Behind the scenes: the translation profession

Do you know how language service providers operate?

Today, I would like to take you on a brief journey behind the scenes of the translation profession and market. Our profession is often referred to as the “translation industry”, but we as professionals are actually not very keen on this. Why? Because we feel that it might give a wrong impression: “industry” might invoke the image of large-scale productivity (especially in the age of AI and machine translation), when there is actually only very limited possibility to scale our productivity.

In this month’s blog post, I would like to shed a light on things that I have found only few outsiders know and that might be useful for you to navigate the market.

Structure

The way the translation sector is set up is actually very simple. There is a large number of translation agencies (also often called “LSPs”—Language Service Providers), both big and small, and then there is an even larger number of self-employed translators and interpreters.

Most agencies employ only a handful of translators. Their business model is focused much more around outsourcing work to self-employed translators and interpreters and dealing with the project management aspect of language-related work.

Large translation agencies

According to the European Commission, the top 100 largest LSPs currently own only 15% of the global market share, whereas the remaining 85% belong to small to mid-size companies.

Nevertheless, they are usually the ones to pop up first on online searches. Due to their size, they have the budget necessary to invest in advertising and search engine optimisation. Large LSPs’ unique selling points include their vast network, which allows them to cover almost any language combination and subject area as well as to speed up large translation projects by splitting them across multiple people.

In most cases, projects handed off to these large LSPs are not actually completed in-house, but outsourced not only to self-employed translators, but also smaller translations agencies, who in turn outsource work to self-employed translators or other agencies. This may lead to inconsistencies due to a long and complicated supply chain.

Boutique translation agencies

One step down the chain are smaller LSPs with a dozen or so employees, operating on a more regional and less global scale. They often take on work from the large agencies, but also have direct clients of their own and often focus on a certain specialisation or on specific language combinations. Their unique selling points are a more personal service, specialist knowledge of a certain field and better consistency in their products.

Self-employed language professionals

The vast majority of the translation work done will, in the end, filter down to self-employed translators and interpreters. They usually specialise on one or two language combinations and a narrow field such as engineering, medicine or tourism and gain a lot of experience in the language used in their field of choice. While they only cover a very narrow scope of work, they are experts who know what the field requires.

A lot of them will get regular influx of work from agencies both large and small, but many also work with their own direct clients who appreciate the personal one-on-one service and the fact that the language professional of their choice knows exactly what the company needs.

Market size

With more and more companies striving to increase their footprint by going international, the language market has been experiencing constant growth over the years.

According to Statista research published in January 2023, the market doubled in size in the last decade. Even during the COVID pandemic, it continued to grow, unlike many other sectors of the global economy.

As mentioned above, one of the main drivers of this development is increasing globalisation and interconnection. But there is more to it: digitising increasing amounts of content, more personal and globalised customer support and the consumption of media are just some examples.

But what about emerging technologies? Are human translators about to go extinct?

Current trends and changes

Machine translation

Did you know that machine translation (MT) is not actually a recent invention? It dates back to the 1950s, when the first statistical models were developed to automate transfer of content from one language to another. And even back then, the scientists involved were convinced that human translators would soon be replaced.

Fast-forward over half a century, human translators were still very much in business. The quality and technology behind machine translation had steadily increased over the years, but the first real jump in fluidity came around 2014 with the introduction of neural networks and machine learning. But even though many translation engine developers imply that they have reached human parity, we are still far from eliminating human translators from the workflow—in fact, more and more people are needed to edit said machine output to bring it up to the required quality. (Which, incidentally, has been shown to be inferior to the quality of a human translation without machine aid.)

If you would like to know why MT needs editing, check out my previous post on the disadvantages of machine translation.

Artificial intelligence

But what about artificial intelligence (AI)? I think it would be fair to say that we were all impressed by how natural ChatGPT’s output sounded when it was first released. It almost feels like the algorithm understands what we are asking it to do. (Spoiler alert: no, it doesn’t. It is just very good at computing the most likely output based on numbers. It does not deal with words at all!)

So, is AI a better translator than MT? Yes and no. It has some distinct advantages compared to normal machine translation engines, e.g. because it allows you to give more context or specific instructions on how to translate a text. But just like in other areas, it can also output faulty text, including hallucinations and other issues. If you would like to know a bit more about this, check out the post I linked above.

Special needs awareness

One of the most positive developments I have personally spotted over the last few years, is a focus on accessibility of content. There is now a host of tools to help authors assess the readability of their texts, but also the demand for simple language versions of texts is growing. A considerable number of language professionals has undergone training in how to produce accessible texts and can now help customers to reach certain audiences better.

Consumption of media

One of the side effects of the COVID pandemic was the increase of interest in video streaming platforms. And with it came an increased consumption of international video content. But while a lot of video content is created in English, the lingua franca of our modern world, not everyone would like to watch it in English. Maybe because they are afraid of missing out on vital information, or maybe because they would like to just enjoy the entertainment after a hard day’s work instead of having to work hard to be able to understand every detail of what is being said on screen.

Whatever the motivation may be, audiences prefer content in their native language. And with an increase in international productions for such streaming platforms comes an increased need for dubbing, subtitling and closed captioning services.