How We Scaled an iGaming Offer Across Tier 2-3: A Real Casino Promotion Case Study

I've been buying traffic for iGaming offers across Tier 2-3 for close to three years now - running 1WIN, Mostbet, and Pin-Up in Africa, Southeast Asia, and Latin America. We got to $4k a day at 120% ROI, and here's exactly what it took.

01The Start: Geo, Offer, and First Hypotheses

We started with a $500 test budget on Indonesia and South Africa as the geos - the offer was a casino app with free spins on the first deposit, paid on a CPA model at $8-12 per first depositor, or RevShare at 30% of the loss. Links went straight to the operator's landing page, no pre-landers in between - conversion runs higher that way.

The first channel hypotheses were Facebook Audience Network (FAN) and TikTok Ads. FAN delivered cheap clicks - $0.03-0.05 - but the traffic quality didn't hold up: almost no deposits came through. TikTok started out at a $2-3 CPM with reach into a young audience, but the geo-targeting worked poorly, and the $500 budget burned through in a week at -40% ROI.

After the first week's failure, we switched approach: dropped FAN for Google UAC - these are App campaigns now, but the name 'UAC' stuck among buyers - and got app installs at $0.4-0.7, though conversion to deposit came in lower than expected: 5-7% against 10-12% on TikTok. Telegram Ads came in as a supplementary cheap-reach channel at $0.3-0.5 per subscriber, but only 2-3% clicked through the link. By the end of the second month, the budget had grown to $3k a day at 70% ROI.

02Channels and Budget Split

After the first week, the budget got reshuffled: 60% into TikTok Ads, 30% into Google UAC for app installs, 10% into Telegram Ads with hand-picked channels. On TikTok, the target was deposit-leads through the pixel; on UAC, install plus first session; Telegram Ads ran narrowly, through referral channels, at a CPA of $0.5 per subscriber.

Average CPC on TikTok ranged $0.12-0.35 depending on the creative and time of day, with CPM holding at $1.8-3.2. Google UAC delivered a CPI of $0.4-0.7, but conversion to deposit ran 30% lower than TikTok's, and Telegram delivered cheap subscribers ($0.3-0.5), of whom only 2-3% clicked through the link. By the end of the first month, the budget had grown to $3k a day at 70% ROI.

In the second month, TikTok scaled to $2.5k a day and immediately ran into bans - three accounts lost in a week to overly aggressive creatives, which meant keeping 7-10 accounts in rotation, warmed up on organic activity first. Google UAC got cut back to 15% of the budget: it was bringing in too many non-target installs, 60% of them kids or people with no payment method. Telegram stayed on, but only for referral programs.

03The Funnel in Numbers: From Click to Deposit

The funnel ran on Keitaro with postbacks from the operator: average ad CTR held at 1.8-2.5%, and 70-80% of clicks actually made it to the landing page - the rest dropped off from slow page loads, especially in South Africa. On the landing page, 25-35% filled out the registration form, and 40-50% of those who registered made a first deposit. End-to-end conversion from click to deposit came out at 5-8%.

CPA on Tier 2-3 ranged $9-14, and with an average deposit of $35 and 30% RevShare, LTV came out around $50 per player - margin held at 70-80% after traffic costs. We tried not to lose late conversions: the operator's postback could arrive with up to a 2-hour delay, so the attribution window stayed at 24 hours.

While scaling toward $4k a day, we started segmenting traffic - by source (TikTok, Google, Telegram) and by creative type (UGC, influencer, stock) - which made bid management much more precise: on UGC with influencers, CPC held at $0.15-0.25 with 12-15% conversion, while stock creatives pushed CPC up to $0.4 and conversion fell to 6%.

04What Actually Drove Growth

The first thing that actually worked was live UGC: 'players' at the roulette table shot on an iPhone, subtitles added in the local language - Swahili, Indonesian, Portuguese - and those clips ran a 40% lower CPM than stock footage. The second was GenAI for mass-producing variations - background, close-up, color grading; a day's work turned out 50-100 creatives, each tested on small budgets of $10-20.

The third was collaborations with local TikTok micro-influencers, 10-50k followers: an integration cost $20-50 a post, and conversion from those posts ran 2-3x higher than standard ads - audience trust was doing the work. The fourth was segmenting depositors and retargeting the ones who registered but never put money down: one extra touch, a push notification with a free-spin offer, added +15% deposits.

The fifth was optimizing ad-serving hours: on TikTok in Indonesia, the best CTR landed at 19:00-23:00 local time, when CPM dropped to $1.5-2.0 and CR climbed to 14%; in South Africa, the peak shifted to 18:00-21:00. We restricted delivery to those windows alone, and CPA dropped 20%.

05What Didn't Work - Honestly

Google UAC on Tier 2-3 turned out to be an expensive, messy channel: a $0.4-0.7 CPI, but 60% of installs landed on a non-target audience - kids and people with no access to payments - and the channel got switched off after a month. Telegram Ads did fine for cheap reach, but conversion to deposit was negligible, so it stayed in the mix for branding only.

Another mistake was creatives too aggressive with direct copy like 'deposit - bonus': TikTok would block those within 2-3 hours and ban the account - three accounts lost that way in a single week. We moved to a softer approach - video showing the game in play, with no direct call to deposit.

The retargeting hypothesis through the Google Display Network also failed: CPM $5-8, conversion 1-2%, and a $500 budget burned through in 3 days at -80% ROI. Animated creatives didn't land either - pricier to produce ($50-100 apiece) and running a noticeably higher CPM than UGC.

06Conclusions and What We Kept

Casino promotion on Tier 2-3 works if the focus stays on local creative and combos get swapped out fast. TikTok remains the main volume driver, but it needs 5-10 accounts in reserve; Google UAC only makes sense for core markets with an audience that can actually pay; Telegram works better as a channel for referral networks than as a source of direct CPA.

We never let ourselves lean on one channel or one creative: we ran through 200-300 creatives a week and cut anything that wasn't hitting 100% ROI within 3 days. We kept Keitaro and its postbacks running the whole time - without them, there's no seeing the real conversion.

One thing I took from this case: if a channel doesn't clear 80%+ ROI on a $300 test, I don't scale it. No result in 3 days, and I swap the combo - we killed 2-3 hypotheses a day until we found the ones that worked.

07FAQ

How much budget does it take to start?

A minimum test runs $500-1,000 per geo, and if ROI clears 50%+ within a week, the budget can double. Scaling steadily up to $4k a day takes roughly $15-20k in working capital.

How do you deal with TikTok bans?

Keep 5-10 accounts warmed up on organic activity, and build creatives with no direct call to deposit: the video centers on gameplay, without showing the offer outright, and the operator's logo never sits in the first frame.

Which tracker works best for iGaming?

Keitaro or Binom - both accept s2s postbacks and calculate ROI in real time. Set the attribution window to 24-48 hours: the postback can arrive with a delay.

Is RevShare worth using?

Yes, for a longer play: RevShare at 30-40% with an average LTV of $50-80 delivers margin above 100%. For quick cash, CPA at $8-15 per deposit fits better.

Ioann Putevoy

Head of Traffic & growth lead. I build AI-first products and take them to market - see the portfolio.

Read next: Traffic Arbitrage in 2026: Funnels, Channels, and Unit Economics · Adult traffic in 2026: channels, compliance and what actually scales · iGaming user acquisition: the acquisition-and-retention playbook · Traffic to OnlyFans: the channels, creative, and funnels that still pull