The Elusive Uberization of Translation

The title of this post is borrowed shamelessly from the excellent article by Florian Faes found here.

Over the last decade, I’ve lost track of the number of times that the translation industry was gonig to be “reinvented” by this that or the other new model, usually one based on some sort of new technology. And each “reinvention” is accompanied by breathless hype and wonder in more blogs and posts by industry “experts” than you can count, many of whom should know better. Each “reinvention” is based on some plausible-sounding theory about the nature of the translation market that turns out to be just wrong.

These false theories and assumptions have no end. There would be an infinite demand for translation if only it were cheaper. There would be an infinite supply of translators if only we could tap the millions of bilingual people our there. Or no, the real problem is the efficiency of translation tools. Machine translation (MT) will eventually take over the lion’s share of translation work. Or it won’t. What we really need is to make it easier to order translation; where is my one-click translation button?

Problem is, none of these assumptions have proven to be true. The largest translation companies in the world continue to derive most of their revenue from large contracts with enterprises that have ongoing needs for translated documentation and software, carried out via well-defined processes and requiring well-defined quality levels. Startups based on the latest idea for a shiny new translation gadget flounder, ending up doing down-rounds to keep the lights on. Mid-tier translation companies continue to dominate the business in revenue terms, most doing semi-specialized medium-sized projects for medium-sized clients.

So how do we uberize this? The fundamental problem is to what extent translation is a commodity. Ride-hailing is the ultimate commodity: a car comes and takes you from point A to point B. There may be different sizes of cars, or special services like handicapped accessibility, but these amount to nothing more than different grades of coal. Ride-hailing is commoditizable because it is a commodity. We can aggregate demand, aggregate supply, and then tweak the economics of the business to death. To speak of uberizing translation implies that translation is a commodity.

Wikipedia defines a commodity as a “substantially fungible” item. It says a commodity is “an economic good or service when the demand for it has no qualitative differentiation across a market”. So to what extent is translation a commodity? I would answer very little or not at all.

First, translation has distinctly different quality levels. One relatively recent entrant into the market dubs its quality levels as “standard” and “business”. “Standard” apparently means “you can pretty much understand what it says”. Inexplicably, this company then quotes prices for translating a website at the lowest “standard rate”; more bizarrely, it also uses this price for estimating the cost of translating the microcontent found in an app. But in any case, “standard” and “business” can still be thought of as levels of a commodity. But interestingly, this company has now given up on the highest, third “premium” tier it once used, and now simply calls it “additional services” with the note to “contact sales”. In other words, this leader in new translation models has essentially admitted that translation is not a commodity, and that anything to be translated to a meaningful quality level requires contacting sales and getting a quote, and presumably, the company making special arrangements to select and price the right translators for the job. So much for commoditization.

Second, translation is inextricably related to the subject matter. Unlike Uber which offers three types of cars, it is meaningful to distinguish when translating among hundreds of subject areas, or more. A translation by someone not versed in the subject area is likely to range from unusable, to embarrassing, to damaging. And we cannot depend on translators to self-qualify themselves by subject area; the only way to qualify them is either through tests, or more realistically through evaluations of actual work that they perform. Since the client cannot reliably evaluate the quality of a translation into a language which they do not speak, and in many cases do not want to take the time and trouble to evaluate quality, the translation company needs to have a system to do internal evaluations, often using third-party experts. This involves more overhead and cost. We are getting further and further away from translation as a commodity.

Third, there is a learning effect over time in translation. Whereas neither you, nor the driver, nor Uber, learn anything from the first 99 rides that affects the 100th ride, translation is an ongoing learning process for the customer, the vendor, and the individual translator. This is the process which results in usable, viable translations. Whereas a key aspect of a commodity is low switching costs–I buy coal from Company A this time and Company B next time–in the translation world, switching, whether a customer switching a vendor, or a vendor switching a translator, comes with tangible costs in the form of discarding the accumulated learning of the past, whether that be about desired style, terminology, knowledge of the content, or project management approaches.

So translation is not a commodity, and therefore it is not Uberizable.

But wait. Isn’t the translation of tweets, or chats, a commodity? It is certainly closer to a commodity than other types of translations. But I believe the market for translating such snippets is vastly overestimated. In comparison to the value to Microsoft of translating its documentation–a requirement for selling its products into a foreign market–what is the value of having my tweet available in Chinese? Remember also that the majority of potential consumers of a translated tweet may well have enough English knowledge to understand a one-sentence tweet anyway.

It is with some puzzlement that I look at CSOFT’s Stepes app, discussed in the article referenced at the beginning of this post. According to the article:

CSOFT sees huge untapped demand for translation services that can be unlocked only if prices are reduced dramatically by using a crowdsourcing approach. Yao called this the long tail of translation demand.

But virtually every phrase in this quote raises questions. Where is this huge untapped demand? From what sector? Who has determined that all these users are price sensitive? How will crowdsourcing reduce prices? People place a certain value on their time, whether they are translating 50K words per month or 50 words per month. Do they actually understand the actual theory of the long tail, which can be very long but still have a very low area underneath it?

Translators on Stepes use a chat-based user interface (UI) that allows them to work line by line in a way that looks like they are texting.

So the key innovation in the translation industry is going to be a new UI for translators? We’re going to upend the world by moving from a document-based approach, or a table-based translation approach, to a chat-based approach? I know that chat-based interfaces are all the rage, with people putting up Slack apps to order lunch, but do translators want to translate as if they were chatting? What if they want to go back and fix something? What if they want to review a few paragraphs ahead in order to gain some context? Let me go out on a limb and say that a chat-based interface, whatever its positives for other usages, is not a good translation interface whatsoever. Even if it were, in and of itself it is a minor element of innovation in the translation business.

So let me on this first day of 2016 give my predictions for what the new year will bring for the translation industry. Bottom line: not much. Things will change at about the same pace as they have for the last 30 years, which is to say, not very fast at all. All the changes will be incremental. There will be incremental changes in translation tools. There will be incremental changes in processes. There will be incremental changes in machine translation. There will be incremental changes in marketing approaches. There will be incremental changes in translation company valuations. The industry will remain, for the foreseeable future, immune to radical innovation.

One Response to “The Elusive Uberization of Translation”

  1. Alex M Says:

    Good questions Bob. Please see Stepes’ response here:

Leave a Reply