Where to play

Where to play next

Investor / advisor · 2026-05-10

We are deep in Telegram (home), but the stack is built to port. After 13 ecosystems audited live in May 2026, two near-term candidates stand out, TikTok Minis (post-divestment, 1.9B MAU, codebase portable) and Discord Activities (gaming-creator influencer match). LINE Taiwan + Thailand for year-2. World, WeChat, Douyin as long-bets behind language / partner / sanctions barriers.

The «cross-ecosystem doctrine» follows directly from social graph fragmentation: we don't build a new graph from zero, we surface inside graphs that already exist. The question is which graphs to enter, in what order, and at what cost. We audited thirteen messenger / superapp ecosystems live in May 2026. The shortlist is sharper than we initially thought.

The headline finding

Stay deep in Telegram. Two strong near-term candidates emerged: TikTok Minis (post-divestment, 1.9B MAU, codebase portable) and Discord Activities (gaming-creator pipeline match, IPO narrative). LINE Taiwan-first remains attractive as the third priority. Snapchat is dropped (Snap Minis killed 2022). LinkedIn is demoted from drop to watch-list. VK is a Russia hedge only. Worldcoin is strategic narrative. WeChat / Douyin / Zalo / Kakao are parked behind language, partner, or sanctions barriers.

The candidate list to actively name on the deck has narrowed from six to three or four, recommended: TikTok + Discord + LINE, optionally World or Zalo.

Three biggest surprises from the live research round

Snap Minis was killed in August 2022. Snap is no longer a candidate, they discontinued the HTML5 mini-app surface as part of their 20%-cut. AR Lens economy remains, but it's a separate product category. Drop.
TikTok divestment closed January 22, 2026, Oracle / Silver Lake / MGX / a16z own ~80%, ByteDance ~20%, USDS operates the algorithm. The geopolitical kill-switch that previously kneecapped TikTok-as-target is gone. TikTok Minis is at the same maturity inflection Telegram was in 2022, no entrenched «Apps Pro for TikTok» incumbent. HTML5 + JS bridge directly portable from our codebase.
MiniAI (miniai.vn, Vietnam), direct multi-ecosystem competitor building mini-apps tooling across Zalo + MoMo + WeChat + Telegram + Kakao from Vietnam outward. Vietnam strategy probably needs to be partner-or-acquire MiniAI rather than head-to-head.

Tier-ranked candidate set

Tier 1, Primary focus

Ecosystem	Verdict	Why
Telegram	✅ Stay deep	1B MAU / 500M DAU; 15M Premium subs; $870M H1 2025 revenue (+65%); 3 platform tailwinds in past 5 weeks (Bot API 9.6, TON direct control May 4, Notspy premium-limit lift); home court; vertical-integration risk lowest of any ecosystem

Tier 2, Active near-term candidates (2026 H2 → 2027)

Ecosystem	MAU	Mini-app surface	Port fit	Window
TikTok Minis	~1.9B	TikTok Minis (2024-2025), low-thousands of apps, partner-curated, FYP discovery	HTML5+JS bridge, directly portable from TG codebase	NOW (greenfield)
Discord Activities	230–260M / 30M DAU (94 min/day!)	Embedded App SDK GA Mar 2024, Apps Directory	6 of 10 Apps Pro blocks port directly	2026 H2 (~2 quarters with 2-3 engineers)
LINE Taiwan / Thailand	22M TW / 54M TH (LINE 194M global)	LIFF + LINE MINI App; TW + TH channel creation OPEN to anyone since March 11, 2026	~70% port; needs zh-TW + th localisation, LINE Pay TW adapter	2027 Q1-Q2 (TW first)

Why TikTok jumped to a near-term candidate: divestment removed the kill-switch; mini-app surface is greenfield (no entrenched tooling player); 1.9B MAU is roughly 2× Telegram; codebase ports directly. Highest-EV expansion candidate after Telegram-deepening.

Tier 3, Strategic optionality / long-bets (2027+)

World, ~33M App users, 18M+ verified, #1 self-custody wallet by MAU. But 5+ market bans (Brazil, Kenya, HK, Spain, Indonesia). Strategic narrative asset (proof-of-personhood) > near-term revenue. Enter only after US rollout proves stickiness AND remaining bans resolve.
Douyin, ~750–900M MAU, $480B+ commerce GMV, $95B revenue. Requires Chinese WFOE / JV / Mandarin product / ICP. Defer to 2027+.
WeChat, ceiling-of-ecosystem (mini-program DAU 764M > Telegram total MAU; 4.3M apps; 68 min/day). Same Chinese-co-founder / JV requirement. Defer 2027+.

Tier 4, Watch-list / hedge

LinkedIn, Games hit 3.5M DAU May 2026 (+75% in 4 months), 80–84% next-day retention, but first-party only, no SDK, no rev-share, no announced opening. Watch-list. ~1 person-week / quarter to monitor.
VK, VK ~94M MAU ~75M MAU (state-mandated pre-install Sept 2025), same SDK as VK. Russia hedge if TG block becomes permanent. Sanctions: rouble-out problem for Western entity. Build VK Bridge architectural readiness for 4-8 week ship; don't enter directly in 2026.
YouTube Playables, third-party SDK + cert, but curated, no Stars-equivalent. Gemini-3 Playables Builder (Dec 2025 beta). 24-36 month optionality bet.
Zalo (Vietnam), 79.6M MAU, ~3.5–6K active Mini Apps, 13M monthly users. MiniAI direct competitor. Vietnamese-language wall + partner-or-acquire strategy. $20–80M ARR ceiling if pursued.

Tier 5, Drop / not viable

Snapchat, Snap Minis killed August 2022. No mini-app surface exists.
Roblox, DAU declining 2 quarters in a row: 152M → 144M → 132M (Q1 2026). Russia ban Dec 2025 + age-check Jan 2026 hit acquisition. UGC games platform, not messenger.
Kakao, 49.1M MAU saturated (97% market share). No general-purpose mini-app surface, only Channels (chatbots) + webview-via-Login. $50M ARR ceiling.

What we look for in a candidate

Not vibes, five concrete criteria:

Active users at scale. Tens of millions of monthly active users minimum; Tier-1 candidates are >100M.
A native mini-app surface or embedded-app SDK. No surface, no entry. Snapchat disqualified for exactly this reason.
Code port-fit ≥ 50%. HTML5 / JS-runtime ecosystems cost weeks. Native-app or proprietary-runtime ecosystems (Roblox Lua, AR Lens) cost much more.
Native pay rail or payout primitive. TikTok Coins, Discord Nitro, LINE Pay, Stars, TON. Without this, monetisation is grafted.
Foreign-team barrier ≤ medium. WFOE / mandatory local partners / mandatory Mandarin product / sanctions-related rouble-out are red flags.

Each tier is a verdict on these five.

The indexation playbook (cross-ecosystem entry pattern)

The same pattern applies in every new ecosystem we enter. Articulated by Mark on the 2026-05-16 CTO sync:

Index the ecosystem. Run our crawler on the new mini-app surface. Build the equivalent of the @appss catalog for that ecosystem, what's live, what's growing, what's monetising.
Generate the Market Research talmud. Same AI agent (with new-ecosystem connectors) writes the «what's working / what's missing» research, scoped to that ecosystem.
Stand up Builder templates. Pro Builder + Builder API gain a new template for the ecosystem (e.g. LINE-MINI-App scaffold-bot scaffold, etc.). Schema-aware.
Open Remix to that ecosystem's catalog. Once indexed apps exist, creators can remix them through the same UX they use for Telegram.
Wire the local pay rail. Stars-equivalent on the host platform.

This isn't theoretical, Mark explicitly named (Russian messenger) as a near-term target on 2026-05-16 (see decisions/2026-05-16-max-messenger-entry-strategy.md), with an explicit play: open a RU legal entity, register bots under it, hire a lobbyist, ship the Builder -bot template. The cross-ecosystem doctrine is moving from doctrine to schedule.

, promoted from Tier-4 watch-list to Tier-2 candidate (2026-05-16 update)

Originally classified as «Russia hedge if TG block becomes permanent», defensive position. Mark's 2026-05-16 framing flips it offensive:

«Мы откроем юр.лицо в РФ, под которое будут регистрироваться все эти боты и апы. Мы будем проводить модерацию. Нанять лоббиста, который будет следить чтобы наших ботов не банили. И мы оторвались от Telegram-only, при этом в Telegram у нас всё круто работает.»

What changes vs. the «hedge» framing: - Active build target, not insurance. Builder team is asked to deliver a -bot template alongside Telegram-bot. - Concrete entity-+-lobbyist roadmap. Not «when Russia blocks TG, we'll figure something out», there's a sequenced plan. - Bot-only at v1. doesn't have a Mini App surface yet (only bots + strict verification). So initial entry is bot-shaped. If/when ships Mini Apps, we re-cast. - Investor narrative shift. «What about Russia-block risk?», answer evolves from «views recover via VPN» (defensive) to «we're on too via a RU entity» (offensive).

Open questions in the decision file: legal-entity owner, lobbyist sourcing, RU revenue-rail integration, regulatory reporting requirements.