Meta在游戏中的角色：Muse、VR与UA

A month after its personal productivity agent Muse hit number one on the App Store, Meta used its Connect 2026 event to state its priorities: AI first, glasses second, VR a distant third. But that list is missing arguably the most consequential story for game developers: Meta’s slow return as a preeminent gaming UA channel. Growing engagement with video products like Reels is a double-edged sword for gaming: Facebook and Instagram’s growth lets publishers reach more potential players, but those same products compete relentlessly with games for users’ attention.

Right now, Meta’s gaming-oriented metaverse investments are in a down cycle: the Quest hardware and Horizon UGC platform are clearly deprioritized, while its role in gaming UA is gaining momentum. And while glasses and Muse are clearly areas of investment today, the company has yet to connect them to gaming use cases.

Meta Is (Back as) a Games UA Engine

Since Apple introduced App Tracking Transparency (ATT), Meta has not been the dominant growth channel for mobile gaming. That is changing rapidly: in 2025, Meta grew its share of gaming ad impressions on social media from 54% to 69%, while TikTok’s share nearly halved (21.0→12.7%) and YouTube fell at a similar rate (20.0→12.0%). The company is outcompeting its social media rivals not only by growing impressions but also by improving price per ad (across all verticals, not just gaming), which rose 12% in 2026. Meta’s ad business is growing in both volume and efficiency, an impressive feat that suggests its underlying Recommendation System is becoming more efficient at driving engagement and allocating ads.

ATT hit Meta particularly hard; it relied on the deprecated IDFA for targeting gaming ads far more than competitors like AppLovin, and ultimately ceded much of the vertical to them, winding down its once-grand dedicated gaming ads team and prioritizing e-commerce until quite recently.

*Meta’s gaming ad impressions in 2026 YTD already exceed those of any full year since 2020 | Source: Pathmatics (Sensor Tower)*

Since 2024, Meta has been building the capabilities needed to recapture a large share of gaming UA. These include incrementality optimization (Q4 2024), LTVision (2024), AMM (reinstated in July 2025 after being shuttered in October 2021), and pLTV postback / “Value Optimization for Non-Purchase Events” (beta in 2025); Andromeda retrieval/ranking and Advantage+ as the default for App Promotion; and, importantly, rebuilding the dedicated gaming ads organization, now based in London, to work directly with publishers globally.

What does Meta’s return mean for gaming publishers? On the bright side, a new, high-performing channel is welcome. Unlike AppLovin, Unity, and other gaming-focused ad networks, Meta delivers impressions outside of games. Meanwhile, the biggest drawback of gaming networks is that they may shuffle an existing pool of game players. Many game studios also hide ads from their highest-LTV players, creating the risk of adverse selection.

Meta avoids this problem structurally. Everyone sees ads on social media, not just low-LTV players. Meta’s post-ATT ad products also help expose gaming ads to users most likely to spend in mobile games. Of course, Meta is not the only player in social media, but its share of social media engagement and gaming ad impressions has increased significantly at the expense of rivals like TikTok and YouTube, making Meta an increasingly important channel for reaching the most valuable players.

The VR Platform Dream Keeps Fizzling

*Meta’s upcoming VR Glasses | Source: IGN*

Meta’s ads business is newly re-optimized for serving game publishers, but the same cannot be said for its Reality Labs business, which houses both its VR gaming and Horizon UGC efforts. Meta not only changed its name to signal its long-term bet on metaverse-like products but also sank $88B into the division between 2020 and 2026. Since then, it has hedged this bet considerably. Of the nine VR / gaming studios it acquired (such as Sanzaru, Twisted Pixel, and Armature), most have been closed or reduced to a skeleton crew. Sequels to VR “hits” like Arkham Shadow were canceled, and 2026 Connect had “no big-budget exclusive of the kind Meta used to put in the Quest box.”

Meta has been refocusing its hardware away from the gaming-oriented Quest line of headsets and towards glasses. Meta’s augmented-reality Ray-Ban glasses have sold well and received heavy promotion since their September 2023 launch. The most recent product, announced at this year’s Connect, is the aptly named “VR Glasses.”

*Meta’s VR Glasses prioritize media and productivity | Source: Meta*

Although still a VR device, the VR Glasses are positioned as “a different product for a different market,” according to VP of VR Ryan Cairns. The VR Glasses are clearly a competitor to the Apple Vision Pro and are meant for casual VR entertainment that requires little local computing power. This helps explain why the announcement barely mentioned gaming, with only one sentence: “more than 75 games at launch that you can play solely with hand gestures, and with Xbox Cloud Gaming, you can stream hundreds more.”

The developer page for the product explicitly recommends casual and social mechanics over “virtual locomotion or gun-based gameplay.” This reflects the expectation that most users will rely on hand-tracking rather than dedicated controllers. In Meta’s current catalog, there are 123 titles that support hand-tracking input (including 19 ‘Coming Soon’ titles featuring well-known IP such as Tetris, Two Point Hospital, Ace Attorney, and Project Hail Mary). But ten titles account for 82% of the 265K star ratings for these games, and every title with more than 20K ratings launched between 2019 and 2020. On the bright side, the 2025 cohort is the biggest ever, with 29 games, but the median game has 73 ratings; the most recent hit has been Forever Pets (F2P, 7.7K ratings, 4.9★).

Meta announced no new Quest headset, though a forthcoming Quest 4 is rumored to be explicitly gaming-focused. This would segment the market with the VR Glasses serving the video + apps consumer, while the Quest doubles down on gaming.

The Ray-Ban glasses are Meta’s most accessible hardware products and likely to gain the most consumer traction over time. They still lack full native support for third-party apps and games, but Meta is clearly moving in that direction. At Connect, Meta announced Explore, “a new on-glasses app to browse and use… experiences built for your glasses… to learn, explore, and play.” This looks like the embryo of a store. It follows the opening of Ray-Ban Display to developers in May 2026 via the Wearables Device Access Toolkit, a native SDK that adds a “glasses layer” to existing iOS- and Android-supported apps and supports standalone web apps. For now, there is still no public publishing, reviewing, or monetization process, unlike the process for Meta’s virtual reality headset.

Meta also fully unbundled its Horizon UGC ecosystem from the Quest hardware in February as part of a pivot toward mobile (which it has since somewhat walked back), and Horizon’s latest sign of life is Horizon Create: a prompt-to-UGC long shot against Roblox. Create turns Horizon into an Instagram-style feed of instantly playable AI-generated games. Anyone can prompt games into existence, remix them with friends in real time, and publish them to Facebook, Instagram, or Horizon in two taps. A public rollout is slated for early 2027.

As Horizon moves away from hardware, Meta’s latest software product, Muse, is likely to become even more bundled with these devices: as Ben Thompson at Stratechery put it, this hardware may be little more than “Muse Delivery Vehicles.” Meta’s new agent will likely be deeply integrated into this hardware, but it’s worth examining in its own right.

What Could Muse Mean for Games?

*Muse’s anthropomorphized character | Source: The New York Times*

Meta is poised to launch a physical Muse “charm” device, while also opening up Muse to other hardware makers. On the surface, it clearly parallels Tamagotchi, and the anthropomorphized agent (like OpenAI’s Dots) was designed with that concept in mind.

At its core, Muse is a goal-based personal agent: it connects to Meta and third-party apps, fills out online forms, uses the web, makes purchases, and keeps working after you close it. It offers a free tier (unlike Dots), plus $20 and $100 tiers. Heavily promoted on Facebook and Instagram, Muse became the No. 1 free app on the U.S. App Store, surpassing ChatGPT. By late September, it had reached 3.4M downloads with ~55% day-over-day growth. Early retention metrics put Muse’s D14 retention at 22%, comparable to ChatGPT — a strong start, though the app has drawn notable privacy and safety criticism.

While Muse has been marketed entirely around personal productivity, some users have used it for entertainment (of the adult variety). Gaming use cases have yet to emerge, and it is unclear whether Meta intends to support them in the near future, but there are several implications for what Muse or similar products could do:

Discovery and UA disruption: Muse, Instinct, and other personal agents already deliver e-commerce ads; game discovery fits here. An agent that discovers, installs, tries, and even pays for a game on a user’s behalf would sit on top of app stores and even above Meta’s feed-based UA engine. Will gaming UA start targeting players’ agents?
Agentic commerce within games: Muse already transacts on users’ behalf. Beyond discovering and buying games, Muse could purchase things within games a user already plays, potentially reshaping the relationship between players and spenders, as well as in-game monetization.
Companion / play-on-your-behalf: Muse or other agents could grind endlessly inside a game for their users, raising design, monetization, and anti-cheat questions.

Agents such as Muse are poised to change how people live, work, and play digitally. Their potential impact on gaming habits deserves particular attention.

A Complicated History and an Unpredictable Future

Meta tried at least twice to be a gaming platform. Its first attempt succeeded during the Mafia Wars and FarmVille era, which famously ended when Zuckerberg felt Zynga’s growth tactics threatened Facebook’s user engagement. After that, Meta bought Oculus and several game studios, such as Ready at Dawn, to build a virtual reality metaverse. More recently, it tried to pivot into a Roblox-like mobile UGC platform, without notable success.

AR and VR remain limited opportunities for game developers. The VR Glasses compete against Apple’s Vision Pro for an extremely niche audience, with gaming as an afterthought. The Quest 4, which has not been officially announced, is unlikely to do better than its predecessors. Perhaps there is a light at the end of the tunnel in the eventual opening of the Ray-Ban devices as a platform for third-party developers, and in VR Glasses finding an audience.

Meta may be realizing that its most valuable role is the one it has quietly played all along: UA infrastructure that routes players to games on other platforms, using its increasingly sticky social apps and recommendation algorithms to become gaming’s largest and most efficient UA channel. Meta’s growth engine is Reels time. Reels time delivers more gaming ads to valuable potential players, but it is also competing with games for players’ time. Meta doesn’t need to kill games’ distribution the way Facebook did in 2012; it just needs to keep winning the attention war that games are losing, and then sell game publishers a way back in.

As Meta’s VR and platform bets lose money while its UA roars, Muse is the wild card. It has no clear direct connection to gaming yet, but will likely have the most transformative impact on the industry by changing how players discover and play games.

Sponsor: Better Matches, Steadier Spikes: Whatwapp’s Matchmaking Case

Whatwapp used Heroic Labs to move from synchronous to asynchronous matchmaking. This change delivered better-balanced matches and steadier handling of traffic spikes, reshaping how their team approaches matchmaking design.

Learn More

In Other News

Funding and M&A:

Business & Product:

Miscellaneous:

Sponsor: The Payments Infrastructure That Nets You More Revenue

Every failed transaction, false decline, and unresolved dispute is revenue a studio never sees. Appcharge closes those gaps – both on Web Stores and Payment Links that move IAPs off app store rails.

Playstudios’ MGM Slots Live saw DTC Payment Links alone reach 39% of overall game revenue within weeks, and top-grossing titles across genres are seeing similar results.

That performance scales to $1BN annually, with multi-PSP architecture, local acquiring across 10+ regions, and a smart routing engine that delivers a 96% Order Success Rate and a 0.1% dispute rate, versus an industry standard of 0.83%. Add full Merchant of Record coverage, and partners keep more net revenue from every DTC transaction, at scale.

Payments optimized for mobile games are where the real margin lives.

See the payments data

Content Worth Consuming

Aream & Co. Video Game Market Update Q3 2026 (aream.co): “The report highlights key trends across three major sectors of the gaming industry: Gaming Market Environment: Steam continues to grow steadily with a 14% YoY increase in LTM revenue driven by new IP like Wardogs, whereas console revenue experienced its steepest quarterly drop in two years (down 12% to $14.1bn) and mobile IAP spend slipped 7% to $19.6bn. Dealmaking Activity: M&A activity remained strong with 54 deals worth $57.8bn, heavily driven by the $55bn EA take-private deal in August, alongside notable content-focused acquisitions and private equity investments concentrated in Asia. Capital Markets: Private investment more than tripled YoY to $2.8bn across 137 deals—led by late-stage rounds and major fundraising by AI gaming companies—while public offerings rose 122% YoY to $1.3bn with more listings on the horizon.”

Star Fox Developer Velan: How to Survive as a North American Game Studio (The Game Business): “This week, we speak to Velan Studios leaders Guha and Karthik Bala, who run the studio behind the new Star Fox remake on Nintendo Switch 2. The Balas, who had previously founded Vicarious Visions, discuss how they won over Nintendo, how they keep costs low as an independent US developer, working with big IP, taking risky bets and finding a ‘middle way’.”

Global Gaming Report Q3 2026: EA Goes Private, Warner Bros. Clears, Gaming Stocks Surge ~15% and AI & Tools Capture $1.3B of $1.6B Raised (drakestar.com): “The Drake Star Global Gaming Report Q3 2026 analyzes gaming dealmaking and public markets in the third quarter of 2026 and outlines what to expect for the rest of the year and beyond. 3’26 belonged to the megadeals. The $55B take-private of Electronic Arts by PIF, Silver Lake and Affinity Partners, the largest all-cash sponsor take-private in history, closed in August. Paramount’s $110B acquisition of Warner Bros. Discovery, home of Warner Bros. Games, cleared its final hurdle in September, and the combined company now trades as Skydance Corporation (NYSE: SKYD).”

GMN Source Files — Composer Austin Wintory (The AIAS Game Maker’s Notebook): “Grammy and Bafta winning composer Austin Wintory stars in the premiere episode of Source Files, a new series in the Game Maker’s Notebook podcast family where we ask developers what games were most influential in their own careers. Austin recounts which three games had a profound influence on his own work on titles such as Journey, The Banner Saga series, and the Grammy award winning Sword of the Sea.”

Why Players Judge Your Game by Its Worst Part (Building Better Games): “Building a game from scratch is brutally difficult, and game developers often debate how much to invest in genre fundamentals versus chasing new, shiny innovations during early pre-production. While some push for an ultra-lean approach to reach the experimental phase faster, skipping or rushing core genre foundations creates an unstable prototype, misleads early player validation, and forces massive, costly technical rewrites later in development. In this second instalment of our ongoing episode series with Leif, Leif and Ben dissect why mastering “minimum lovable” genre foundations early is essential for setting up innovation, shielding player credibility, and maintaining team momentum.”

The Happiest Man In Gaming Hates Generative AI (gamesindustry.biz): “It’s impossible to not be swept up in Tarn Adams’ enthusiasm. To speak to him is to feel like you’re talking to the happiest man in gaming. Along with his brother Zach, he’s been working on the wonderfully chaotic procedurally generated world of Dwarf Fortress since 2002, and he is far from tired of it. If anything, he’s more enthusiastic than ever. ‘There’s just so many cool things to work on when you’re writing a sort of expansive, scope-bloated simulation game,’ he tells GamesIndustry.biz. We can feel the waves of excitement coming off him as he tells us about the latest addition to Dwarf Fortress: procedurally generated magic systems.”

The post Meta’s Role in Gaming Today: Muse, VR, and UA appeared first on Naavik.