Understanding the Lara Translate Value Proposition

Rome – April 10, 2025

For the last twenty years or so, we have all been exposed to many different machine translation solutions, and we know that today, these generic public MT portals are used by hundreds of millions of people daily. However, the large majority of these MT portals provide a very static interaction with MT technology, where a user has little or no control over how the technology responds to unique translation requests. Translated has been a pioneer since the SMT days, in providing MT solutions that could be easily steered by enterprise users interested in delivering multilingual content at scale to an increasingly global customer base. ModernMT, an Adaptive MT pioneer, is a recognized MT quality leader amongst Neural MT (NMT) options available today and is considered among the most adaptive and flexible MT technology solutions, even in 2025.

Lara is an evolution and builds on this proven market-tested long-term MT experience. What distinguishes Lara is its foundation in Translated's long-established philosophy and approach: AI-driven, yet human-optimized technology where specialized data and tightly integrated human feedback work symbiotically to produce continuous learning. Lara produces remarkably more fluent and natural-sounding translations, and importantly, it offers greater control to the translator. We've always seen technology as a tool to assist experts in complex cognitive tasks. Language translation, in particular, has long been one of AI's most formidable challenges. However, the progress we've witnessed in the last five years is truly remarkable—machines can now translate at a level of competency that was once unimaginable. Lara represents a significant advancement in machine translation output quality, bringing us closer than ever to the quality and nuance of professional human translation.

We've taken the latest advances in LLM technology and meticulously refined and tuned them to perform optimally for translation tasks. This isn't just about automation and better COMET scores; it's also about enhancing the synergy between AI and human expertise. Lara produces remarkably more fluent and natural-sounding translations, and importantly, it offers greater control to the translator, or any user interested in getting a more flexible and accurate outcome in an MT interaction. 

Lara Translate offers significantly increased control, flexibility, and accuracy for casual, business, and professional translators by providing a range of easily applied adaptation options. These enhanced capabilities result in substantial efficiency gains and a more satisfying user experience in terms of control and quality. By enabling the translation industry to scale the production of high-quality, multilingual content, Lara ultimately facilitates broader global communication and superior global customer experiences.

There are at least three specific benefits that come to mind:
Context-rich, Accurate Business and Professional Translation.

Lara excels at providing contextually relevant translations and can often do this with little need for translation memory, which previous generations of MT and NMT required. We have noted that the improvement in MT output quality averages around 20% for Tier 1 languages compared to well-tuned NMT engines.  Tier 1 languages include the following: English, Deutsch (German), Español (Spanish), Français (French), Italiano (Italian), 日本語 (Japanese), 한국어 (Korean), Português (BR) (Brazilian Portuguese), Русский (Russian), 简体中文 (Simplified Chinese) and 繁體中文 (Traditional Chinese).

In specialized applications, where Lara is utilized with meticulously curated reference data, we have observed instances where its raw output quality surpasses that of average human translators and approaches the level of expert linguists (shown above).
While we project more moderate quality enhancements for less frequently utilized and low data-resourced languages (Tier 2, Tier 3, and Tier 4), we are actively leveraging Large Language Model capabilities to expedite progress for these languages. These advancements are designed to enable localization teams to achieve publishable quality with greater speed and efficiency.
The categorization of languages into tiers is not fixed. The table below considers factors like the share of global online economic potential, the number of native speakers and internet users, and strategic market expansion considerations. While Tier 1 preferences may be similar across major global enterprises, the specific language tier rankings can vary slightly for each company.

The error rates for Tier 2 and Tier 3 languages are expected to be higher initially. Still, they will improve over time as a continuous improvement process is embedded into the design of the Lara technology infrastructure . Historical experience and learning suggest that the rate of improvement will be faster as Tier 1 data and learning could also be leveraged to accelerate the ongoing progress.

"It’s wonderful and nerve-racking seeing people turn on TaaF (translation as a feature) all over their enterprises. It’s been a dream for decades. But, the issues with untrained, generic machine translation and LLMs multiply when you do them in a dozen languages that most employees don’t understand — in functions that know almost nothing about how language works. The next generation of localization is AI-centric and ubiquitous. I really hope that AI and IT leaders step up and work with localization teams to bring all the planes in, to a smooth landing in their orgs."

Kathleen Pierce, Principal Analyst at Forrester, Content Strategy and Operations

Deployment of Enterprise-Wide, Secure, Adaptive Instant Translation

Lara offers a user-friendly, universal translation tool accessible to executives and managers across all departments. This empowers them to quickly and efficiently translate essential business materials—emails, presentations, contracts, plans, real-time chats, research, and PDFs—directly within their daily workflow. By providing instant multilingual access to critical content, Lara streamlines international business communications and accelerates strategic initiatives.

The precision and quality of this universal translation function can be significantly augmented through the provision of curated data and expert guidance by linguistically and technologically proficient localization teams. This expert input enables Lara to achieve elevated quality standards for enterprise-specific content.

Broad deployment of this adaptive AI technology has the potential to foster increased collaboration between localization teams and their counterparts in operations, product management, engineering, marketing, and sales departments, particularly when focused on global business initiatives.

This expanded engagement allows localization teams to play a more pivotal role in driving positive international business outcomes and enhancing the global customer experience. Notably, the implementation of a high-quality, reliable, rapid, and universally accessible enterprise translation solution for key global stakeholders may strategically position localization managers to assume a more influential role in the formulation of international business strategies.

"And I should really start a collection jar with a nickel for every time I hear "we made MT or LLM localization widely available in our org, and we can't believe how many different functions are using it." Yeah, by anyone who works with a colleague, customer, partner, or investor who uses a different language."

Kathleen Pierce, Principal Analyst at Forrester, Content Strategy and Operations

Improving Global CX with More Relevant Content

A persistent challenge in global business is the limited access international customers have to relevant content, primarily due to translation costs and the unceasing creation of new content in a single enterprise language. Research confirms that the existence of straightforward and robust multilingual content production capacity can positively impact the decision to increase the volume of information made available to global customers, thereby more fully addressing global customer informational needs throughout the buyer and customer journey.

There is widespread recognition across industries, that customer-created content has grown increasingly more important in the overall marketing mix. "User-generated reviews on your landing page significantly increase trust, with 83% of people considering a business more credible if it has such content. And when it comes to trust, content from peers is preferred to brand content every time", according to a report from the Forbes Council. This highlights the clear preference for peer opinions over corporate messaging. While this preference is clear for basic consumer items, the influence is even more pronounced for high-ticket items, with a 380% conversion rate boost when reviews are present, according to the Medill Spiegel Research Center (SRC).

McKinsey and Forrester have reported that this is also true in B2B scenarios. Both firms emphasize that B2B buyers conduct extensive online research before engaging with sales representatives. This research includes consulting peer reviews, online forums, and other forms of UGC. Forrester, in particular, has tracked the "digital body language" of B2B buyers, highlighting their reliance on online information.

The benefits of providing superior CX are increasingly clear:

• Consumers will pay a 16% price premium for a great customer experience AND are more likely to be loyal to the brand. – PwC
• Among US consumers, 63% say they’d share more information with a company that offers a great experience. - PwC
• “Customer Experience leaders grow revenue faster than CX laggards, drive higher brand preference, and can charge more for their products.” – Forrester’s Rick Parish.
• 79% of consumers who had a positive support experience would recommend the company to others. - Zoom

The consequences of providing what is regarded as inferior CX are also quite clear:

• One in three consumers says they will walk away from a brand they love after just one bad experience. This figure is even higher in Latin America, at 49%. – PwC
• 92% would completely abandon a company after 2 or 3 negative experiences. – PwC
• Customers will switch to a competitor after a single bad experience; however, this varies by industry – Qualtrics XM Institute

Providing relevant content to customers before, and after, the purchase where product or service issues may arise is critical to success. While global customers favor self-service options for product and service information, many organizations face a significant information gap between their primary language content and that offered in their international markets. Lara provides a practical approach to bridging this gap by enabling the rapid and cost-effective generation of translated content, which market experience demonstrates is widely accepted by customers. We now see that the fastest-growing markets are in emerging economies in the Global South and many new languages will need more content.

As the translation focus moves to service high-volume, fast-flowing social media, UGC, real-time chat, email, and collaboration around knowledge content, it makes much more sense for the systems containing this content to connect directly into adaptive AI with integrated corrective feedback (learn-fast, improve quickly) systems, built around data interchange layers designed for 100X+ the scale of most traditional TMS systems which are less important when AI translation is trusted.

The evolution of enterprise CX infrastructure demands a paradigm shift towards more integrated, agile, and direct interaction with both automated and human translation services. This enhanced synergy is essential to facilitate the delivery of translations that are precisely tailored to their intended purpose and context. The incorporation of scalable enterprise translation technology, exemplified by platforms like Lara, heralds a future where the constraints of translation are dramatically reduced. This vision transcends the current limitations of translating only what is deemed essential, and instead, envisions a reality where comprehensive translation is achievable - a future where we can, quite literally, "translate everything."