Geographic Framework: Why AI Multiplies Hub Advantages—and How Transcend Captures Them
This framework explains how we identify winners before they become obvious. The decade's data reveals a clear pattern: cost advantage without domain knowledge produces outsourcing traps. Cost advantage with F2P expertise—monetization design, LiveOps mastery, player psychology—produces $5B+ exits. AI amplifies this gap.
(F2P IQ + Institutional Knowledge) × AI Leverage = Output
Result: The gap WIDENS, not narrows. AI democratizes production; it does not democratize success.
Dream Games at Seed (2019): 5 Peak Games executives with puzzle genre expertise, attacking the same category they knew, leveraging Turkey's cost window + innovation edge + accessible acquirers. Result: $5B valuation in 5 years.
What We Look For Today: Proven team from successful studio (Peak, Supercell, Playtika, King alumni) + AI-native workflow from Day 1 + attacking genre where they have domain expertise + in market with cost window AND high F2P IQ.
The central question for Fund III: Does AI democratize hit-making (benefiting low-cost markets) or compound existing advantages (benefiting established hubs)? We've analyzed this extensively. Conclusion: AI compounds, it does not equalize.
| Market Attribute | Pre-AI Importance | AI Impact |
|---|---|---|
| F2P IQ (monetization knowledge) | Critical | AMPLIFIED – AI multiplies winning formulas |
| Execution IQ (code/art) | Important | COMMODITIZED – AI makes this table stakes |
| Seeding Events (alumni networks) | Required | PARTIALLY SUBSTITUTABLE – AI transfers explicit knowledge, not tacit |
| Cost Arbitrage | Useful | CHANGED – Value shifts to "AI direction ability" not raw cost |
| English Fluency | Important | REDUCED – AI translation near-instant |
Our Fund III AI thesis identifies verified productivity gains across the value chain:
Critical insight: These gains MULTIPLY existing F2P IQ advantage. A Turkish studio with 10 years of puzzle monetization expertise iterates 4x faster on proven winning concepts. A Vietnamese studio iterates 4x faster on cloned concepts without the monetization DNA. The gap widens.
| Scenario | Probability | Implication |
|---|---|---|
| AI Compounds (Turkey/Israel extend lead) | 60% | High F2P IQ × AI = faster iteration on winning concepts. Our core thesis holds. |
| AI Commoditizes Production (tools win) | 25% | We're positioned on supply side via Kinoa investment. Tool providers capture value; our portfolio companies gain leverage. |
| AI Democratizes (Vietnam/India breakout) | 10% | Requires AI to solve "Product Taste" gap. LLMs are regression-to-mean engines. |
| Big Tech Captures | 5% | Innovator's Dilemma. Big Tech uses AI for cost cuts; startups use AI for risk-taking. |
At 60% probability that AI compounds existing hub advantages, we maintain primary allocation to Turkey/Israel. For the 25% scenario where AI tools capture value, we're positioned on the supply side via our Kinoa investment. Emerging markets (Vietnam/India) remain "Watch" for Fund III—we monitor for AI-native breakout signals that could justify Fund IV allocation.
This analysis covers ex-China mobile gaming markets investable by Western VCs:
Canada and UK gaming studios are grouped with US ecosystem due to tax incentive structures that attract US capital, frequent US ownership, and exit paths flowing through US/EU acquirers.
China is excluded from investable universe due to CFIUS constraints for US-based funds, state-adjacent industrial model, and acquirer misalignment (Chinese acquirers create exit paths incompatible with Western LP structures).
The standard narrative claims: "Low labor costs create gaming hubs. Find cheap developers, build studios, profit." This thesis has driven capital into low-cost markets for 15 years—and it has consistently failed to produce exits.
| Market | COLA Index (vs NYC=100) | Gaming Value ($B) | Top 50 Global Studios |
|---|---|---|---|
| Vietnam | ~25 (lowest) | ~$0.1B (gaming IP) | 0 |
| India | ~28 | $0.6B (mostly RMG) | 0 |
| Turkey | 56 (senior dollarizing) | $2.2B+ realized, $5B+ total | 3+ |
| Israel | 75 (offshore model) | $15B+ | 5+ |
| Finland | 85 | $12.5B+ | 3+ |
Vietnam and India maintained cost advantages for 15+ years. Neither produced globally-scaled mobile gaming IP with sustainable enterprise value. Turkey—higher absolute costs but meaningful junior/mid-level arbitrage—created $5B+ in total ecosystem value ($2.2B+ realized exits). Finland—among the highest costs in Europe—generated $12.5B+. Cost predicts nothing. F2P IQ predicts everything.
Cost is necessary but not sufficient. A cost window creates temporal opportunity—nothing more. Without an innovation edge, cost advantage produces:
| Factor | Turkey | Vietnam |
|---|---|---|
| Cost window | Yes (35-40% at junior/mid) | Yes (lowest globally) |
| Innovation edge | Yes (puzzle/casual attack) | Limited (volatile hits) |
| Seeding event | Peak Games ($1.8B, 2020) | VNG = Tencent affiliate (43% shares) |
| Acquirer gravity | US/EU/MENA accessible | China (CFIUS complications) |
| Enterprise value | $5B+ total | Minimal retainable |
Revenue per employee reveals the true story—high F2P IQ + small team = extraordinary efficiency. This is what AI amplifies:
| Studio | Revenue/Employee | Notes |
|---|---|---|
| Supercell (Finland) | $8.8M | Gold standard efficiency |
| Dream Games (Turkey) | $5.4M | Single-game focus, peak efficiency |
| Peak Games (Turkey) | $2.1M | Pre-AI benchmark |
| VNG (Vietnam) | $20K | Publishing model (no IP leverage) |
Key insight: The 400x gap between Supercell ($8.8M) and VNG ($20K) is not cost—it's F2P IQ. AI multiplies Supercell's efficiency further; it doesn't close the gap.
| Finding | Correlation | p-value | Status |
|---|---|---|---|
| Domestic Market Trap | r = -0.845 | 0.001 | ROBUST |
| Cost Alone Insufficient | r = +0.724 | 0.012 | ROBUST |
| Outsourcing Trap | r = -0.582 | 0.060 | Notable |
| B2C Orientation | r = +0.566 | 0.069 | Weak |
| Seeding Events | r = +0.837 | 0.001 | CIRCULAR (defined by outcomes) |
Note: Seeding Events correlation is high but circular—seeding events are often identified retrospectively. We use this as pattern recognition, not prediction.
The Lifecycle Model explains why some cost windows convert to exits while others stagnate. Three components must align:
A cost window is a temporal opportunity created by cost advantage relative to incumbent markets. Windows close as wages rise post-success—Israel developer salaries increased from ~$55K to ~$110K+ after validation.
Senior Unity/Unreal engineers: Wages increasingly USD-pegged to retain talent
Junior/mid-level talent: Significant arbitrage remains (30-40% advantage)
Operational expenses: Full arbitrage benefits (rent, utilities, support staff)
The innovation edge is what you're attacking—the genre, platform, or mechanic that creates differentiated value.
| Era | Innovation Edge | Winners |
|---|---|---|
| 2010-2012 | Mobile-first (vs. web/console) | Supercell, Rovio (Finland) |
| 2012-2015 | F2P monetization + marketing science | Playtika, Moon Active (Israel) |
| 2014-2018 | Puzzle/casual genre depth | Peak Games, Dream Games (Turkey) |
| 2017-2020 | Hypercasual + UA optimization | Rollic (Turkey), Voodoo (France) |
| 2020-2022 | Web3/P2E (volatile) | Sky Mavis (Vietnam—unsustained) |
| 2024+ | AI-enhanced development | TBD |
Value flows TO where acquirers are headquartered, not where development teams sit.
| Acquirer Region | Key Players | Notable Acquisitions |
|---|---|---|
| US | Zynga, EA, Take-Two, AppLovin | Peak Games ($1.8B), Rollic ($168M+), Playfish → EA ($400M) |
| UK | King (Activision) | King → Activision ($5.9B) |
| Saudi Arabia/MENA | Savvy Games Group (PIF) | Scopely ($4.9B), EA stake ($2.1B) |
| Singapore | Sea Ltd (Garena) | SEA publishing, Moonton |
| China | Tencent, NetEase | VNG stake, SEA publishing |
| Korea | NCsoft, Krafton, Netmarble | Indygo/Lihuhu ($103.8M, 2025), Kabam ($800M) |
Savvy Games Group emerged as the most aggressive acquirer in 2023-2025. Key implications: Geography-agnostic capital expands options for Turkish studios. Geopolitical considerations may limit Israeli-Saudi deal flow, reinforcing US acquirer gravity for Israel.
| Cost Window | Meaningful (30-40% at junior/mid-level; senior wages dollarizing) |
| Innovation Edge | Proven (puzzle/casual heritage) + AI capability emerging |
| Acquirer Gravity | Excellent (Zynga, King, Take-Two; Savvy Games compatible) |
| Phase | Growth → Maturation |
| Value Creation | $5B+ Total ($2.2B realized; $2.75B unrealized incl. Dream Games) |
Turkey is performing like early-stage Israel but priced like Poland—mispriced relative to our framework. Peak Games exit ($1.8B to Zynga, 2020) created deep talent network. The "Peak Alumni Cluster" now drives Dream Games ($5B valuation), Spyke Games, Ace Games, and 20+ studios. We have first-call relationships with this network.
Critical timing: Peak Games succeeded (2020 exit) before Turkish gaming subsidies began (2022-2024). Turkey's gaming ecosystem is organic—built on talent networks and F2P expertise, not government incentives. The subsidies arrived after the flywheel was already spinning. This validates our thesis: cost window + F2P IQ + accessible acquirers drive value creation, not policy intervention.
Turkish tech ecosystem increasingly overlaps with gaming. We've seen direct evidence: one Istanbul studio reduced background art team from 5 to 1 using generative AI, reallocating budget to UA. AI × Turkish F2P IQ = the next Dream Games.
| Cost Window | Moderate (offshore team structures maintain advantage) |
| Innovation Edge | Proven (social casino, marketing science) + strong AI/ML capability |
| Acquirer Gravity | Excellent (US-dominant; Savvy complicated but alternatives abundant) |
| Phase | Maturation |
| Verified Exits | $15B+ (Playtika: $4.4B acq 2016, $11B IPO 2021; Moon Active $5B+) |
Mature ecosystem with $15B+ in proven exits. Higher entry valuations are offset by dramatically lower execution risk—we know what works here. Marketing science heritage (Unit 8200 veterans, adtech) transfers directly to AI/ML capability.
Israel has the strongest AI advantage in gaming. Deep bench of data scientists from adtech (ironSource, AppsFlyer) and gaming analytics (Playtika) enables predictive UA, dynamic pricing, and personalization at scale. These teams adopted AI tools first because they already think in data terms.
| Cost Window | None (among highest in Europe) |
| Innovation Edge | Heritage (mobile-first pioneers: Supercell, King, Rovio) |
| Acquirer Gravity | Proven exit paths to US/UK acquirers |
| Phase | Consolidation |
$12.5B+ in exits proves the ecosystem works. But cost window closed years ago—Supercell alumni command premium salaries. Investment thesis is pure team quality and IP potential, not geographic advantage. Reserve for exceptional second-time founders from Supercell, King, or Rovio alumni networks.
| Cost Window | None (highest global costs; UK cost window closed post-King acquisition) |
| Innovation Edge | Created the category (US: Zynga, Kabam, Scopely; UK: King, Playfish); now consolidation phase |
| Acquirer Gravity | SOURCE of acquirers (Zynga, EA, Take-Two, AppLovin, Activision) + target for Savvy/PIF ($4.9B Scopely) |
| Phase | Consolidation (acquirers, not startups) |
The US and UK created mobile F2P gaming—Zynga, Kabam, Scopely, Jam City (US) and King, Playfish (UK) pioneered the category. But these markets are now where acquirers sit, not where startups with cost advantages emerge. "Avoid the middle": we don't invest in average US/UK opportunities. Premium valuations + no cost arbitrage = only exceptional founding teams with unique defensibility warrant allocation. Middle-tier deals in these markets offer neither the upside of emerging hubs nor the risk mitigation of proven ecosystems.
We know these markets from the inside: Andrew Sheppard led Kabam's growth as VP Marketing during its scale from startup to $400M+ revenue and eventual Netmarble acquisition. Shanti Bergel built Playfish's marketing through its acquisition by EA. We understand why US/UK work for consolidators but not for seed-stage geographic arbitrage.
| Cost Window | Strong (-28% COLA, RU/BY exodus adding talent) |
| Innovation Edge | Emerging (PC/console heritage transitioning) |
| Acquirer Gravity | Good (EU/US access, Savvy active) |
| Phase | Emergence → Growth |
RU/BY developer exodus (2022+) is potential seeding event—experienced talent relocating to Warsaw, Krakow, Prague. Mobile-native breakout hit needed to validate ecosystem transition. Fund IV primary candidate if validation emerges.
| Cost Window | Strongest (lowest global costs, 15+ years) |
| Innovation Edge | Present but volatile (Flappy Bird, Axie—unsustained; Lihuhu acquisition promising) |
| Acquirer Gravity | Evolving (NCsoft $103.8M acquisition Dec 2025 provides Korean path; VNG remains Tencent-linked) |
| Phase | Early Emergence (first notable M&A exit Dec 2025) |
VNG operates as Tencent's primary publishing affiliate in Vietnam (43% shares, 23% voting rights). Vietnam remains a low-cost, high-execution gaming market that has not yet demonstrated a repeatable, globally scalable IP flywheel. While Hanoi plays an important role in policy, telecom infrastructure, and state AI research, Vietnam's true commercial gaming production, export, and executive leadership hub is Ho Chi Minh City—home to VNG, Sky Mavis, Amanotes, Sparx*/Glass Egg (Virtuos), and most export-focused studios. Flight time Shenzhen ↔ HCMC is 2h45m ("commuter route")—explaining satellite status.
NCsoft's $103.8M acquisition of Indygo Group (Lihuhu's Singapore-based parent) represents the first traditional M&A exit for Vietnam-origin gaming IP—modest by global standards (compare Peak Games at $1.8B), but notable as proof of exit path. Unlike Axie's token-based volatility, this is traditional M&A with a non-Chinese acquirer. The Singapore holding structure validates the "M&A-ready" corporate architecture path. However, one acquisition does not create an ecosystem—Turkey required Peak Games plus 20+ alumni studios before the flywheel validated. Vietnam now has Domino #1; we monitor for Domino #2.
Brett's Vietnamese fluency enables monitoring HCMC's gaming and AI dev scene before Western VCs arrive. We track Amanotes-style casual exporters and emerging hybrid-casual studios pivoting to midcore with AI-native teams. The NCsoft/Indygo deal increases our monitoring intensity on Lihuhu alumni movement.
| Cost Window | Strong (low costs) |
| Innovation Edge | Domestic only (RMG/fantasy sports) |
| Acquirer Gravity | Galapagos market—only Korean bridge (Krafton) |
| Phase | Isolated from Asian supply chain |
"Galapagos" market—isolated entirely from Asian B2B networks. No organic connection to Vietnam, Thailand, or China post-2020. Krafton is the ONLY bridge. Large domestic market (600M+ gamers) creates "local trap" with low ARPU. Enterprise/service DNA (Infosys, Wipro) doesn't translate to F2P product creation.
Team member's ethnic background + cultural familiarity enables identifying outliers breaking domestic trap. We watch for Games24x7 or Dream11 alumni founding AI-native global studios.
| Cost Window | Moderate (LATAM competitive) |
| Innovation Edge | Wildlife only ($3B exit, isolated success) |
| Acquirer Gravity | US accessible but no ecosystem |
| Phase | Pre-ecosystem |
Creative talent exists but business environment (taxes, labor law) makes scaling difficult. Wildlife ($3B exit) remains the only major success; no ecosystem formed around it. Wildlife alumni network not yet visible.
Portuguese is one translation hop from Spanish (our LP network in LATAM). AI translation reduces linguistic barrier significantly.
LPs often ask: "Why not China? Korea? Japan?" These are massive gaming markets. We've thought carefully about this—and in some cases, have direct operating experience that informs our view.
| Market Size | Largest globally (~$50B mobile) |
| Innovation | World-class (Tencent, NetEase, miHoYo) |
| Why Not | CFIUS constraints, state-adjacent model, LP structure incompatibility |
Brett worked with FunPlus in China—we understand the market's sophistication firsthand. Chinese studios are world-class at F2P design; they're not a "watch list" opportunity, they're a fundamentally different investment universe. CFIUS restrictions make direct investment impossible for US-based funds. Chinese acquirers create exit paths incompatible with Western LP structures. This isn't ignorance—it's structural incompatibility.
We monitor China for innovation trends that will reach Western markets. Chinese studios often pioneer mechanics (gacha, battle pass, social features) that Turkish and Israeli teams then adapt for Western audiences. Our Mandarin-speaking team member tracks this flow.
| Market Size | Top 5 globally (~$8B mobile) |
| Innovation | Strong (Krafton, Netmarble, NCsoft) |
| Why Not Primary | Mature ecosystem, domestic consolidation, limited seed opportunities |
Korea's gaming ecosystem is mature and domestically consolidated. The major players (Krafton, Netmarble, NCsoft, Kakao) dominate deal flow. Seed-stage opportunities are rare; when they exist, Korean VCs have first access. The value for Transcend is as LP relationship (Krafton) and co-investment pipeline, not as primary allocation target.
Krafton is our LP—we co-invest on deals they source. Shanti and Brett's Japanese fluency + regional relationships create access Western VCs lack. But we're a co-investor here, not a lead.
| Market Size | Top 3 globally (~$20B mobile) |
| Innovation | IP-driven (Sony, Nintendo, Bandai Namco, Sega) |
| Why Not Primary | Domestic focus, IP-centric model, cultural barriers to Western integration |
Japan's mobile gaming success is built on domestic IP (anime, manga, console franchises). The model doesn't translate to seed-stage F2P investing—Japanese studios optimize for Japanese audiences with Japanese IP. Global expansion is an afterthought, not a strategy. Western-style F2P mechanics are often culturally adapted in ways that limit global portability.
Bandai Namco is our LP—we access Japanese studios through this relationship. Shanti and Brett speak Japanese, enabling direct relationship building. But like Korea, we're positioned for co-investment and partnership, not lead investment.
The US invented mobile F2P gaming. Zynga (founded 2007) pioneered social gaming monetization on Facebook. Kabam (founded 2006) proved mid-core F2P could work on mobile with games like Kingdoms of Camelot. Scopely (founded 2011) scaled to $4.9B exit to Savvy Games (2023)—though notably with significant international talent (Barcelona, Dublin, Israel studios).
The Scaling Story: Kabam grew from startup to $400M+ annual revenue before being acquired by Netmarble ($800M, 2017). This era established the playbook—F2P monetization, whale economics, LiveOps cadence—that every subsequent market learned from. Scopely's success reinforces the pattern: US HQ for capital and acquirer access, international talent for cost efficiency.
Key Insight: No cost window, but first-mover innovation edge. US defined the rules, then other markets (Finland, Israel, Turkey) executed with cost advantages. Today the US is where acquirers sit (Zynga, EA, Take-Two, AppLovin, Savvy/PIF), not where seed-stage arbitrage exists. Transcend's Andrew Sheppard led Kabam's growth; Shanti built Playfish through EA acquisition—we understand this market from the inside.
Nokia's collapse (2011-2013) released thousands of engineers. Rovio's Angry Birds (2009) and Supercell's Clash of Clans (2012) demonstrated mobile-first could exceed console heritage.
Key Insight: Nokia collapse created forced seeding event. Small domestic market (5M) forced immediate global focus. Innovation edge was mobile-first execution before incumbents recognized platform shift. Finnish puzzle games demonstrated D30 retention 15-20% above global average.
Lowest costs globally (~35,000 developers, ~400 active game companies). Strong talent. VNG became first Vietnamese unicorn—but as Tencent's primary publishing affiliate (43% shares, 23% voting rights). VNG is headquartered in Ho Chi Minh City, Vietnam's true commercial gaming hub.
City-Level Dynamics: Ho Chi Minh City hosts the largest concentration of work-for-hire, art-outsourcing, and export-focused studios: VNG, Sky Mavis (Axie Infinity), Amanotes (#1 global music game publisher, 3.5B+ downloads), Sparx*/Glass Egg (Virtuos, ~1,000 employees combined), Gameloft Saigon, and Lihuhu (Indygo Group). Hanoi serves as a secondary hub anchored by publishers (Appota/Gamota, SohaGame, VTC) and state-linked policy/telecom functions—not the commercial gaming center.
The Sustainability Problem: Vietnam produced genuine innovations. Flappy Bird (2013) was global phenomenon but creator shut it down—no enterprise value captured. Sky Mavis pioneered P2E with $4.6B peak, but token collapse reduced value by >90%—volatility, not sustainability.
December 2025 Inflection: NCsoft's $103.8M acquisition of Indygo Group (Lihuhu parent, Singapore-domiciled) marks the first traditional M&A exit for Vietnam-origin gaming IP—modest by global standards, but proof that a non-Tencent exit path exists. The critical test: whether Lihuhu alumni spawn new studios (replicating Turkey's Peak Games pattern) or remain within NCsoft.
Key Insight: Vietnam's first notable M&A exit signals potential phase transition from satellite trap to independent hub. However, one acquisition does not validate an ecosystem—Turkey required Peak Games ($1.8B) plus 20+ alumni studios before the flywheel produced Dream Games ($5B). Cost window present 15+ years; NCsoft/Indygo demonstrates the exit path exists. The Tencent–VNG relationship remains the dominant structural force for most Vietnamese studios. We monitor Lihuhu alumni movement as the key indicator for Fund IV allocation.
Social casino gaming required sophisticated marketing analytics—CAC/LTV optimization, whale identification. Israeli ecosystem (Unit 8200 veterans, adtech) had competitive advantage.
Key Insight: Innovation edge was marketing science applied to gaming. Studios competed on player acquisition and monetization optimization, not pure game design. Cost window via offshore team structures. Acquirer gravity firmly US (Playtika: $4.4B acquisition 2016, $11B IPO 2021).
Peak Games (founded 2010) attacked Facebook/mobile casual games. Currency dynamics created cost window during F2P expansion.
Key Insight: Peak Games founder Sidar Sahin and team created "Peak Alumni Cluster"—experienced operators now at Dream Games, Spyke Games, Ace Games, and 20+ studios. Innovation edge was puzzle/casual genre depth with superior retention mechanics. Cost window + innovation edge + accessible acquirers = $5B+ enterprise value.
Geographic allocation frameworks are worthless without execution capability. Here's why Transcend has unique access to execute this thesis:
| Market | Transcend Edge | Competitive Moat |
|---|---|---|
| Turkey | Peak/Dream alumni network via portfolio | First-call access to second-time founders |
| Israel | Unit 8200 veteran deal flow; existing relationships | Deep ecosystem presence |
| Japan/Korea | Shanti + Brett Japanese fluency; Krafton/Bandai Namco LP relationships | Co-investment access Western VCs lack |
| Vietnam | Brett's Vietnamese fluency + AI-powered deal flow scanning | Early signal detection before Western capital |
| India | Team member ethnic background + cultural network | Ability to identify outliers breaking domestic trap |
| Greater China | Team member Mandarin fluency + ethnic network | China-adjacent deal monitoring |
| Poland/EU | EU proximity; RU/BY exodus monitoring; Brett's German/French | Talent migration tracking |
Gaming talent doesn't distribute evenly across countries—it clusters in specific cities where seeding events occurred. Our network maps to these hubs specifically:
This granularity matters: a "Turkish gaming company" headquartered in Ankara is a red flag. The talent network is in Istanbul. Our relationships are hub-specific, not country-level.
Important limitation: Transcend's regional expertise is genre-specific—we source mobile F2P opportunities, not all gaming. Our Turkish network, for example, is deep in puzzle/casual mobile but has limited reach into PC/console development. When we've looked for Turkish talent in PC/console, the network doesn't extend there—the talent pools are largely separate.
This is a feature, not a bug: mobile F2P is our thesis. But LPs should understand that "Turkish gaming expertise" means "Turkish mobile F2P expertise." The same applies across our regional networks—they're built around mobile-first alumni clusters, not general gaming industry contacts.
We use AI extensively in operations—the same 4-5x productivity gains we seek in portfolio companies:
This is not theoretical; it's how we operate. Our thesis about AI multiplying productivity is validated daily in our own workflow.
Our LP base creates proprietary deal flow unavailable to competitors:
These relationships create "first look" opportunities in markets where language and relationships gate access.
Natural Hedge Structure: Portfolio companies earn revenue in USD/EUR. Operational expenses (rent, utilities, support staff) denominated in local currency. Labor increasingly bifurcated—senior talent often dollarized, junior/mid arbitrage remains.
Yield Floor Defense: The thesis assumes a 2027+ M&A supercycle. If that window closes or delays, AI-native studios with 40-50% operating margins can distribute profits as dividends. Unlike traditional gaming studios operating at 20-25% margins with reinvestment pressure, high-margin AI-native studios generate meaningful free cash flow that can return capital to investors independent of exit timing. This yield floor provides downside protection if M&A markets remain constrained.
| Tier | Markets | Allocation | Rationale |
|---|---|---|---|
| PRIMARY | Turkey | 25-30% | Peak/Dream ecosystem + cost window + AI compounds F2P IQ |
| PRIMARY | Israel | 20-25% | Playtika/Moon Active ecosystem + Unit 8200 talent + proven exits |
| SECONDARY | Finland/Sweden | 15-20% | Supercell/King alumni + high F2P IQ (high cost offset by quality) |
| EXCEPTIONAL | USA & UK | 0-5% | "Avoid the middle"—category pioneers now acquirer markets; cost windows closed; exceptional teams only |
| OPPORTUNISTIC | Poland | 5-10% | Watch for mobile-native breakout; RU/BY talent influx |
| WATCH | Vietnam/India | 0% | Fund IV candidates; monitoring for AI-native breakout signals |
| AVOID | Canada | 0% | Branch office economy; no change expected |
Vietnam and India remain "Watch" for Fund III. For Fund IV, we'll evaluate AI-native micro-studios meeting these criteria:
Why Not Now: These markets lack the F2P IQ infrastructure and acquirer gravity required for Fund III returns. We invested in Kinoa to own the supply side of AI tools—if AI eventually enables emerging market studios to close the F2P IQ gap, we're positioned as the tool provider, not betting against our own thesis.
| Market | Current Status | Trigger for Escalation |
|---|---|---|
| Poland | Watch 5-10% | Mobile-native Top 50 hit; or AI-native breakout from RU/BY exodus talent |
| Vietnam | Watch 0% | Partial (Dec 2025): Non-Tencent exit achieved (NCsoft $103.8M). Remaining: AI-native US Top 50 hit; Lihuhu alumni flywheel formation |
| India | Watch 0% | Tier-1 geo revenue majority; Krafton-style bridge |
| Brazil | Watch 0% | Wildlife alumni ecosystem; sub-5-person unicorn |
| Saudi Arabia | Monitor | Savvy Games portfolio graduates; sovereign wealth patience |
If AI commoditizes production quality, the market becomes a pure probability game:
We monitor for signs that "good enough" AI-produced games become indistinguishable from "great" games. If this happens, lowest-COLA markets win.
Current Assessment: Not yet. F2P IQ (monetization design, LiveOps) remains non-commoditized. AI amplifies expertise; it doesn't replace it. But if AI tools do eventually solve the "Product Taste" gap, we're positioned on the supply side via Kinoa—we'd rather be the company providing those tools than betting against them.
Data Sources: Numbeo COLA Index, data.ai, Sensor Tower, CB Insights, Pitchbook, Company filings