The State of The Language Industry

LangOps is gaining traction lately, sparking debates across many platforms.  For some, it represents the advent of the Post-Localization Era; others refuse to go beyond translation, quickly dismissing it as a hype, similar to blockchain. I believe it would help the discussion to add some context and have a closer look at the state of the translation industry and the forces driving its transformation.

Elephant in the Room

The big elephant in the room is, of course, AI. Neural Machine Translation, now often also labeled (with some merit) as AI, should have already turned the industry upside down eight years ago with its human parity claims. With Generative AI on the rise, the paradigm shift is now undeniable. When ChatGPT launched, translators were listed among the jobs most likely to be impacted. While some pragmatists embraced the new technology early on, integrating it into their production workflows, others rather joined a lively debate with positions ranging from “it’s all hype” to “AGI will soon send us all home.”

The notion of AI as mere hype is fading, replaced by an acknowledgment of its rapid evolution. Whether we are on the brink of AGI remains debated. Sam Altmann recently suggested that OpenAI already has something close in the drawer. Others have started a rather academic discussion about what AGI actually is. I believe you can confidently draw a progress line: whatever doesn’t work so well today will work tomorrow. Revealingly, many experts measure the time to AGI in years, not decades.

Will the industry step up to lead the LangOps transformation? We are committed to enabling this transformation with cutting-edge technology.

The Quality Debate, yet again?

Currently, the main argument in favor of linguists over AI is quality. It’s a somewhat self-serving mantra in the industry that only humans can achieve the required quality. But the question, “When will AI match human translators?” is flawed, it assumes that human output is perfect, the golden standard. The reality is far from it, often due to a lack of proper resourcing or briefing.

The more relevant question is: “Can AI match the quality typically delivered in the market today?” Increasingly the answer is yes. Furthermore, AI errors can be consistently reduced by smart uses of NLP and Auto-correction using Retrieval Augmented Generation (RAG).

At the end of the day, quality is defined by buyers – not providers. Clients prioritize outcomes that meet business needs over linguistic perfection. The industry’s fixation on embarrassing translation fails is overlooking the truth: most buyers require “good enough” solutions that align with acceptable risk and cost.

Race to the Bottom

AI-driven efficiency has led to a dramatic drop in translation rates. The industry is experiencing a race to the bottom, putting pressure on revenue and profits. Buyers, including public institutions, use AI themselves to drive word rates down. To still win contracts translation services are offered from one year to the other at half or even a third of the price. LSPs and in-house translation departments alike are squeezed between cost-conscious buyers and the income linguists need to make a living.

Money Talks

These developments are leaving ugly marks on balance sheets and stock prices (5-year performance: RWS -74%, Straker -52%, Appen -90%). The industry is fragmented, with only a few players publicly traded, so it’s hard to get good data. Still, most will agree that the days of comfortable revenue growth driven by globalization are over.

The once-thriving M&A market has notably cooled down. In the past, translators-turned-entrepreneurs who built boutique language service providers (LSPs) with one or two key clients could sell their businesses for 1-3x revenue, often enough to secure a comfortable retirement. These opportunities for a lucrative exit have largely disappeared. Likewise, larger LSPs are now facing significant declines in valuations.

In a recent adjacent transaction, the most significant challenge was the technical due diligence, which has shifted from being a routine formality to a potential deal-breaker. Investors, focused on future returns, are getting nervous. VC funds typically operate on a seven-year horizon, while private equity funds aim for five years. With AGI potentially on the horizon within a few years, it is no surprise that investors are becoming increasingly hesitant to commit to this industry. The message is clear: the old localization business model is no longer a safe bet.

Hiding in the Translation Box

Faced with these challenges, many companies retreat to familiar territory – sticking to the “translation box”. They find comfort in influencers who push the “nothing to see here” narrative, predicting that at LocWorld100 the same topics will be on the agenda: translation tools, TMS, project management, pricing models, buyer needs, etc. This defensive posture may well provide short-term comfort, but who knows how long this box will continue to feed LSPs and translation tool/TMS vendors? I believe going forward, just “doing words” won’t create much value for investors, shareholders, or employees anymore.

Drive the Transformation

At a recent invitation-only conference, an executive of a leading LSP showed a slide featuring the masked Hannibal Lecter. The slightly spooked audience was drawn by the provocative metaphor: “Cannibalization”. The advice was blunt – disrupt yourself before the market does it to you. The entire legacy ecosystem – freelancers, boutique LSPs, SLVs, MLVs, tech providers, and even “buyers” in corporate translation departments – is under threat.

In fast-changing times, self-cannibalization can be a tempting strategy, yet a rather uncomfortable one. A more strategic approach is transformation. AI’s potential does not materialize in isolation. It thrives on high-quality, domain- and company-specific data – for instance, multilingual knowledge graphs and content repositories for term- or content-level RAG. Who sits on these multilingual assets (or on legacy formats like TMs and term databases that can be elevated)? Who creates and curates them? Who knows how to tap into a global network of experts delivering data in dozens or even hundreds of languages? We do!

So, the question is not whether AI will replace linguists. It is rather whether the industry will step up to lead this transformation.

Enter LangOps

A leading LSP’s tagline says “…ensuring you are understood anywhere”. If you switch that passive voice to active, “…ensuring you understand every customer, anywhere,” you step outside the translation box and take the first step toward a broad LangOps strategy.

This principle of understanding all customers is in fact the cornerstone of LangOps. By embracing AI-driven solutions and leveraging multilingual assets, companies can move beyond translation to deliver meaningful engagement across languages and cultures.

The language industry stands at a crossroads. The choice is fight or flight: cling to legacy models or embrace innovation and transformation. LangOps is not just a strategy, it is a call to action for the industry to redefine itself and create lasting value in a rapidly changing world. At ESTeam and Coreon, we are committed to enabling this transformation with cutting-edge technology. I believe the future belongs to those willing to step outside the box and lead the charge toward LangOps.