Vertical: SOI sweepstakes
Offer: Decathlon/Mediamarkt/ Worten giftcards
GEO: IT, DE, ES
Period: September of 2018 - June of 2019
Traffic source: FB
Network: BBA (Big Bang Ads)

Important introduction.
By the time we decided to start SOI sweepstakes, we were doing really solid volume on financial offers (crypto/binary options). We never considered sweepstakes as a good offer simply because we did not believe they could be profitable with such low payouts and we never met any people who run them. Once we met a guy from Israel who was making around $1 000 000 monthly on sweepstakes. He did not share any insights but we were dead sure that ther's a lot of money in this vertical and we should try it. So we did....

A good reminder.
As a big fan of testing I consider SOI sweeps a great vertical because low payouts mean less money for tests to gain statistical significance. In crypto you need to burn few hundreds to understand if the lander is good or not while in sweeps you can understand it with 5 dollars.

Traffic game itself:
The funnel was really simple and "classic" on the one hand but one the other hand we literally did hundreds of tests to polish it to the point where we had these crazy ROIs. By "classic" I mean the following: you target specific age on FB and run these users to a lander page which "explains" why these people are special and why they are about to receive the giftcard. The lander in its core is a 3 questions quiz (to engage users). If you do all correctly, you should have the lander CTR of around 60-80%. The pixel was placed on the last answer for the quiz. So FB would optimize for people who got enaged into the lander and wants to see the offer page.

Now comes the interesting part. CPL we were getting on Facebook was around 2-4! cent. This is something you can hardly imagine as CPC these days when you run EU on any traffic source (even push). Anything below 300% ROI was considered as a failed campaign for us and we would stop it and re-setup.

Also worth mentioning:
We started working with few popular CPA netoworks, split testing them, looking for higher payouts and lower "shaving" but from our crypto experience we know that real money is made when you work with direct (first hand buyers) avoiding middle men. So we started looking for the "real owner" of sweepstakes. This is how we came across BBA which is the inner network of Egentic, the company who invented sweeps monetization model more than 20 years ago. To our surprise BBA offered us lower payouts but once we started running with them, we started having higher CVR on the offer page and thus higher EPC. So rule learned - do not fall for higher payouts cause they are normally achieved by increasing the "shave rate" on the CPA network side.

Few more useful tools that helped us to make money with sweeps:
Tracker - Funnel Flux
Cloaker - Just Cloak It
Landers and ads copy fast and cheap translations - transey.com
Webflow - site builder
Anti detect browser - MLA
Domains - namecheap
Hosting - Digital ocean

Some outdated (maybe not) information on Facebook accs. Most of the accounts were coming from our own farm. These are few things we believe helped us to gain more trust and make account live and spend longer.

1) We were using real residential proxies. We did not buy them anywhere online. We rented a small apartment in the middle of the densely populated city district than has few ISPs and we made a deal with ISPs to have multiple clean residential IPs "delievered" to the apartment where we had a server that made it possible to farm FB accounts on these residential IPs.
2) We had our own 4G farm with 2 major carriers.
3) We did not use cheap 1-2 dollar domains. We used aged domains from namecheap (min 6 months).
4) We had real people farming our accounts for 2 weeks minimum (scrolling the news feed, answering the messages, playing games).
5) We always used real plastic cards as a payment method. (we almost had serious problem with tax authorities because of that but that's another story) On the one hand we had a super high (compared to online sellers) cost of the account before it starts spending, but on the other hand it was worth it while we were able to keep these crazy ROIs of 300-1000%.

If you ask me would I start the same set up for farming again, then my answer is "NO"=).
Back then we had energy and desire to invest that much into farming.