As a Business Development Manager at localization, voiceover, and QA service provider Alpha Games, I’ve spent years working with major publishers on titles like World of Warcraft, The Witcher III, Diablo III, and RuneScape. We’ve seen firsthand how high-quality localization can elevate a game’s success in its target territories, turning regional launches into global phenomena. That’s why Valve’s August 2025 update to Steam’s review system, introducing language-specific review scores, feels like a pivotal moment for both the games and localization industries.
For games with over 2000 public reviews (and at least 200 in a single language), Steam now defaults to showing a review score calculated solely from reviews written in the user’s primary language. This is more than just a UI tweak – it’s a direct spotlight on localization efforts, forcing developers and publishers to prioritize quality or else risk a backlash in key markets.
Previously, Steam’s overall review score aggregated everything globally, meaning issues like poor translations or cultural missteps in one region could drag down a game’s rating everywhere. Now, with language silos enabled by default, problems in one language might not drag down the overall review scores like they used to. That said, it also means positive reviews in one language can hide problems elsewhere.
- A low score in one language (e.g.: “Mixed” or “Mostly Negative”) won’t tank the global rating, but it will dominate the experience for players in that region. Potential buyers scrolling in Simplified Chinese, for instance, will see the localized score. This could potentially deter sales in a market that now represents Steam’s largest user base. Developers can no longer ‘hide’ sub-par translations behind strong English-language praise.
- Publishers must now treat localization as a core launch priority, not an afterthought. Speedy machine translations done on the cheap or over-reliance on under-resourced teams risk visible failure. Partnering with a specialist provider like Alpha Games can ensure cultural nuance, idiomatic accuracy, and QA testing that resonates with local players.
- On the flip side, strong localization can sometimes shield the overall score from unrelated review bombs (e.g., political or off-topic outrage in one region). But a persistent low score in a major language such as Simplified Chinese can still hurt revenue; China’s players are vocal, demanding, and influential.
In short, this update raises the bar. Poor localization in one language can cap a game’s potential in that market, even if it’s a hit elsewhere. It encourages publishers to budget properly for expert localization from day one, ultimately leading to better games and happier players worldwide.
Take, as an example, Hollow Knight: Silksong, the long-awaited sequel that launched to massive hype in September 2025. Globally, it earned “Very Positive” or “Overwhelmingly Positive” ratings in most languages, with peaks over 90% in English. But in Simplified Chinese, it plummeted to “Mixed” or “Mostly Negative,” with approval rates as low as 42-50% from tens of thousands of reviews.
Why the review bomb? The Simplified Chinese translation was widely criticized as “rubbish”; overly archaic, pretentious, and tonally off instead of matching the original’s whimsical, immersive style. Silksong credited just two translators for the Chinese translation (versus six for the original Hollow Knight), and players felt it ruined immersion, story, and dialogue. Many reviews praised the gameplay but criticized the game heavily for the localization, with complaints about unreadable fonts and mismatched tone.
The impact on Team Cherry and publishers?
- Chinese players, now Steam’s biggest demographic, saw the low localized score immediately, potentially stunting word-of-mouth and revenue in a huge market.
- Team Cherry’s Matthew Griffin publicly acknowledged the issues and promised fixes “over the coming weeks,” turning what could have been quiet patches into a high-profile attempt at damage control.
- Without Steam’s new system, this could have dragged the global score down (as it briefly did before separation). The studio managed to isolate the problem but, ultimately, the experience shows how one weak link (localization) can overshadow years of development in a key territory.
The Silksong experience underscores the risk of cutting corners on localization for a high-profile release; such a move invites a backlash that directly affects discoverability and sales in the target language’s ecosystem.
A “Mixed” rating in a major language does more than reflect poorly on the game, going so far as to erode trust in future titles from the same studio or publisher. Players associate the brand with subpar experiences in their region, reducing wishlist adds, launch hype, and long-term engagement. In markets like China or Japan (where reviewers are often harsher), this can mean delayed or skipped localizations for sequels, perpetuating a cycle of neglect.
Ultimately, Steam’s update is a net positive, as it empowers players with relevant feedback and pushes the industry towards excellence. For developers and publishers eyeing global success, partnering with proven localization experts isn’t optional anymore; it’s essential insurance against fragmented ratings and lost markets.
If you’re a publisher planning a multi-language launch, let’s talk. We’ve helped icons like Blizzard and CD Projekt RED nail their localizations, ensuring every player feels the magic, no matter their language. Ultimately, Steam has sounded the wake-up call that our industry needed.