time 2 days ago

finished day 22 though, starting day 23 and will update this post...

edit1: did the FTP setup and edit 'iphone 7' to 'iphone 10' exercise on my work Mac. crossFTP crashed once and can't open Sublime from that app, will see how it works on personal Mac and consider using a different FTP app if needed.

been 5 yrs since I coded in JavaScript, but hopefully I remember some..

Question: should I expect networks to ask me a ton of questions for the application process?

---

Question: Where on STM can I ask these kinds of code questions?? Here's an example:

Not really a newbie question, more code specific:

So I get the gist of what is going on, if the user is a bot or logs into Facebook on the lander -- then this is recorded via the lp_update_tokens function and then gets set into an "img" element that is created.

But then how do I get access to this info?? It seems like it's set into the html, and then the o var/ "img" element is never again to be seen or used..

---

quick update, I have 5 optimized landers and will start day 28 of the tutorial tomorrow.

planning to read a few other follow alongs, could use a bit inspiration right now.

I posted a question in Question and Answers in the newbie zone but it only got 4 views, so hope it's OK to ask again here:

---

From what I understand they want you to replace the value 1 from ext_id parameter with your clickid token.
Apart from that, I don't recommend to run this offer.
Seems like a cc submit, it's in a very expensive and competitive Geo and has a very high payout.
Testing 5 landers without previous stats of good and bad placements can get very expensive.

---

well everyone, I think I made a dumb mistake and ripped landers for a geo/prize pair that doesn't have many offers for it..

also I can't seem to get accepted by any affiliate networks that have a lot non tier 1 sweeps offers.

so my plan is to

1. apply to more networks from here https://stmforum.com/forum/showthrea...-in-march-2018 and https://stmforum.com/forum/showthrea...ut-of-the-Dark and from other follow alongs
2. run a camp with the couple offers I have with the 5 landers (a few of them are very aggressive, we'll see if propellerads approves...)

lastly thank you to twinaxe for your constant help, much appreciated


--

update:
https://imgur.com/a/du6cQKh

testing 4 landers - i think 1 of them will break rules of network - just waiting for propellerads to approve/reject

---

sorry for being so bad at updating. have some stuff to share:

1. I'm working with a few other affiliate networks now, talked to AMs (which I was nervous about, but they understood I'm a newbie) and am excited to try new offers. got some recommended

2. I had a couple conversions from a Mobidea offer as well (tier 2 too), which was exciting!!!

https://imgur.com/a/gwFQUWq

[note1: I am dumb and selected EUR as the currency in Mobidea on the postback page - though the offers settings are all in USD and so is Binom. so revenue is slightly higher.]

[note2: I can't post images directly here for some reason anymore. I know @axeman117 also couldn't create a thread in this forum before, maybe some setting changed? @vortex]

in particular, I really need help understanding 1) did I setup the bot script correctly from the screenshot? 2) how do I read the data and understand how to cut placements? the bot script also causes the alerts - that show up right when the user hits the lander - to close by themselves. i feel this could really affect CTR, and the 2 landers which I got conversions from I didn't use the bot script. so maybe i'll give up on that for now. I'm thinking cutting placements will happen regardless of if i know it's a result of bots or just certain audience

otoh, i am efficient at ripping landers and have plenty of offers to play around with. expect more from me this weekend!

clickloss:

PropellerAds - 9,685
Binom - 8,299
Mobidea - 336 people reaching the offer link


I'm going to try removing the 2 landers that didn't convert and add 2 more into the mix, then remove the bot script. wish me luck

--

edit: I realized one of my landers converted even with a typo. hopefully that's a good sign? the offer I selected passes the criteria in day 28-31 and I used the stats tool from there to cut the 2 worst landers. so I'm going to find which is the best lander then mass test offers using my new networks.

Question: how strict are affiliate networks with the types of landers you use? I'm asking my AMs and sending example screenshots as needed, but is putting 'Google'/logos/'FREE'/'you've won' generally OK/not OK?

---

I'll answer this with a part of the revamped 40-day tutorial which I'll be posting in a few days:

"The networks you've applied to above will try to contact you! Keep an eye on your email and chat messengers, and pick up on phone calls - basically monitor all the different methods of contact you've left on the application forms.

Some of the networks may want to have a chat with you before deciding on approving your account. Just be honest about your experience level - all the networks above are newbie-friendly.

Main things to remember are 1)be professional, 2)make it clear that you'll do your best to succeed and that you're not just dabbling, 3)you're a member of STM and following this tutorial, and that you can get help from this forum when you need to, and that 4)you'll abide by their terms of service.

The top concerns of affiliate network reps are: 1)they want to know that their time spent with you will result in profits, 2)they don't have to do a lot of hand-holding, and 3)you won't do anything shady to get them into trouble (both with the affiliate network and with the advertisers a.k.a. the offer owners).

After you're approved into an affiliate network, ask your affiliate manager ("AM" for short) for their skype/telegram/whatsapp whatever and add them - but it's also possible that they may have already added you in order to conduct the interview.

Also log into your affiliate dashboard after getting approved - on some of the networks you'll find the contact info of your AM there as well."


Regrettably, I suck at coding - can't really help you out much there! But do feel free to post questions in this section:

https://stmforum.com/forum/forumdisp...rs-amp-Scripts

If you really can't figure something out - just move onto the next ripped lander. In the future, you can always either partner up with someone who's strong in coding, or just hire a freelance coder.


As twinaxe has pointed out, running such a high-payout offer when you don't even have a proven lander, and before you've cut bad placements, is not a good idea.

It would be a LOT less expensive to run a low-payout offer to generate enough conversions to find a winning lander, then use that lander to test offers to find a good one, and then use this promising offer+lander combination to cut placements.

When you have a good lander, and have cut the worst placements, THAT would be the time to test higher-payout offers. Even then you may need to invest a good amount of money into optimizing the campaign to profitability, given the high payout - but at least you'll have minimized the costs using the initial effort as described above.

As for the test link: It sounds like the aff network is wanting you to add the offer and lander(s) to the tracker, then create a campaign on the tracker, and give them the campaign url. They may want to verify that your setup is correct. But really, once you start running traffic, you'll be able to verify whether you've set up tracking correctly, by doing the following:

1)Logging into your aff network dashboard and looking at offer stats - if traffic is showing up for the offer you were intending to run, you're good. If not, then you've either entered your aff link into your tracker incorrectly, or are targeting traffic that isn't accepted by the offer (and therefore the aff network has redirected all the visitors to offers that DO accept them).

2)When you see a conversion in aff network stats, go to your tracker stats to see if it's showing up there. (Sometimes there can be a delay in the aff network posting conversions to the tracker though, so if you don't see it right away, wait an hour or two.) If it is, then you're good. If not, then you haven't set up your postback url correctly.

Speaking of postback url: Like twinaxe said, you'd need to replace the "1" in the offer link with the tracker's clickid - for Binom it's {clickid}, so you should add this offer link to Binom:

http://clicks.2trck.xyz/?aff_id=X&offer_id=Y&ext_id={clickid}

You'd also need to get the postback url from binom, substitute the tokens with those of the aff network, and save that at the aff network (in the "global postback" section if they have one). I could verify the postback url for you if you like! Simple provide the following information:

1)The original postback url you copied from binom

2)Screenshot of the postback tokens at your aff network (log into the aff network, look for the page where you need to save your postback - that's where the tokens would typically be found).

3)The postback url you saved at the aff network.


That's a common experience - next time you'll know to hunt down some offers first, and THEN rip and fix up landers. No biggie.

Alternatively, since you already have the landers ready, look into other geos that speak the same language, to see whether there are offers for that particular prize.

Have you tried applying to Gotzha? They're newbie-friendly and have a lot of sweeps offers.


Thanks so much for letting me know! I've recently changed some forum sections around - perhaps it's a permissions issue. Will do some troubleshooting right after I finish replying to your post, and report back!


I can't really tell from the screenshot whether or not you've set up the bot script correctly without verifying all the lander code was implemented correctly. Also, you did implement the bot script only for LP3 and LP4, correct?

A part of the stats does look weird: The number of impressions (or "Clicks" in binom) for the actual offer for LP3 and LP4 are low. Each lander should have gotten half of the total traffic to path 2 (i.e. 4110/2 = 2055 impressions).

I have yet to run a bot test using binom - will need to do that soon anyways when I adapt the tutorial for binom. But for now, I would suggest to just run the bot test separately. Run a bot test FIRST, using 1 lander and 1 offer, cut high-bot placements, and THEN run the "real" campaign using landers that don't contain the bot script. I can probably offer a more efficient solution later - but as mentioned, will need to test it first. (I'm thinking that when testing multiple landers AND offers in one campaign, we should be able to just set up a separate tracker PATH for each offer + all landers. Just need to run a quick test to verify that this setup will work.)

You're correct in saying that cutting placements will happen as you run traffic as normal - but the idea of running a bot test first is to spend a lots less money to weed out the obvious culprits. Caurmen told me that just be running 60 impressions to a placement will be statistically significant in judging bot %. Say you're bidding $1 CPM for traffic, and your payout is $2. Running a bot test to each placement will only cost you around $0.05, whereas if you had wanted to judge the same placement based on number of conversions, you'd have to run at least 0.5x-2x payout's worth of traffic to it = between $1 and $4. That's a big difference that can potentially result in significant cost-savings.


Looking forward!


The 2 landers that didn't convert, could be due to the bot test not having been executed correctly. You may want to consider just starting from scratch, i.e. doing a bot test using 1 lander + 1 offer to weed out high-bot placements, then testing all landers again with the bot scripts removed.

TIP: It may make it less confusing to have 2 different versions of each lander on your server: One with bot script and one without. And have separate entries in the tracker. This will result in less confusion than to manually edit lander files to add/delete/add/delete bot script code. When you look at lander stats in the tracker, it would be harder to tell which stats were generated by the bot script version and which generated by the version without the bot script.


Sounds like a plan!


It would really depend on the aff network and the offer owner (aka "advertiser").

Ask your AM how aggressive your landers are allowed to be. AMs want to make money with you, and more aggressive = more money, so if at all possible they WILL allow you to run aggressive.

Usually the only situation where they won't allow it, is when the advertiser explicitly tells the network they won't pay up if they catch an affiliate using aggressive landers.

And then there are TOSs at the traffic source that we need to be aware of. Ideally you'd want to push the boundaries and get stuff that is borderline-acceptable approved.

Either that, or we can cloak. As a mod I can't go into this topic. There are definitely risks involved and no matter how advanced the cloaking tactics are that we use, there's still a chance we could get caught. Depending on who catches us and how aggressive the landers are and the lead quality we're generating, we'd be looking at non-payment of commissions, and/or banning of aff network / traffic network accounts. Not fun.

Hope that answers your question! Have a productive weekend - but I hope you'll get some rest as well.




Amy

---

quick update: not much to share, switching geos and running bot tests (for wifi + carrier) using Binom's script (can be found in their documentation) which doesn't need an iframe to use. link: https://docs.binom.org/events.php

I'm running traffic to an offer that my AM recommended, but after looking on Adplexity I can't find that GEO on the filter list (so I'm just using a lander from another geo). is that a bad sign?

will provide a better update tomorrow, right now I'm so tired and I almost put the wrong lander in my campaign setup

I talked to an AM who told me tier 3/4 usually only has 1 offer per geo/prize pair. is that true? he recommended that I stick with tier 2, but would it be feasible to split test tier 3 offers across multiple networks to optimize for best offer?

---

Not finding a geo on adplexity doesn't mean you can't make money from it - Adplexity does not cover all countries AFAIK (although it already covers a TON). And actually by targeting a geo that isn't covered by adplexity, you may be able to avoid being spied on - so can innovate more on landers etc.

You could still rip landers from other geos that speak the same language though!

1 offer per geo/prize pair? Not last I checked - definitely depends on the geo and the particular point in time but I'm sure for the bigger geos you can find multiple offers per geo+prize. Perhaps they were just referring to their network?

Thanks for getting to the bottom on how to do bot testing on binom!

And I hope you got some rest! Health should be a priority whatever you do!



Amy

---

thanks Amy for the advice! yeah I think they were referring to within a singular network

--

So using Binom's script, I can continuously get bots % and split test.

I read other follow alongs and started splitting wifi and carrier traffic in separate camps. I also am running 2 geos at once (and burning thru $ at 2x the speed).

The 1st camp is running on wifi in a tier 2 geo and the 3rd and 5th camp are running on carrier traffic in a tier 3 geo.

I'm going to test more landers in both geos, but since the tier 3 geo only accepts traffic from one carrier, it's tough to get more traffic. I'm going to try PopAds very soon (ideally around the same time Vortex's tutorial on that goes up), so maybe I can get more traffic in that geo. I saw PopAds has a traffic estimator feature, so we will see.

The tier 2 geo, in addition to testing more landers I will also add another offer with same geo/prize in rotation with the current top performing lander -- to see if it converts better perhaps (+ get on the good side of that network's AM).

There's a 3rd geo that has a lot of sweep offers for the same prize, so I'm going to rip landers over the next 1-2 days and start running camps there.


questions:
1. does it ever make sense to redo the bot test again for zones previously blacklisted?
2. if someone had a list of all blacklisted zones, is it possible to share that information with others purchasing traffic on the same traffic source?
3. does using Caurmen's iframe method for bot detection violate some offer rules that say 'no iframe'? I'm going to ask my AM this, but keep forgetting to

---

took 1.5 days break, now back at it for the week

so scratch everything in my last post, I'm going thru the days 5-6 process flow chart for one of the geos -- will cut placements and see how much I can maximize ROI. the other geo I have a showdown between 2 landers, then hopefully can try to optimize that lander/offer pair too.

my AM said I'm doing great for that 2nd offer, which is nice to hear. hopefully a pay bump is on the way soon?

what I learned: lander CTR != conversion rate. the lander with the highest CTR got eliminated via the stats calculator

another thing I learned: I might be shooting myself in the foot because I tend to start campaigns at 12-2am New York time and then the campaign budget is reached some 12 hours later. but then I'm missing around 12pm-12am New York time's data for my campaigns.

I figured this out by breaking down campaign stats by hour

so I need to make a conscious effort to check my camps every 8 hours and act quickly to get a better overview of what's going on.

---

one offer I've been optimizing a camp for had a huge plummet in conversion rate. I'm asking if there's any problems with the offer, holding off on it until I hear back.

the other one I'm slowly optimizing - there's $20 of traffic per day but not much more than that. my AM seems satisfied with lead quality, am considering asking for pay bump -- when is the best time to ask for that? have done 14 usd in revenue with around 20-25 conversions if that matters

---

Did it fall for just one day or has been consistent for a few days now? Reason is even for most affiliates some days are going to be obscure anomalies to the conversion rates you see more consistently

---

just one day, would you recommend I give it another day's worth of traffic?

I was in the middle of the traffic flow chart from Vortex's tutorial days 5-6, went below -50% roi so wasn't sure what to do

---

The lesson on PopAds is out! However, PopAds isn't known for Voluum - chances are for your geos, you won't find more volume there than on Propeller.

You can try PopCash and Adcash and Zeropark though! But I wouldn't do that until/unless you have a good offer+lander. Split-test offers and landers on good-quality traffic (Propeller/PopAds) first, then scale the winners to the other sources and focus on cutting placements and testing bids. That way you save on budget.


Great! You've spent time preparing and optimizing the landers, so it would be to your advantage to test more offers - PLUS, testing offers (especially ones that have a good track record) is one of the best ways to improve a campaign.

If finding offers is an issue, it may be better to just ask all your AMs for recommendations, compile a list of offers, see how you can target the most number of offers with the same set of landers (same language+prize), and test those first.

I LOL'ed at the getting on the good side of your AM comment.


1. Please see https://stmforum.com/forum/showthrea...l=1#post358386

2. If it's for the same geo, worth a shot! But retesting would be good once you have a green campaign or promising offer+lander combo. If you could team up with a few people (don't remember if you're in a mastermind) and test separate geos and share blacklists, it would save everyone some money. Just have to find team mates that are willing to share and not just give.

3. The iframed part is only the little piece of code at the bottom. Neither your lander nor offer will be loaded in an iframe. So no need to worry.


Be careful about using that flow chart for anything but direct-linked 1/2-click offers.

If your geo has a ton of traffic you'll need to cut a LOT more aggressively than I suggested in that flow chart - between 0.5x-2x payout without converting. Even for a smaller geo, I would suggest cutting at 2x in loss or more.

If the showdown between the 2 landers keeps up for more than a few conversions, just randomly pick a winner instead of wasting money.


The fact that they said you were doing great, means you need to ask for a pay bump!


Important observation/discovery!


If you want to control your budget but still get the benefit of round-the-clock traffic, just throttle the traffic to spread it throughout the day - most traffic networks allow you to do that. PropellerAds and PopAds both do.


https://stmforum.com/forum/showthrea...formance-dives


That flowchart doesn't really account for "abnormalities" - if it did it would be infinitely more complicated. It's really just a crutch.

I'll be writing a more-useful (but more complicated) lesson on optimization hopefully next week. It's not a straight-forward topic but I'll do my best to explain what I know.

In the meantime - we can analyze your specific case here. But do check out the link above and try some of those things to see if your conversions come back.




Amy

---

Thank you Amy for the detailed feedback! I'll be looking forward to the optimization lesson for sure

I'm still alive and doing my thing, tackling a GEO with a ton of traffic but lots of green segments are sticking out.

Gave up on all the camps/geos I was previously running, as a recap:

• my AMs know I'm serious about this, as I come prepared to discussions and am making them $
• I did discuss a paybump with that one AM, but I couldn't get any lander/offer pair to stick
• part of that is the GEO traffic limitations, both in terms of volume and price
• another issue was the lack of offers out there for the prize I selected
• yet another issue was the top performing lander was all over Adplexity! the lander was getting OLD

---

Lots of green segments are always easy on the eyes!

Regarding the lack of suitable offers: Perhaps a better approach would be to start by talking to all your AMs and compiling a list of good sweeps offers, then forming a strategy as to how you could prepare the fewest lander sets to run the most number of offers. Also, for some landers it's easy to "switch out" the prize - that way you can use one set of landers for multiple prizes just by switching the prize names and images.

Also: Now that you have more experience, you can branch into other verticals. Again, talk to your AMs to see what's hot right now, browse Adplexity to see which verticals are receiving the most traffic in each geo, and test from there.

Regarding landers getting old: The fact that you're seeing them everywhere means THEY WORK. Yes - it always means they're approaching saturation more and more quickly, but you can innovate based on them! Using the most popular landers to establish a benchmark was the smartest thing you could do, but now that you have a benchmark, feel free to test variations or even completely new landers you come up with yourself.

Have fun!



Amy

---

Hey Amy, are you suggesting I try to move out of sweeps?

I was thinking of not spreading myself too thin as I'm learning the ropes. Sweeps isn't my first choice, but I've seen tons of warnings for affiliates who give up in verticals too soon and never become a master of any.

in any case, I'll definitely try out different prizes. I've been doing the same one this entire time, courtesy of AM's recommendations
--

quick update: from reading @Wakeboarder 's thread, I've become inspired to try a new GEO every 1-2 days and see what happens. budget's not too big of a problem for me, I'm just trying to have fun with this and gather as much data as possible. if you ever read this, thanks for your FA!

---

quick update:

• the big green segments ended up disappearing. I think I may have gotten a whitelist camp to get enough traffic to work. it's kind of borderline, but I'll give it a day and see. had to comb thru a ton of data to find the green in this GEO. but there's also some work I can do to improve the lander, so I'm going to iterate on that then use some coding tricks to make it difficult to rip
• a GEO my AM recommended is annoying because it has 2 main languages. I'm already using Binom rules to redirect traffic, but I think it's going to end up being like running 2 camps in one. @axeman117 and I are trading landers, he has other languages' landers but he's asleep now I think

if the camp is green by tomorrow, I'll post the negative ROI stats + the green camp as a thank you to the STM community, + hopefully inspire some other newbies..

---

Right now you're just ripping landers and fixing them up and using them as they are. You could be doing the same for another vertical.

I would just recommend to stay away from antivirus - it mostly requires cloaking to succeed, and comes with all sorts of risks.

But yes there are things you're learning about sweeps that would give you an edge. For example offer conversion flows, what types of landers work better, etc.

It's really up to you. But if you have a good feeling about an offer, try not to let the fact that it's not "your vertical" keep you from testing it. Sometimes it's better to go with gut-feeling. I know many people that have tried multiple verticals before finding their "true calling". Just make sure it's not because of shiny-object syndrome or because you're encountering a bit of difficulty with the current vertical that you're switching. And sticking with something for a while before trying something else is always good too.


If you're going to do that, pick tier 3/4 geos. The more difficult tier 1/2 geos will require extensive testing of landers and offers, then extensive cutting of placements etc. - all of which will require capital upfront + time. Picking up a new geo like that every 1-2 days may be hard on the wallet, and it would be easy to drop the ball when juggling so many big geos.

Off the top of my head, here are some geos that aren't as competitive, that have good traffic volume, that may be good to try for a newbie:

ID
TH
VN
PH
NG
BD
SA, AE and other arabic-speaking geos (the bigger ones)
Latin American geos (the bigger ones)

Fun is what it should be about! Wish you lots of it!


You can also try to set up 2 campaigns, each targeting a separate language. Either would work.

I think it's fantastic that you and axeman are working together - 2 people can be more efficient than one!


Keeping my fingers crossed for you! But at the rate you're planning on testing, you'll see a green camp soon.



Amy

---

thanks Amy! I'm just thinking since I would be bored over American Thanksgiving anyhow, may as well be super productive this week and then see what happens haha

damn this whole time I didn't realize you could select Language in propellerads ._. stupid me

---

PropellerAds hasn't been treating me well, I'm going to split the giant GEO into several segments then put each segment in a camp.

I'm thinking the problem is 1 or many of: 1) bid varies immensely for each segment and so I'm bidding too low for some 2) some placements only drive traffic for a set # of hours a day 3) PropellerAds is just wonky recently, I saw GregMorrison seeing similar issues

I had a green camp out of the box with 0 cutting of placements, landers, offers (3 each):


having trouble getting enough traffic though.. so will have to figure out what to do about that

---

edit update:

actually the leads are coming back super delayed from the CPA network it seems, hours later now I got this:


30% roi with 0 optimizations is promising, now just need to see if I can get more traffic..

---

I have the same issue.
Increased my bid today from $1.0 -> $1.5 -> $2.0 and now I'm sitting at $2.5 SCPM And still haven't got more than 9000 impressions per day.
Tried other traffic sources, but not converting as well

---

quick update:

fulfilled social obligations for American Thanksgiving and gained probably 5kg in the process

I have 3 green camps but volume from PropellerAds continues being a big issue. my plan to address this is 1) target GEOs that Amy suggested and 2) try out different traffic networks ASAP

as an example, I got double the # of impressions for these last week during the same time of day:

---

Very promising indeed!


I've seen several people report this - it may be a problem with PropellerAds, but it may also be because of Black Friday and Cyber Monday. It may also be because end-of-year is approaching, and some advertisers have annual budgets they want to use up. Either of these can result in overbidding which will drive up prices, leading to lower volumes for us unless we bid higher.

If this IS indeed the case, then targeting tier 3/4 geos instead of tier 1/2 geos should help a bit. Hopefully though, it's a glitch and things will return to normal in a few days.


I LOLed at that - thanks!



Amy

---

I've tested maybe near 10 geos, with 3-5 landers each:



Nothing really to write home about. The 2 geos I got green, I needed to whitelist for one and the other I couldn't get enough traffic for the green segment.

I'm planning to go thru another 5-10 geos then decide where to go from there.

Honestly, I'm finding pops and sweeps (phone prizes) to be kind of dull.

Some things I'm thinking about are:

1. exploring vouchers/other prizes
2. switching vertical (Q: what other verticals are good for pops?)
3. switching traffic source (native seems appealing)
4. maybe paying for a course, STM already costs $99 and I feel like after 2 months I was able to learn a ton from Amy's tutorial
5. my background is software engineering, so other options could be A) selling friends' software products (get a % cut of recurring billing) B) sell my own software/apps - create my own app, run pops to get app install, then sell ads on my own app? just thinking out loud here

If anyone else has gone thru the 40 days and felt similar afterwards, I'd love to connect with you.

Thanks for reading

---

First of all, those stats are EXCITING! Some initial thoughts:

-Magic happens when you're willing to test extensively - which you have! Next step though would be to test further, i.e. go deep on the promising camps.

-I don't know how many offers you were testing per vertical+geo, but with 3-5 landers, you haven't spent enough on some of the campaigns to even find a winning lander. Once you've identified a good lander, you could then test more offers. If you see an offer+lander combo doing well enough to make a large-enough portion of your total traffic profitable, you can then focus on cutting the unprofitable portions. Another important task would be to scale the promising campaign to lots of other traffic networks. All this will take running more traffic to achieve.

-At this point I would focus on milking the 2 geos instead of testing 10 more. In pop everything is short-lived. When you find something that works, milk it immediately to recoup your losses and more. Test more offers, scale to more networks.

-Pop isn't glamorous by any means, but it IS the easiest to get up and running. If you feel that you want to switch to another type of traffic, you have my blessings (not that you need it!) Push and native are natural progressions from pop. However, if you want to try native, I would recommend that you stick with pop for a bit to learn more first, just so you won't need to pay 10 times the money to learn the same lessons on native.

-Another good model for newbies is FB+sweeps, but you'll need to know how to get accounts, because there IS a risk of getting accounts banned. It really depends on the angles you're using in your creatives. There are people that never or almost never get account banned, by using compliant angles. Look up tutorials/guides by stickupkid, and also FB+sweeps follow-alongs in the FB subforum.


Have fun no matter which road you choose!




Amy

---

Home > > Newbie Follow-Alongs