GGL ASMR: robots are hard at work, kitty is happy 🥰
Kittens, I've been promising to show you how GGL works — Ildar's and my brainchild for registering and farming Google accounts. It's time to relax and watch the automation work itself, accompanied by pleasant music.
- Account registration: https://youtu.be/hr6wj3L8moA
- Activating 2FA: https://youtu.be/VOWvj6gqFcY
- Signing in account with 2FA: https://youtu.be/rFPfQ-OCUbw
- Signing in account with email: https://youtu.be/6ek1p5flW3c
- Fetching statistics: https://youtu.be/SNzSgBqGs7c
- Automatic farming: https://youtu.be/lbD2UUnu1CM
The whole process is in this half-hour video: https://youtu.be/qMNwQ_FsLro
You can endlessly admire how the fire burns, how the water flows, and how someone else works, but I still do not recommend that you set fire to your farmers or pee on them...
Kittens, I've been promising to show you how GGL works — Ildar's and my brainchild for registering and farming Google accounts. It's time to relax and watch the automation work itself, accompanied by pleasant music.
- Account registration: https://youtu.be/hr6wj3L8moA
- Activating 2FA: https://youtu.be/VOWvj6gqFcY
- Signing in account with 2FA: https://youtu.be/rFPfQ-OCUbw
- Signing in account with email: https://youtu.be/6ek1p5flW3c
- Fetching statistics: https://youtu.be/SNzSgBqGs7c
- Automatic farming: https://youtu.be/lbD2UUnu1CM
The whole process is in this half-hour video: https://youtu.be/qMNwQ_FsLro
You can endlessly admire how the fire burns, how the water flows, and how someone else works, but I still do not recommend that you set fire to your farmers or pee on them...
Please open Telegram to view this post
VIEW IN TELEGRAM
Fundamentals Pit.
How a tracker works: click probability distribution
The post starts from the fifth paragraph. You’re welcome.
Kitties, by the time you’re reading this, a gorgeous keyboard from HardCore Affiliate Club has already arrived at my place. And that means it’s the perfect moment to talk about probabilities.
You know we have the HardCore community, and inside it there’s a funny little mini-casino. We play with internal points, then use the winnings to buy merch from sponsors. The jackpot this season was a custom mechanical keyboard. Mr. Bayan and I set up a friendly match: whoever hits the sacred 77,777 points first wins. He started with 12k, I had only 3k. Sounds like my loss was inevitable — same playing method, same probabilities…
But at the very last moment I not only closed the gap, I pulled ahead. Strategy? Nope. A miracle.
Let’s take ten thousand arbitrage guys, give each of them a thousand bucks and a revolver with one bullet. Each shot either doubles his money or sends him straight to the graveyard. By step ten, we’ll definitely have a couple thousand lucky bastards with a mega-buck, and by step twenty there’ll be a few dozen billionaires — and a massive cemetery behind us.
You’d think: one bullet out of six is lethal, so every six steps there must be a shot…
But nope. Some people get lucky on step ten, on step twenty, and even by step thirty there’s still a chance of ending up with a trillionaire. Because probability is not a guarantee — and random events do, in fact, happen.
A very similar misconception hits many arbitrage newbies when they first launch split testing. They expect that with two paths and a 50/50 probability, the first click goes to path one, the second to path two, the third back to path one, and so on in neat rotation. Bullshit. Absolutely not.
Every time the tracker chooses a path in a split, it has zero clue what happened on previous visits. That’s a standard optimization principle — every new click is evaluated from scratch.
The tracker receives a set of paths that passed filtering for the current visitor. Every time, it sees them as if for the first time, along with their probabilities. Maybe some fancy multi-armed bandit algorithm has already scribbled something naughty onto those probabilities, and instead of an even distribution we’re looking at a smart weighted list.
The tracker then gathers all probabilities together and stretches them on a Procrustean bed until they sum up to 10,000 — this is called normalization. Because you might end up with three paths: one at 300% and two at 146% — no guarantee they’ll sum to 100.
Then we simply pick a random number from 1 to 10,000. A brand-new one every time. And every time it can land in the slice owned by one path or another. Hit the first 50% slice twice in a row? Easy. Four times? Sure. Eight? Why not. What happened on the previous click stays in Vegas.
When you see 8 clicks on the first path and 2 on the second — that’s normal. That’s how probability works. Every click is a new choice. You’ll get something close to even around the fifth hundred clicks, and even then 235/265 is still perfectly normal.
Remember, kitties: the tracker does not distribute clicks evenly. Every time, it decides from scratch where to send the visitor. Probability is just a guideline, not a guarantee. And you can only approach it over distance.
At least ten thousand clicks or so…
How a tracker works: click probability distribution
The post starts from the fifth paragraph. You’re welcome.
Kitties, by the time you’re reading this, a gorgeous keyboard from HardCore Affiliate Club has already arrived at my place. And that means it’s the perfect moment to talk about probabilities.
You know we have the HardCore community, and inside it there’s a funny little mini-casino. We play with internal points, then use the winnings to buy merch from sponsors. The jackpot this season was a custom mechanical keyboard. Mr. Bayan and I set up a friendly match: whoever hits the sacred 77,777 points first wins. He started with 12k, I had only 3k. Sounds like my loss was inevitable — same playing method, same probabilities…
But at the very last moment I not only closed the gap, I pulled ahead. Strategy? Nope. A miracle.
Let’s take ten thousand arbitrage guys, give each of them a thousand bucks and a revolver with one bullet. Each shot either doubles his money or sends him straight to the graveyard. By step ten, we’ll definitely have a couple thousand lucky bastards with a mega-buck, and by step twenty there’ll be a few dozen billionaires — and a massive cemetery behind us.
You’d think: one bullet out of six is lethal, so every six steps there must be a shot…
But nope. Some people get lucky on step ten, on step twenty, and even by step thirty there’s still a chance of ending up with a trillionaire. Because probability is not a guarantee — and random events do, in fact, happen.
A very similar misconception hits many arbitrage newbies when they first launch split testing. They expect that with two paths and a 50/50 probability, the first click goes to path one, the second to path two, the third back to path one, and so on in neat rotation. Bullshit. Absolutely not.
Every time the tracker chooses a path in a split, it has zero clue what happened on previous visits. That’s a standard optimization principle — every new click is evaluated from scratch.
The tracker receives a set of paths that passed filtering for the current visitor. Every time, it sees them as if for the first time, along with their probabilities. Maybe some fancy multi-armed bandit algorithm has already scribbled something naughty onto those probabilities, and instead of an even distribution we’re looking at a smart weighted list.
The tracker then gathers all probabilities together and stretches them on a Procrustean bed until they sum up to 10,000 — this is called normalization. Because you might end up with three paths: one at 300% and two at 146% — no guarantee they’ll sum to 100.
Then we simply pick a random number from 1 to 10,000. A brand-new one every time. And every time it can land in the slice owned by one path or another. Hit the first 50% slice twice in a row? Easy. Four times? Sure. Eight? Why not. What happened on the previous click stays in Vegas.
When you see 8 clicks on the first path and 2 on the second — that’s normal. That’s how probability works. Every click is a new choice. You’ll get something close to even around the fifth hundred clicks, and even then 235/265 is still perfectly normal.
Remember, kitties: the tracker does not distribute clicks evenly. Every time, it decides from scratch where to send the visitor. Probability is just a guideline, not a guarantee. And you can only approach it over distance.
At least ten thousand clicks or so…
Is it possible to make a perfectly even split test?
No, it’s not. At least, I don’t know a single piece of software where this functionality is actually implemented.
Can it be done technically? “Yep, but what a fuck for?”
All you need is a counter that very quickly cycles through 1, 2, 3, 1, 2, 3, and based on that value the click is sent sequentially to the corresponding path.
Such a contraption would only work under extremely limited conditions:
➡️ All paths must have perfectly even probabilities.
50/50, 33/33/33/33, 25/25/25/25, etc.
The moment you want 60/40 — you’re instantly fucked, for very obvious mathematical reasons. You could get kinky and send the first 6 clicks to one path and the next 4 to another. Or rotate evenly, but give the last 2 in every ten to the first path.
But what about 63% and 37%? That’s a total shitshow. And why the hell would you even need this?
➡️ Filters won’t work.
Want to route different geos/devices to their own paths and still use a counter? Not happening.
Let’s say the first path supports all devices, and the second only Apple. How much traffic do we send to the second path? Half — we overload it. Less — oops, there goes your perfect evenness.
Sure, you could go full pervert and create a separate counter for every condition. But… hell why?
➡️ Smart split testing or any kind of optimization won’t work.
Because there’s an algorithm there that directly conflicts with the counter logic. And we can’t magically jump into the future to know exactly how much traffic will arrive and from where, so we can distribute it in the “correct” percentages.
And here you can’t even get creative. And again — hell why?
So in the end, if you try to implement this idea, you’ll get either:
a dumb little tool for one-off edge cases, or
a maximally perverted monstrosity that’s still impossible to optimize, or…
probabilistic distribution, which is exactly what you were trying to escape from in the first place.
So why the hell bother with all this?
No, it’s not. At least, I don’t know a single piece of software where this functionality is actually implemented.
Can it be done technically? “Yep, but what a fuck for?”
All you need is a counter that very quickly cycles through 1, 2, 3, 1, 2, 3, and based on that value the click is sent sequentially to the corresponding path.
Such a contraption would only work under extremely limited conditions:
50/50, 33/33/33/33, 25/25/25/25, etc.
The moment you want 60/40 — you’re instantly fucked, for very obvious mathematical reasons. You could get kinky and send the first 6 clicks to one path and the next 4 to another. Or rotate evenly, but give the last 2 in every ten to the first path.
But what about 63% and 37%? That’s a total shitshow. And why the hell would you even need this?
Want to route different geos/devices to their own paths and still use a counter? Not happening.
Let’s say the first path supports all devices, and the second only Apple. How much traffic do we send to the second path? Half — we overload it. Less — oops, there goes your perfect evenness.
Sure, you could go full pervert and create a separate counter for every condition. But… hell why?
Because there’s an algorithm there that directly conflicts with the counter logic. And we can’t magically jump into the future to know exactly how much traffic will arrive and from where, so we can distribute it in the “correct” percentages.
And here you can’t even get creative. And again — hell why?
So in the end, if you try to implement this idea, you’ll get either:
a dumb little tool for one-off edge cases, or
a maximally perverted monstrosity that’s still impossible to optimize, or…
probabilistic distribution, which is exactly what you were trying to escape from in the first place.
So why the hell bother with all this?
Please open Telegram to view this post
VIEW IN TELEGRAM
Alright, you 10 freaks — here you go!
About ten people messaged me saying they really need a mega-gadget for perfectly even click distribution across paths. I have absolutely no idea why the hell you need this — but I built it.
Why?
Because even the desire of ten perverted readers is already enough motivation for me to go and code some unique feature bullshit.
The implemented approach sends clicks sequentially to each available path, completely ignoring their probabilities, and distributes an almost perfectly equal number of clicks to every path. It will only work in split tests without any additional filters on the paths.
And yes, keep in mind: even this split can sometimes send a few extra clicks to one of the paths. But I hope you’ll survive…
And still.
Write to me and explain — why the hell do you need this?
About ten people messaged me saying they really need a mega-gadget for perfectly even click distribution across paths. I have absolutely no idea why the hell you need this — but I built it.
Why?
Because even the desire of ten perverted readers is already enough motivation for me to go and code some unique feature bullshit.
The implemented approach sends clicks sequentially to each available path, completely ignoring their probabilities, and distributes an almost perfectly equal number of clicks to every path. It will only work in split tests without any additional filters on the paths.
And yes, keep in mind: even this split can sometimes send a few extra clicks to one of the paths. But I hope you’ll survive…
And still.
Write to me and explain — why the hell do you need this?
Tracker Updates Digest. Part 1: Priorities 🔄
Over the past month, the tracker has picked up a few fun new features. Today I want to share one — the most important and the coolest one so far: Priority handling.
The logic is simple: you create a flow with multiple ways for every possible scenario, and then show the system the order and priority in which clicks should be sent to each way.
If you set up 5 ways, with 2 at the highest priority and 3 at normal priority — as long as it’s possible to send a click to a way from the highest-priority pool, that’s exactly where it will go.
So when does traffic not go to the highest priority? There are two cases:
1. Display sequence. For example, a user just fell into our paws — the system sends them to the highest-priority way. If they show up again, the system will send them to the next way in the hierarchy. And so on, until it runs out.
2. Cap drain. Pretty straightforward. Once caps on the highest priority ways are exhausted, traffic spills over to the next priority level, and so on.
How many priorities are there? Seven. You can use as many as you need.
How is a way chosen within the same priority? Plain old random or a multi-armed bandit algorithm. In short — business as usual.
Who is this for? Anyone building funnels that are even slightly more complex than a default redirect or a basic split test. This system improves traffic control and helps squeeze more conversions out of leads — or simply fill caps correctly🔄
Over the past month, the tracker has picked up a few fun new features. Today I want to share one — the most important and the coolest one so far: Priority handling.
The logic is simple: you create a flow with multiple ways for every possible scenario, and then show the system the order and priority in which clicks should be sent to each way.
If you set up 5 ways, with 2 at the highest priority and 3 at normal priority — as long as it’s possible to send a click to a way from the highest-priority pool, that’s exactly where it will go.
So when does traffic not go to the highest priority? There are two cases:
1. Display sequence. For example, a user just fell into our paws — the system sends them to the highest-priority way. If they show up again, the system will send them to the next way in the hierarchy. And so on, until it runs out.
2. Cap drain. Pretty straightforward. Once caps on the highest priority ways are exhausted, traffic spills over to the next priority level, and so on.
How many priorities are there? Seven. You can use as many as you need.
How is a way chosen within the same priority? Plain old random or a multi-armed bandit algorithm. In short — business as usual.
Who is this for? Anyone building funnels that are even slightly more complex than a default redirect or a basic split test. This system improves traffic control and helps squeeze more conversions out of leads — or simply fill caps correctly
Please open Telegram to view this post
VIEW IN TELEGRAM
Tracker Updates Digest. Part 2: Fun Little Shit ☝️
I also brought in some nice and fun little tweaks. First of all, you’ll now have way less domain-fucking to deal with:
1. In the Domains section, we’ve added WHOIS data, similar to the AlterCPA platform. Now you can clearly see what’s working, what’s broken, and what’s slowly bleeding out.
2. In Domain settings, you can now enable notifications for domain status changes or expiration dates.
You can get alerts via email or straight to Telegram. So you find out about a dead or expiring domain instantly, not later from stats showing zero EPC.
I also shipped a proper freak feature — even distribution in split tests.
So if you’re one of those people who want to compare stats on 40 clicks, you can now send exactly 20 clicks to each route, instead of getting 17/23 with the usual random distribution.
Useless? Maybe.
Fun? Absolutely.
And nobody else has this.
As always, we’re the most inclusive freaks on the market 👯
I also brought in some nice and fun little tweaks. First of all, you’ll now have way less domain-fucking to deal with:
1. In the Domains section, we’ve added WHOIS data, similar to the AlterCPA platform. Now you can clearly see what’s working, what’s broken, and what’s slowly bleeding out.
2. In Domain settings, you can now enable notifications for domain status changes or expiration dates.
You can get alerts via email or straight to Telegram. So you find out about a dead or expiring domain instantly, not later from stats showing zero EPC.
I also shipped a proper freak feature — even distribution in split tests.
So if you’re one of those people who want to compare stats on 40 clicks, you can now send exactly 20 clicks to each route, instead of getting 17/23 with the usual random distribution.
Useless? Maybe.
Fun? Absolutely.
And nobody else has this.
As always, we’re the most inclusive freaks on the market 👯
Please open Telegram to view this post
VIEW IN TELEGRAM
Who am I? A bourgeois or a kitty working for quality…
Today I want to talk about support. And about the model of its existence that I’ve chosen.
Support has been running in this mode for many years now, and I’m not going to change it — even though, from a marketing point of view, I’m getting absolutely wrecked by competitors.
But I don’t give a fuck. And I’m ready to explain why.
How it works now: Paid support with fixed working hours, and double pay if something is on fire during my kittens’ off-hours. There are packages, different options, different price tags.
How the market is used to it: A statement on the landing page: «Free 24/7 support.»
Why will I never write that?
Thesis 1: In practice — not on a landing page — 24/7 support is provided either by a bot or by a human with expertise at the level of «Have you tried restarting your router?»
We, on the other hand, have professionals of a galactically-fucking-awesome level, and they need to be paid properly — not like cashiers at a discount grocery store.
Many also use AI bots at night and sell it as round-the-clock support. No comments here.
And the fact that with half of these «awesome» supports you wait for replies for hours or days — even during business hours — I won’t elaborate on either (I already did).
Thesis 2: Years of practice show that we simply don’t need 24/7 support.
All the software works stably and can only break if you abuse it in a truly perverted way — with three black dicks and two white horses.
Thesis 3: Free support is never free.
The salaries of idle freeloaders are just baked into the pricing.
Given that half of our clients are companies with their own tech people, there’s no point in dumping unnecessary costs on them.
We give a choice: your engineers — or ours on outsourcing, for a reasonable amount of money.
And now a question for you:
Am I right?
Do you like this approach more, or the marketing one?
Would you choose «free» but shitty support included in the price — or excellent support, optional and for an extra fee?
Today I want to talk about support. And about the model of its existence that I’ve chosen.
Support has been running in this mode for many years now, and I’m not going to change it — even though, from a marketing point of view, I’m getting absolutely wrecked by competitors.
But I don’t give a fuck. And I’m ready to explain why.
How it works now: Paid support with fixed working hours, and double pay if something is on fire during my kittens’ off-hours. There are packages, different options, different price tags.
How the market is used to it: A statement on the landing page: «Free 24/7 support.»
Why will I never write that?
Thesis 1: In practice — not on a landing page — 24/7 support is provided either by a bot or by a human with expertise at the level of «Have you tried restarting your router?»
We, on the other hand, have professionals of a galactically-fucking-awesome level, and they need to be paid properly — not like cashiers at a discount grocery store.
Many also use AI bots at night and sell it as round-the-clock support. No comments here.
And the fact that with half of these «awesome» supports you wait for replies for hours or days — even during business hours — I won’t elaborate on either (I already did).
Thesis 2: Years of practice show that we simply don’t need 24/7 support.
All the software works stably and can only break if you abuse it in a truly perverted way — with three black dicks and two white horses.
Thesis 3: Free support is never free.
The salaries of idle freeloaders are just baked into the pricing.
Given that half of our clients are companies with their own tech people, there’s no point in dumping unnecessary costs on them.
We give a choice: your engineers — or ours on outsourcing, for a reasonable amount of money.
And now a question for you:
Am I right?
Do you like this approach more, or the marketing one?
Would you choose «free» but shitty support included in the price — or excellent support, optional and for an extra fee?
Which support is better?
Anonymous Poll
67%
Paid and actually works
33%
Free bullshit, 24/7 (allegedly)
Pricing & Support Updates in AlterCPA 🔄
Kitties, after throwing a well-known substance onto the fan for everyone to see, let me break down how things work on our side. Especially since we’ve rolled out a bunch of changes that I hope and pray you’ll like and actually find useful.
🔸 Lifetime licenses are gone. Sad but true. I warned you — those who wanted in, made it in time. On an individual basis in DMs it’s still possible… but expensive. Veeery expensive. DOHUYA!
🔸 Monthly plans for server products now differ by support level.
🔸 The pricing and support logic of the platform has also been implemented in AlterCPA Lite.
🔸 The good old basic support is back in the plans. It’s been a while — and it gives you just that small, handy amount of help when you need it.
For AlterCPA Pro:
➡️ $500. Our classic. Harsh Enterprise for those who build with their own hands… and paws 👣 .
➡️ $950. The optimal plan. Basic support included, plus 20 simple good-old integrations, sites, offers.
➡️ $1450. The premium plan. Complex integrations, much higher limits on everything. And if you somehow outgrow those — no worries, we can always tailor a bigger package for you.
For AlterCPA Lite:
➡️ $111. Our classic. Once again, Harsh Enterprise for those who build with their own hands — license stays with you, no sneaky extra payments.
➡️ $333. The optimal plan. Basic support, assistance for 50 flows, 25 sites, and all sorts of mysterious stuff you probably don’t fully understand.
➡️ $555. The premium plan. 100 flows, 50 sites, server and system health monitoring. Basically everything you need to work in peace and pleasure.
◀️ Also rolled out separate service plans for my kittens ◀️
Out-of-plan support for AlterCPA Pro users
Out-of-plan support for AlterCPA Lite users
So here’s the bottom line. Enterprise folks can still just grab the software at a fair price without overpaying for support they don’t need. Need help? There are convenient plans that include the cat-hours of my team.
Overall, I genuinely believe this support-based pricing logic is as fair as it gets: you pay for what you actually need — and not for someone else.
Meow?
Kitties, after throwing a well-known substance onto the fan for everyone to see, let me break down how things work on our side. Especially since we’ve rolled out a bunch of changes that I hope and pray you’ll like and actually find useful.
For AlterCPA Pro:
For AlterCPA Lite:
Out-of-plan support for AlterCPA Pro users
Out-of-plan support for AlterCPA Lite users
So here’s the bottom line. Enterprise folks can still just grab the software at a fair price without overpaying for support they don’t need. Need help? There are convenient plans that include the cat-hours of my team.
Overall, I genuinely believe this support-based pricing logic is as fair as it gets: you pay for what you actually need — and not for someone else.
Meow?
Please open Telegram to view this post
VIEW IN TELEGRAM
God or Loser? Gambling Partner Network Edition
Every network wants to be a god of traffic. To take in tons of clicks, send them where payouts are higher and conversion is better, pay their subjects (affs) enough to grab a bite, and just enjoy the ride.
In reality, though, most networks end up playing the role of some broke loser with an outstretched hand — begging for drops of traffic and hoping it converts… and maybe even shows decent quality. And if it doesn’t — the aff just moves to a luckier competitor with the same offers.
So how do you turn a loser into a god? Well, you take traffic management into your own hands. You need to build a proper gambling-style smartlink.
Here are a few takes about this little beast:
➡️ It’s easiest to build a smartlink within a single GEO. But if you feel creative — you can configure it however you like.
➡️ One link can hide a whole bunch of different offer links inside.
➡️ Inside the link, you run an ML-powered split test that optimizes for the best EPC — and that’s exactly what makes your aff happy.
➡️ Once a cap runs out, the system automatically switches to a new cap for that product (if available) or rotates to another offer — and the split test keeps going.
➡️ One more important thing: smartlink solves the classic situation where you have 5 capped links for an offer the aff likes. The aff loses some, ignores others, or sends traffic to your competitors instead. Not great. With a smartlink — you just keep adding caps into one link, and it becomes super convenient for the aff. And when you don’t have to do anything — that’s peak convenience. And convenience is the best LTV booster.
In the end, it’s convenient for the aff to send you traffic — one link, solid EPC (assuming you’re not an idiot and actually watch your stats). It’s convenient for you — because your turnover grows💂♀️
So who are you after that? Right. A god of traffic!
And how can you become one?Send 100 USDT to wallet TMG7Ui… DM me @altercpapro
Every network wants to be a god of traffic. To take in tons of clicks, send them where payouts are higher and conversion is better, pay their subjects (affs) enough to grab a bite, and just enjoy the ride.
In reality, though, most networks end up playing the role of some broke loser with an outstretched hand — begging for drops of traffic and hoping it converts… and maybe even shows decent quality. And if it doesn’t — the aff just moves to a luckier competitor with the same offers.
So how do you turn a loser into a god? Well, you take traffic management into your own hands. You need to build a proper gambling-style smartlink.
Here are a few takes about this little beast:
In the end, it’s convenient for the aff to send you traffic — one link, solid EPC (assuming you’re not an idiot and actually watch your stats). It’s convenient for you — because your turnover grows
So who are you after that? Right. A god of traffic!
And how can you become one?
Please open Telegram to view this post
VIEW IN TELEGRAM
Media is too big
VIEW IN TELEGRAM
And here’s a short video guide. Unfortunately, it’s in Russian, but a lot of it will still be clear.
AlterCPA launches traffic give-away! 💵
Well, not exactly AlterCPA. And not exactly giving it away. But the traffic is real!
Our feline friends from CashFactories are gods of nutra and kings of local traffic. They’ve got plenty of this good stuff — and it’s quality. With money, approvals, and high redemption rates.I’d know — I see all the stats from the Sweaty Factory myself 😊
The offer is unique and personalized. They’re ready to share traffic with advertisers working on AlterCPA. Normally, they’re advertisers themselves and happily consume their own traffic. But for my cats, they’ve made a special opportunity.
And I personally recommend taking advantage of it — so your dick creams and green coffee beans don’t sit rotting in warehouses, but instead make their way to happy owners of micro-dicks and belly folds.
Welcome to CashFactories support➡️ https://t.me/CashFactories_bot
Well, not exactly AlterCPA. And not exactly giving it away. But the traffic is real!
Our feline friends from CashFactories are gods of nutra and kings of local traffic. They’ve got plenty of this good stuff — and it’s quality. With money, approvals, and high redemption rates.
The offer is unique and personalized. They’re ready to share traffic with advertisers working on AlterCPA. Normally, they’re advertisers themselves and happily consume their own traffic. But for my cats, they’ve made a special opportunity.
And I personally recommend taking advantage of it — so your dick creams and green coffee beans don’t sit rotting in warehouses, but instead make their way to happy owners of micro-dicks and belly folds.
Welcome to CashFactories support
Please open Telegram to view this post
VIEW IN TELEGRAM
What’s new in PrivateFlare 🔥
Our feline PF unit has done a massive amount of work — and I’m genuinely impressed. The software has clearly leveled up, and even at a glance you can see the product has matured in a serious way.
The new PrivateFlare features make managing domains and servers much easier: faster access to data, more flexible protection settings, better traffic control, and site acceleration via GeoDNS — all from a single dashboard. But let’s go step by step…
Working with tables is now way more удобнее
➡️ You can now export domain and server lists directly from the dashboard in CSV and JSON formats. The export respects active filters and selected columns — so you get exactly what you see on screen. It includes all records (not just the current page), and domain tags are preserved, which is super handy for analytics and integrations.
➡️ Customizable columns are here. You decide what data is displayed in tables. Settings are saved in your browser, so next time the interface looks exactly how you left it.
➡️ Domain filtering got a serious upgrade. Now you can filter by multiple parameters at once:
- online / offline;
- with SSL / without SSL;
- enabled / disabled.
Combined with text search, this lets you instantly find what you need — even in massive lists.
➡️ Using the same filters often? Save them as presets. Set your parameters, name it — and next time open the exact selection in one click.
Less annoying interface
The welcome notification in the dashboard can now be dismissed forever. If it used to pop up every time — now you close it once and forget about it.
Advanced domain protection settings
The «Protection» section got a major upgrade — now it contains all key tools for security and traffic control.
➡️ JS Challenge. Similar to Cloudflare’s Under Attack mode. New visitors see a verification page and only get access after passing the check. This effectively filters out bots and automated attacks. Returning users won’t see it again.
➡️ Basic protection (WAF). A set of rules blocking common attack types: SQL injections, XSS, vulnerability scans, and other typical threats.
➡️ Domain-to-server binding. Server selection is now directly inside the protection tab. You can route domain traffic through specific nodes or use automatic distribution across the network.
➡️ Bandwidth Limit. Set a maximum content delivery speed (KB/s). Useful for cost control, abuse prevention, and more even load distribution. Value 0 means no limits.
➡️ Geo traffic filtering. Flexible country-based access control:
- allow all countries;
- allow only selected;
- block selected.
Perfect for restricting access to target GEOs or blocking unwanted traffic regions.
Own DNS service with GeoDNS
PrivateFlare now provides its own nameservers:
ns1.privateflare.com
ns2.privateflare.com
You can point your domains to our DNS and manage all records directly from the panel. Supported types: A, AAAA, CNAME, MX, TXT, NS, SOA, SRV. Changes propagate in seconds.
The key feature — GeoDNS. You can assign different IPs for the same domain depending on the user’s country or continent. The system automatically routes visitors to the nearest server, reducing latency and speeding up load times.
🐈 🐈 🐈 PrivateFlare — protecting and accelerating your web resources 🐈 🐈 🐈
Welcome to the crew in Telegram➡️ @privateflarenews
Our feline PF unit has done a massive amount of work — and I’m genuinely impressed. The software has clearly leveled up, and even at a glance you can see the product has matured in a serious way.
The new PrivateFlare features make managing domains and servers much easier: faster access to data, more flexible protection settings, better traffic control, and site acceleration via GeoDNS — all from a single dashboard. But let’s go step by step…
Working with tables is now way more удобнее
- online / offline;
- with SSL / without SSL;
- enabled / disabled.
Combined with text search, this lets you instantly find what you need — even in massive lists.
Less annoying interface
The welcome notification in the dashboard can now be dismissed forever. If it used to pop up every time — now you close it once and forget about it.
Advanced domain protection settings
The «Protection» section got a major upgrade — now it contains all key tools for security and traffic control.
- allow all countries;
- allow only selected;
- block selected.
Perfect for restricting access to target GEOs or blocking unwanted traffic regions.
Own DNS service with GeoDNS
PrivateFlare now provides its own nameservers:
ns1.privateflare.com
ns2.privateflare.com
You can point your domains to our DNS and manage all records directly from the panel. Supported types: A, AAAA, CNAME, MX, TXT, NS, SOA, SRV. Changes propagate in seconds.
The key feature — GeoDNS. You can assign different IPs for the same domain depending on the user’s country or continent. The system automatically routes visitors to the nearest server, reducing latency and speeding up load times.
Welcome to the crew in Telegram
Please open Telegram to view this post
VIEW IN TELEGRAM
AlterCPA
Private Flare - сервис защиты сайтов
Защита ваших серверов и сайтов. Скроем ваш сервер от злоумышленников, защитим его от атак и ускорим загрузку сайтов.
Just a quick note about a nice new feature: if you resend the command /license LICENSE_CODE in your support chat, Pupurr will start sending reminders in the chat when your license is about to expire. Meow 🐈
Please open Telegram to view this post
VIEW IN TELEGRAM
We need to go DEEPER! Or: how not to build an API (a true smoker-tier API) 🚬
Kittens, today I’m gonna show you how not to build an API — even if someone’s holding a shotgun to your head and shoving a good old Soviet soldering iron up your ass (you know, the kind every household has). And we’ll be talking about renting phone numbers in such a wonderful service as TextVerified.
We all know how number rental APIs are supposed to work. The algorithm couldn’t be simpler: you send the required service and GEO — and get back a rental ID plus the phone number. What could possibly go wrong? We request a rental from TextVerified, and here’s what we get:
Alright, fine. Maybe they created the rental and immediately gave us an API link to fetch its data — where everything should be. Questionable, but okay. Not the first time an API decides to redirect instead of actually returning results — but let’s not kick yet another well-known crypto dumpster.
Excited, we hit the link hoping to finally get our rental data, and…
Where the hell did I click? I see an ID, but where’s the phone number?! Oh wait — this isn’t rental data. This is the order for the rental. Not a big deal, although… well, holy shit.. We’ve got another link, and surely this time we’ll get the actual rental!
Seriously? Another redirect?! You literally just gave me a link to the rental! I got a link, inside a link, inside another link — and here we go again. Fine, one more try, screw it…
Praise the Omnissiah — finally the rental data! On the fourth step of manual redirects, each with different payloads, we’ve finally reached what we were looking for — the actual phone number.
Oh, and by the way — even that’s fucked. No country code included.
Sometimes it’s a shame that the creators of such masterpieces don’t believe in hell — wishing them a separate boiler full of ever-changing redirects would make perfect sense. Or maybe this endless chain of redirects is the embodiment of Samsara?
We’ll never know.
Just… don’t build APIs like this ever again.
Kittens, today I’m gonna show you how not to build an API — even if someone’s holding a shotgun to your head and shoving a good old Soviet soldering iron up your ass (you know, the kind every household has). And we’ll be talking about renting phone numbers in such a wonderful service as TextVerified.
We all know how number rental APIs are supposed to work. The algorithm couldn’t be simpler: you send the required service and GEO — and get back a rental ID plus the phone number. What could possibly go wrong? We request a rental from TextVerified, and here’s what we get:
{
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/sales/rs_012JDM74PTH9B3944"
}
Alright, fine. Maybe they created the rental and immediately gave us an API link to fetch its data — where everything should be. Questionable, but okay. Not the first time an API decides to redirect instead of actually returning results — but let’s not kick yet another well-known crypto dumpster.
Excited, we hit the link hoping to finally get our rental data, and…
{
"createdAt": "2026-03-19T08:12:03.815099+00:00",
"id": "rs_012JDM74PTH9B3944",
"backOrderReservations": [],
"reservations": [
{
"id": "lr_012JDMW1JEATVX",
"link": {
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/reservations/lr_012JDMW1JEATVX?type=Renewable"
},
"reservationType": "renewable",
"serviceName": "Google"
}
],
"state": "succeeded",
"total": 8.4,
"updatedAt": "2026-03-19T08:12:03.820098+00:00"
}
Where the hell did I click? I see an ID, but where’s the phone number?! Oh wait — this isn’t rental data. This is the order for the rental. Not a big deal, although… well, holy shit.. We’ve got another link, and surely this time we’ll get the actual rental!
{
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/reservations/rental/renewable/lr_012JDMW1JEATVX"
}
Seriously? Another redirect?! You literally just gave me a link to the rental! I got a link, inside a link, inside another link — and here we go again. Fine, one more try, screw it…
{
"sms": {
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/sms?ReservationId=lr_012JDMW1JEATVX"
},
"calls": {
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/calls?ReservationId=lr_012JDMW1JEATVX"
},
"createdAt": "2026-03-19T08:12:03.820072+00:00",
"endsAt": "2026-04-18T08:12:03.820089+00:00",
"id": "lr_012JDMW1JEATVX",
"refund": {
"canRefund": true,
"link": {
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/reservations/rental/renewable/lr_012JDMW1JEATVX/refund"
},
"refundableUntil": "2026-03-19T09:42:03.820072+00:00"
},
"sale": {
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/sales/rs_012JDM74PTH9B3944"
},
"saleId": "rs_012JDM74PTH9B3944",
"serviceName": "Google",
"state": "renewableActive",
"billingCycle": {
"method": "GET",
"href": "https://www.textverified.com/api/pub/v2/billing-cycles/d924a71b-1234-4321-a11e-5c11a5311c1e"
},
"billingCycleId": "d924a71b-1234-4321-a11e-5c11a5311c1e",
"isIncludedForNextRenewal": true,
"number": "9876543210",
"alwaysOn": true
}
Praise the Omnissiah — finally the rental data! On the fourth step of manual redirects, each with different payloads, we’ve finally reached what we were looking for — the actual phone number.
Oh, and by the way — even that’s fucked. No country code included.
Sometimes it’s a shame that the creators of such masterpieces don’t believe in hell — wishing them a separate boiler full of ever-changing redirects would make perfect sense. Or maybe this endless chain of redirects is the embodiment of Samsara?
We’ll never know.
Just… don’t build APIs like this ever again.
Let’s talk about how conversions work in gambling. Part 1 🎰
We spend a lot of time breaking down nutra and all kinds of crypto stuff, but rarely talk about gambling. And yet — it feeds a huge chunk of this community!
Today I’ll break down, in simple terms, how conversions technically work in this vertical — and how you should set up your tracker depending on your needs.
Globally, there are two main approaches to tracking conversions. Most trackers use the first one — the classic model.
In this setup, one conversion equals one lead, which moves through different statuses as the user progresses through the funnel.
Typical flow:
➡️ Registration. A lead is created with status new / waiting.
➡️ First deposit received. The lead moves to hold.
➡️ Deposit confirmed. The lead moves to approved.
➡️ User blocked or canceled. The lead goes to trash / cancelled.
So basically, the entire user journey is stored inside a single lead — it just changes statuses over time.
Advantages:
➡️ No explosion of entities;
➡️ The entire funnel is stored in one lead;
➡️ Easy to track status changes;
➡️ Simple data structure.
Disadvantages:
The main issue appears in verticals where there’s a long gap between registration and deposit. The deposit conversion (the one that actually brings revenue and affects balance) gets recorded at the lead creation date — i.e., the registration date.
Nothing critical happens, but when registration in January, FTD in February:
➡️ The system shows revenue in January, while in reality (and for the advertiser) it belongs to February;
➡️ Caps become harder to track;
➡️ Stats get skewed;
➡️ In short — it’s inconvenient…
If you’re using a full platform like Pro/Cloud instead of just a tracker, this is partially solved. The platform can show multiple dates:
➡️ Lead creation date;
➡️ CRM submission date;
➡️ Approval or rejection date;
➡️ Payout or refund date.
But in a typical tracker (Red/One/Lite), you usually only get one date.
That’s why for some verticals (like gambling), where weeks can pass between lead creation and approval, a different approach is used.
Which I’ll cover in the next post🦊
We spend a lot of time breaking down nutra and all kinds of crypto stuff, but rarely talk about gambling. And yet — it feeds a huge chunk of this community!
Today I’ll break down, in simple terms, how conversions technically work in this vertical — and how you should set up your tracker depending on your needs.
Globally, there are two main approaches to tracking conversions. Most trackers use the first one — the classic model.
In this setup, one conversion equals one lead, which moves through different statuses as the user progresses through the funnel.
Typical flow:
So basically, the entire user journey is stored inside a single lead — it just changes statuses over time.
Advantages:
Disadvantages:
The main issue appears in verticals where there’s a long gap between registration and deposit. The deposit conversion (the one that actually brings revenue and affects balance) gets recorded at the lead creation date — i.e., the registration date.
Nothing critical happens, but when registration in January, FTD in February:
If you’re using a full platform like Pro/Cloud instead of just a tracker, this is partially solved. The platform can show multiple dates:
But in a typical tracker (Red/One/Lite), you usually only get one date.
That’s why for some verticals (like gambling), where weeks can pass between lead creation and approval, a different approach is used.
Which I’ll cover in the next post
Please open Telegram to view this post
VIEW IN TELEGRAM
Let’s talk about how conversions work in gambling. Part 2 🎰
So what do we do with gambling? We use a goal-based model.
The core idea is to create separate flows for different types of conversions.
Simple example:
➡️ Registration
➡️ Deposit
How this model works
When a user registers:
➡️ A lead is created;
➡️ Lead goal — registration;
➡️ Status can be pending or approved.
When the user makes an FTD:
➡️ A new lead is created;
➡️ It has a new external ID;
➡️ Lead goal — deposit;
➡️ The lead date matches the actual deposit date.
Advantages of this approach:
➡️ More accurate conversion tracking;
➡️ Caps are calculated correctly;
➡️ You can clearly see daily deposit inflow;
➡️ Standard statuses (hold, approve) work cleanly for deposit leads;
➡️ You can use separate columns like money in hold and confirmed revenue.
When working with revshare, this second (non-classic) model is the standard.
With RS:
➡️ Registration = separate lead;
➡️ Every payment = separate lead.
Each payment has:
➡️ Its own date;
➡️ Its own amount;
➡️ Its own status.
This allows you to see:
➡️ Full user history;
➡️ All their payments;
➡️ Accurate revenue dynamics.
Goal limits
The number of goals depends on the system:
➡️ Red / One — up to 5 goals;
➡️ AlterCPA Lite — unlimited goals.
Even with goals, you still use statuses.
Registration example:
➡️ Lead is created with goal registration and status pending;
➡️ After confirmation — status changes to approved.
Deposit example:
➡️ Lead is created with goal deposit and status pending;
➡️ After payment confirmation — status changes to approved.
Deposit tracking parameters:
➡️ addcost — adds deposit amount to an existing lead;
➡️ addprice — adds revenue (e.g., for revshare).
Grouping data by customer
If you’re using a multi-lead model (registration, deposits, etc.), it’s best to group data by Customer ID passed by the affiliate network.
If you group by click — you’ll get a massive table with thousands of clicks, and only a small portion will actually contain useful data.
If you group by Customer ID — the table will show:
➡️ Actual users;
➡️ Their registration;
➡️ All their deposits;
➡️ Full user-level stats.
This gives you the full picture of how a user interacts with your system.
Long story short: besides the obvious boost in stats accuracy, if the advertiser shares deposit amounts — you can monitor traffic quality on your side and extract a ton of useful insights to optimize your setups. And revshare will finally make sense.
Pretty cool, right?
Now tell me — did you actually understand these two posts?
And did you get how powerful this approach is for working with gambling through a tracker?
Drop a comment.
If you got it — write it.
If you didn’t — write it too, I’ll explain again😘
So what do we do with gambling? We use a goal-based model.
The core idea is to create separate flows for different types of conversions.
Simple example:
How this model works
When a user registers:
When the user makes an FTD:
Advantages of this approach:
When working with revshare, this second (non-classic) model is the standard.
With RS:
Each payment has:
This allows you to see:
Goal limits
The number of goals depends on the system:
Even with goals, you still use statuses.
Registration example:
Deposit example:
Deposit tracking parameters:
Grouping data by customer
If you’re using a multi-lead model (registration, deposits, etc.), it’s best to group data by Customer ID passed by the affiliate network.
If you group by click — you’ll get a massive table with thousands of clicks, and only a small portion will actually contain useful data.
If you group by Customer ID — the table will show:
This gives you the full picture of how a user interacts with your system.
Long story short: besides the obvious boost in stats accuracy, if the advertiser shares deposit amounts — you can monitor traffic quality on your side and extract a ton of useful insights to optimize your setups. And revshare will finally make sense.
Pretty cool, right?
Now tell me — did you actually understand these two posts?
And did you get how powerful this approach is for working with gambling through a tracker?
Drop a comment.
If you got it — write it.
If you didn’t — write it too, I’ll explain again
Please open Telegram to view this post
VIEW IN TELEGRAM
WANNA CAPTCHAS?! 🔄
Yep, the topic is too old, the functions are yet bit older too😦 But you'll do need it, so check it out! 👇
AlterCPA Pro / Cloud allow you to add a captcha to the registration page to protect the form from bots while preserving a convenient self-registration experience for real users.
To use this feature, you need to register in Google reCAPTCHA service (you will need an account):
1. Go to the registration page.
2. In the reCAPTCHA type, select Challenge (v2) → "I'm not a robot" Checkbox.
3. Specify your network domain (you can use it as the label as well).
4. Accept the terms of service and click Submit.
You will receive a Site key and a Secret key.
Further setup is done on the AlterCPA side:
1. Go to Control → Settings → Basic → Registration and referrals.
2. Check the box Enable reCAPTCHA v2 on the registration page.
3. Enter the previously obtained Site key and Secret key into the corresponding fields.
4. Save the settings.
Yep, the topic is too old, the functions are yet bit older too
AlterCPA Pro / Cloud allow you to add a captcha to the registration page to protect the form from bots while preserving a convenient self-registration experience for real users.
To use this feature, you need to register in Google reCAPTCHA service (you will need an account):
1. Go to the registration page.
2. In the reCAPTCHA type, select Challenge (v2) → "I'm not a robot" Checkbox.
3. Specify your network domain (you can use it as the label as well).
4. Accept the terms of service and click Submit.
You will receive a Site key and a Secret key.
Further setup is done on the AlterCPA side:
1. Go to Control → Settings → Basic → Registration and referrals.
2. Check the box Enable reCAPTCHA v2 on the registration page.
3. Enter the previously obtained Site key and Secret key into the corresponding fields.
4. Save the settings.
Please open Telegram to view this post
VIEW IN TELEGRAM
❤1👀1