"From translation, all science had its offspring." Giordano Bruno.

The ability to translate information and communication significantly impacted the development of humanity. It is not for nothing that the story of Babylon is an example of God's punishment, not a blessing. The information exchange provided the power to share cultural acquisitions, knowledge, and traditions to improve them by mixing. It's hard to imagine our modern life without translation systems, but not so long ago, automated translators didn't exist at all or were imprecise at the dawn of their formation. This article will give you more details about how the translation industry was developed and which trends and impacts it has nowadays.

From the Сold War to Deep Learning

As early as the 17th century, philosophers Gottfried Wilhelm Leibniz and Rene Descartes assumed that there could be a specific code to connect the words of different languages to support people in understanding each other. After centuries of discussions and patenting, in 1954, IBM and Georgetown University (USA) demonstrated machine translation publicly (Georgetown Experiment). The system was primitive - only 250 words, six rules, and 49 pre-selected sentences with a chemical topic.

In the 1960s, there were two systems of Russian-English translation- MARK and GAT. But these systems had low quality and were unprofitable. Machine translation development continued, but with fewer efforts due to funding decrease. The 60s of the last century marked the decline of the MT industry, and everything could remain as it was, but computer technologies came.

The first machine translation systems were created for specific pairs of languages. They were based on complex language modeling processes, the basis of which was analysis, transfer, synthesis, and interlingua.

first model of machine translation

1970–80 - are the "Renaissance" of machine translation. Primarily that was associated with computer technologies development and, as a result, creating the complex dictionary search systems focused on working with natural language data.

*During 1978-93, 20 million dollars in the United States, 70 million in Europe, and 200 million in Japan were spent on researching and developing the machine translation industry. One of the technologies was the translation memory which accumulates source and target segments to form the linguistic database. *

The era of SMT - statistical machine translation. Until 2016, the translation machine was based on different patterns. For example, the pairs of words were matched a few million times to count how many times the word "Das auto" translated as "car" vs. "automobile" vs. "motorcar," and so on. The text tried to split by words and phrases or used syntax-based translation.

2016 - Neural Machine Translation

In September 2016, Google announced the development of the GNMT translation system, which was a revolution. Nowadays, there is a list of such systems like Google, Deepl, and Amazon, and we at Lingohub actively use the combination of them to provide accurate pre-translation. Neural machine translation proceeds to improve day by day. For example, in comparison with 2016, the translation accuracy grew from 80% to 97% for the English-Spanish pair!

Translation Industry Today

The world has become interconnected, and the global language market is experiencing rapid growth. Based on the Fact.MR research, the market value in 2022 is around $60 billion, while in 2023, the predicted figure can reach $96 billion. In the context of geographical share, the estimated size of the language services market in 2020 was next:

  • USA - $18.7 billion
  • Europe - 26.9% billion
  • Asia - $8.1 billion
  • South America - $0.7 billion
  • Australia - $0.5 billion
  • Africa - $0.1 billion

Besides English, the list of other languages has become increasingly popular in the web environment - over 50% of all Google queries are in languages other than English. Compared to the number of visitors (Google visits 89.3 billion times every month), we can clearly understand how many people require content in their native language.

The logical question would be - so what language to translate? We already discussed these questions in our previous article. In this one, let's look at the top 10 languages by native speakers number to give you some direction for reflection.

top languages by number of native speakers

The translation industry develops by leaps and bounds. As in all growing industries, it has trends and new features that are current for some period. A decade ago, translating tools were primarily used for documentation and simple text translating. What about nowadays, when we have robust engines, the "powerful pocket computers" - smartphones, and various information? Let's look at the translation industry trends that are on top now.

Machine translation and post-editing (MPTE)

The time to market for the businesses determines the product's success. Do you have a new feature earlier than the competitors on the market? Without localization, it is mostly irrelevant, especially in sensitive industries like Fintech. Your users will not understand specific terms and will skip your proposition. Here we will come to the MPTE. With machine translation service, linguists save time because their job is simplified by pre-translation. The MT is developed to that level where there is no need to work with simple sentences. Such a lightweight process approaches us to the next trend.

Adaptation instead translation

So, with modern systems, translators can save time and focus on the quality of the result. Our previous article about famous product localization overviewed the importance of adaptation, not translation. It is always a pleasure for users to see or hear culture-close content. But text adaptation requires a native-speaker expert or linguist because the devil is in the details. For instance - the direct questions are mostly impolite in English - the imperative form of the verb in English sounds rough. The same thing to "Are you sure?". When we talk about the Ukrainian language, such a construction with exact translation will be unnatural.

"Are you sure you want to proceed with the installation" the direct translation will sound like "Are you really/exactly sure you want to proceed with the installation?" so, more natural in the example with Ukrainian will be the question like "Proceed the installation?."

Localization in the Metaverse

The Metaverse is an entirely new trend that does not refer to one specific industry. It combines VR, AR, eye-tracking, and other modern technologies. The ways of metaverse usage are still discussed, but some companies have already made a profit by selling their NFTs. In 2022, brands - Nike, Gucci, Adidas with PUNKS Comic, and Bored Ape Yacht Club unveiled special collections for the Metaverse. The result is - Adidas sold more than $22 million in NFTs for the 4 hours.

What is happening in Metaverse now? Will the industry grow?

  • Metaverse Holdings is planning to create copies of Dubai and Abu-Dhabi. This project is estimated at $54.5 million.
  • Renault launches the world's first industrial Metaverse.
  • Beijing has developed a roadmap for developing NFT technologies and Metaverses for 2022-2024.

As you can see, governments and big companies are investigating virtual environments. We predict that they will proceed with their advertising and promotion so that this trend will cover the world sooner or later. Back to localization - the user experience is in the first place. Adidas, Burberry, Gucci, Tommy Hilfiger, Nike, Samsung, Louis Vuitton, etc., are already building their stores in the Metaverse. They need at least to localize items' names, pricing, and currencies (or use crypto) to get a larger audience. The metaverse localization will grow together with technology.


The Covid pandemic has contributed to the development of e-learning. Learning Management Systems, educational applications, online lecture platforms, etc., received increased attention and popularity. The e-learning market can reach 325 billion dollars in 2025 (according to Forbes). The incredible popularity of online education and the high demand for learning from home allows companies to expand their solutions to different countries. The localization in this area requires deep and conscientious work. It is critical to translate terms correctly and adapt the content in a quality manner (expressions, adages, etc.)

Multilingual SEO

The success presented in the international market requires an in-depth view of multilingual SEO. There are two reasons. The first one - consumers always prefer resources in their mother tongue (40% of users will not buy on the website in a non-native language.) The second one, despite the high popularity of English, other languages have billions of native speakers. The content in English wouldn't help to reach global goals. In the long run, a multilingual SEO strategy that includes site content optimization, metadata updating (title, description,) link building, and more provides the requested result. The companies that started to be with users in the same internet environment have a higher chance of success there.

Types of the Translation software

So, we learned a lot about the translation industry's history, current situation, and trends. What does it offer in the context of private or business usage?

Machine Translation.

As we discussed in the first paragraph, the development stages of MT were marked by different approaches to the translation process. As a result, we have the following:

  • Rule-based machine translation. Such types of systems can be used in narrowly focused areas. They require constant manual dictionary updating and source text checking but provide predictable results.
  • Statistical machine translation. They don't need explicit rules because they independently understand the comparability of the texts. The system can automatically select the rules from the set of such examples if the engineer gives it a lot of source and target text.
  • Neural machine translation. The most common-usage systems, like Google Translate, are learned for a long time and improve the quality of translation texts daily.

Google Translate has more than 500 million daily users. It is widely used for work, education, or traveling.

CAT tools (Computer-Assisted Translation)

When discussing professional content translation, localization, or adaptation, we must highlight the instruments supporting linguists and related teams. One of the options is a group of tools that are called CAT. Who are these "cats," and how do they support the translation process?

Such systems have main missions - cut off the time spent on the translation and perform results. It's essentially a text editor with accessibility features for  teams. In such a tool, the manager can prepare tasks from source files, control the execution of tasks, and determine the amount of work. Mostly they have a list of instruments like:

  • Translation memory - the system can remember the translation of some segment and provide the correct one next time.
  • The glossary contains specific terms and rules for their usage.
  • Style guide.
  • Quality checkers.

TMS - translation management system

The next stage is translation management systems. As the name implies, the TMS is targeted at control through the entire localization/translation process. What is the difference between CAT and TMS if both support translation and management processes?

Let's take a deeper look at the Lingohub example. Lingohub is a translation management tool that combines in general three crucial localization things to provide a 360-degree overview and continuous processes:

The Management tools include text edits reports, a general dashboard with all the necessary information, a discussion tool for brief conversations within the projects, contacts, automated transaction calculating, and a progress bar to check each language word and segment status immediately. Moreover, the manager has valuable features like labels, backups, flexible role creation, and permission.

The Automating tools. Forget about manual files back-and-force. The smooth integrations with popular repositories like GitHub, Gitlab, Azure, CMS like WordPress, Storyblock, Intercom Help Center, and applications like Figma allow teams to reduce the time by simple content push and pull. The developers can work with separate branches and keep the project's versions clean.

The Translating tools. CAT features and powerful MT engines provide non-stop localization. The translator may never start from scratch - the prefill function admits to filling text segments from machine translation, translation memory, or existing language. The term base always provides correct term usage and style guide support with a single tone of voice, while context images give a better overview. Have no translators? That's not a problem - you can order the Lingohub professionals who care for your content (more than 40 languages are available).

translation management system


The field of translation is still relatively young, which does not prevent it from bringing many benefits to the world. Without thinking, people daily use translated applications, watch videos with subtitles, and learn languages on convenient and prepared platforms. The translation is standard, and the user is more likely to be confused by its absence than surprised by its presence. In a market with such needs, localization has become a necessary part of the expansion and promotion strategy. You create - Lingohub translates - with this slogan that our application was developed and continues to improve. It doesn't matter if you need to adapt technical documentation or a financial application - there are always convenient tools to help reduce risks and reach the world level. Order a demo to learn more about the product, and let's grow together.

Try lingohub 14 days for free. No credit card. No catch. Cancel anytime