Embark Studios head Patrick Söderlund explains how Arc Raiders was made on "a quarter of the budget" of a AAA title

Arc Raiders has sold over 14 million copies since its October 2025 release, with Wedbush estimating development and marketing costs near $75M (Embark employs ~360 staff). CEO Patrick Söderlund says Embark delivered AAA-caliber output at roughly a quarter of a typical AAA budget by reconfiguring pipelines, using procedural tools/photogrammetry and selective AI as a production aid rather than a replacement for actors. The studio re-recorded some AI-generated voice lines with real actors post-launch and plans to keep The Finals and Arc Raiders active while targeting four total games within five years without expanding headcount to 600–800. Operational discipline and tech-driven efficiencies are positioned as the firm’s competitive strategy, though industry-wide layoff pressures and AI controversies remain risk factors.

Analysis

Embark’s ability to hit AAA output with materially lower spend signals a structural shift: marginal cost of producing high-quality content is falling as procedural pipelines, photogrammetry and lightweight AI are adopted. Expect growing differentiation between firms that own modern, composable toolchains (engines + cloud + content marketplaces) and those that remain dependent on bespoke legacy pipelines—this will compress moats built purely on scale and increase returns to middleware and cloud infra vendors.

Second-order effects: voice/TTS licensing becomes a recurring-revenue product (licensing actors’ voice profiles and usage rights), creating new monetizable intermediaries between studios and performers and increasing demand for rights-management platforms. Simultaneously, talent economics will bifurcate—small, high-skill teams commanding premium per-head output while large publishers face elongated, lumpy capex cycles to modernize, creating a multi-year reallocation of developer headcount and vendor spend.

Tail risks that can reverse the trend include rapid regulation/unionization around generative AI and voice licensing (raising unit production costs), a platform-level backlash from players over synthetic content quality or authenticity, or a failure mode where procedural assets produce homogenized experiences that depress engagement. Key near-term catalysts to watch: engine middleware revenue growth, GPU/cloud bookings from creative customers, and emerging licensing marketplaces for voice/photogrammetry over the next 3–12 months.

Key Decisions for Investors

• Long Unity Software (U), 12–18 months: buy shares or a modest call spread. Thesis: Unity collects software revenue as more small-to-mid studios shift to composable pipelines; expected upside 30–50% if adoption accelerates. Downside: 20% if ad/indie spending contracts; use a 15% stop-loss.
• Long NVIDIA (NVDA) 3–12 months: overweight NVDA or buy 6–12 month call spreads sized for beta. Thesis: GPU demand from AI-assisted content creation and cloud rendering is a direct lever; potential >40% upside if studios increase cloud GPU spend. Tail risk: crypto/AI macro correction could inflict 25–35% drawdown.
• Pair trade — Long Microsoft (MSFT) / Short Electronic Arts (EA), 6–12 months, equal notional: MSFT benefits from Azure + first-party modernization tools and distribution leverage; EA is more exposed to restructuring costs from legacy multi-studio pipelines. Target 1:1 outperformance of MSFT vs EA of 15–25%; stop-loss if spread moves unfavorably by 12%.
• Long Adobe (ADBE), 6–12 months: buy shares or diagonal call structure. Thesis: Creative cloud incumbents will monetize AI-enabled production workflows (photogrammetry, TTS, rapid prototyping), offering steady SaaS upside of ~20–30%. Risk: slower enterprise rollout or higher R&D pushes margins lower by ~5–8% in the near term.