The Invisible Hand

Invisible handIn economics, a free market is a system in which the prices for goods and services are self-regulated based on supply and demand, free from any intervention by a government or other authority.

Free market is often synonym of open market, which actually refers to a market that is accessible to everyone having equal opportunity of entry.

Alongside with the concepts of free and open market goes that of the invisible hand describing the idea that trade and market exchange automatically channeling self-interest toward socially desirable ends.

The translation market is in a state of close-to-perfect competition, where no single player can set the price of the goods produced, which results from the meeting of supply and demand. For being an actual market, as free as imperfect by definition, it has many of the idealizing conditions of perfect competition, except of course for information asymmetry.

The extreme fragmentation is one of the conditions making the translation market an imperfect market, and yet peculiar as a free and open market. It is a consequence of the virtual absence of entry barriers. Having been a genuinely global market per se, especially after the emergence of the Internet, dispersion—way beyond spreading—goes along with fragmentation.

Another, often untold, aspect of the translation industry is its ‘incestuous’ nature. This industry goes way beyond being just based on personal relationships, making who you know much more important than who you are and what you do.

In 1955, Edward C. Banfield observed a self-interested, family-centric society, which sacrificed the public good for the sake of nepotism and the immediate family as the basis not only for a backward society, but for Mafias. Banfield built his most (in)famous book around the postulate that the backwardness could be explained by “the inability of the villagers to act together for their common good or, indeed, for any end transcending the immediate, material interest of the nuclear family”. Those villagers were intimately distrustful, envious, and suspicious of one another and they would refuse to help one another unless their own personal material gain was at stake. Therefore, isolation prevailed, and, with it, the nuclear family, giving birth to an “amoral familism”.

The depicted society is not exclusive to specific and even less backward regions. Indeed, the obstructive potential of its behaviors can be found everywhere. The code of conduct that thwarts any efforts for a “social capital” also prevents working for a “common good”. In the case of the translation industry, this results in a tribal, familyist, almost incestuous conduct. In fact, the extreme fragmentation of the industry also produces a series of peculiarities that makes it impossible to pursue a transversal career in other industries and leads to be confined in a somewhat restricted area, despite the extension of the network of contacts that one can build over the years. The closer, older and stabler are the relationships that one manages to develop, the higher the chance of enjoying the benefits that come from them. This is also one of the reasons for the plethora of industry events that have long seen the same audience of mid-level industry people. In turn, this tells why there has never been any real financial curiosity for the industry and why, only now, the first expressions of interest are being registered, although from PE and VC firms rather than from direct investors or companies from other industries. Indeed, in the past, when having some stake in the industry, these companies have got rid of it.

Despite the freedom and openness of the translation market, the common approach, attitude and conduct of its players have been proving the unsustainability of the invisible hand theory. Should it ever have existed, the amoral familism of the translation industry would have at least been corrected somehow.

Collateral Effects

A major collateral effect is the great difficulty to run surveys and collect reliable results. Firstly, the extreme fragmentation makes it impossible to build up statistically meaningful samples that would inevitably be huge with the resulting data being hard to analyze and interpret. A possible solution would be having local offices run the surveys, but this would turn out to be terribly costly. Perhaps survey reports would be more reliable, but much more expensive, thus even less appealing than the current ones. It is no coincidence that, for a year or two now, analytical reports on the MT market have been proliferating, while the only survey reports on the global translation market are the usual ones from the usual analysis firms, whose reliability can be legitimately disputed given the remarkably different results of their “research”.

Not only do the variations in market estimates and their magnitude suggest that reliable data is hard to find, they also suggest that the research methodology might be badly flawed, by design. Together with the difficulty of defining and differentiating market segments, all this makes market sizing remarkably challenging.

The major flaw in surveys is questionnaires. Not only does everybody lie, it is also impossible to introduce control questions to cross-check responses and allow for corrective measures. Consequently, the only solution would be using the data collected by local inland revenue agencies; this would require having local offices, at least to avoid translating the official reports.

Everybody Lies: The Italian Case

Recently, Italy’s inland revenue agency has introduced synthetic reliability indicators (ISAs, Indici Sintetici di Affidabilità) to replace statistics-based tax assessment (Studi di Settore, SdSs) as a tool to detect the basic parameters of freelancers, self-employed workers and organizations. Like SdSs, ISAs come from the systematic collection of data about the nature of business and the economic environment of operations over several tax periods. Like SdSs, ISAs are used to assess a company’s capability to generate revenues and their consistency.

ISAs express a company’s degree of tax reliability on a scale from 1 to 10. A value up to 6 will express unreliability and have the company listed for tax inspections.

Like it or not, regardless of the Italian habit for tax evasion, ISAs confirm the belief that everybody lies.

In fact, the latest ISAs (2018) report for “Language services and organization of conferences and exhibitions (AG53U)” a total of 6.184 players accounting revenues for € 245,3k on average; 59% record an ISA below 8, which is considered as the benchmark. Freelancers are 3.507, accounting revenues for € 47,7k; only 26% of them record an ISA below 8. Sole proprietorships or partnerships are 609, accounting revenues for € 176,8k; 59% of them record an ISA below 8. There are 2.068 companies accounting revenues for € 600,5k on average; 55% of them record an ISA below 8. Based on conservative estimates, the Italian language industry is worth roughly € 1,52b. Being Italy’s GDP roughly 2,3% of the world’s GDP, the global language service market could be reckoned to be worth roughly € 66b. Considering as conservative and yet fairly accurate the estimate of the translation and localization industry as accounting for roughly 40% of the whole language service industry, the global translation and localization market can fairly be estimate as being worth roughly US$ 31,44b.

Also, as the numbers of the Italian market show, the language service industry is not exactly enticing, and the growth boasted as largely above that of the world’s GDP looks at best very optimistic. It is certainly possible that the LSPs that are or have been able to work on large volumes have seen surges in growth, but these have certainly concerned revenues rather than profits.

It looks like being still in the early 1990’s, when the only trade magazine was filled with articles about amazing companies and brilliant entrepreneurs, with thousands of industry players being left in the shadows for not having the means to access the magazine. It would have come as no surprise if anyone who had detachedly followed events in the industry anticipated the loud L&H debacle, and yet, even today, some still pretend to believe (and would like to make others believe) that it was an unfortunate and unpredictable incident.

The “Good-Enough” Economy

As Shira Ovide wrote in On Tech, the technology newsletter of The New York Times, today “good enough” rules. “Good enough” is underappreciated because it looks like an insult. And it is, as long as mediocrity is admittedly fine. The good-enough economy thrives on the principle that a lot of good enough is better than a little perfect that is most likely hard to find or costs extra. Perfect is the enemy of good.

What most of its players like to think is that the translation industry is not based on the good-enough economy, when in fact it has been since inception in the late 1970s.

While in other industries the perceived quality comes as a result of assessing the finished product and expectations from branding, the boastfulness of many, too many players has a devastating impact on perceived quality, which is all the higher the more insistent is the unique selling proposition based on quality.

Also, while software as a service (SaaS) is “eating the world”, with the market for “vertical SaaS” being particularly strong, tripling in the last decade, the translation industry is still skeptical that the returns might not be lucrative enough to warrant the necessary investment of resources or, worse, it is afraid of any security issues. As a matter of fact, security has always been an issue and a matter of significant concern when it comes to cloud computing even though most people use it all the time without even realizing. And yet, cloud-based systems are generally far more secure than conventional, local ones, and security is a compelling reason to use cloud-based systems rather than avoid them.

Also, many large, sophisticated vendors are now capable of offering several vertical SaaS solutions for a full suite of functions to companies in specific industries, allowing customers to quickly scale and operate more efficiently.

Most translation industry players lean on believing the industry is dynamic and always on the rise, and that they always live exciting times. They argue that there are many perks for people who are looking forward to making a career in translation, because of the sophistication of translation technologies and the many changes that translation and its industry have been going through. Really?

Actually, most of the translation industry is still stuck with local systems focusing on a thirty-year old technology (translation memories) and scared of machine translation, although it is being more and more used outside and—quietly, very quietly—inside the industry. It has, in fact, been adopted as an integral part of the translation process, and this is possibly a major reason of success for PEMT and the proliferation of the relevant services.

Recently, from the event “The Evolution of Localization” a few facts emerged that picture the present of translation and localization more than their evolution and hence their future. Firstly, the impressive volume of 2b words processed per month would represent some 4,8b in yearly revenue, that would, in turn, according to certain estimates, represent 10% of the whole market, whereas ten companies together barely reach 20% of the market, and a remarkable number of licenses too.

In the second half of the event, Semantix’s Britta Aagaard, Supercell’s Silvio Clausen, MIIS’s Max Troyer, and Memsource’s Martin Švestka talked for twenty minutes about the cloud, the benefits of working remotely, of centralization of resources, and of course of automation although it is actually still far from being achieved in full.

Let’s be honest. The translation industry is in its forties and the progress it has been seeing has been noteworthy, in perspective, kind of passing from ice age to the industrial revolution in a flash. But the world has long been in the second half of the chessboard and no one can seriously imagine any serious Silicon Valley business team having a conversation like that one.

This is a major reason if localization is still often seen as a cost rather than an investment because customers cannot see it as adding any value.

And yet, there are still people thinking they understand—and can teach—the principles of marketing and value and ask why a client would not go for a sweatshop translation agency and are prodigal of answers and advice to fellow localizers and translators. The answer is quite simple: A client would resort to localizers and translators for them to solve “a big problem”, which can get bigger according on how localizers and translators frame their services. Again, here is the vexata quaestio of educating the customer. On the other hand, it is still a common belief among industry players that this is an absolute necessity, despite the equally widespread idea that failure always comes from not understanding the customer needs, from focusing on company processes, not on client issues, thus simply dismissing people who are not from the translation industry or familiar with it as to be ‘educated’.

Also, customers are rarely willing to be educated by someone who does not belong to the same class of business. Nor are they willing to hear from this someone that they have a problem and that only this someone can solve it, in a sales approach that is outmoded even in door-to-door.

Economic Effects

The translation industry has all too often been said to be “a growth industry”. Not a growing industry like many others, which may anyway be subject to cycles: The translation industry ostensibly features specific traits, however hard to see, especially for an outsider.

Growth is a major reason for investing, and yet the many uncertainties in sizing the market make investing in the industry very hard. In this respect, an old post in the TAUS blog asked some very good questions like, “What, exactly, does the buyer get when acquiring a translation company? What is the market value of a given translation company as a whole? And how does the market put a price on a company’s individual assets?”

The article plainly and brilliantly pointed at three intangible assets as decisive in an acquisition, well beyond tangible assets (i.e. real-estate, equipment and office supplies):

  • Human resources: The company’s in-house staff, as well as its freelancing personnel. However, a value is hard to assign for these personnel that are often shared with many other companies. Also, the reliability of a vendor base that is often overlooked, especially when growing fast in size, is easy to question. Rarely do LSPs keep contact details of their freelancing personnel up-to-date and even less rate them routinely both in a professional and financial perspective.
  • Clientele: A rich and detailed database of appealing clients makes for an attractive acquisition. However, a key issue before the acquisition is the loyalty of such clients and once the acquisition if finalized, a key issue is how to keep these clients loyal, especially with some major change in management.
  • Technology and patents: Apparently these are potential assets, but it may be an issue when the buying company has its own technology which is conflicting with that of the acquired company. This could also be an issue if the processes of the two companies rely heavily on one or more specific technology. In both cases, this might turn to be not an asset and affect the overall appraisal.

The same article signaled that, for the first time, in 2015, the list of the fastest growing companies published by Inc. magazine also had a European version and that the translation companies in it appeared in the context of other industries. Also, most of them ranked rather low in the list, while the largest industry players, which nonetheless the usual suspects constantly praise for their supposed growth rate, are not mentioned at all on the list.

FT 100, the annual list of Europe’s fastest-growing companies, does not contain any translation or localization company. And so does the Inc. 5000 2020 list.

Is this a possible BS effect? Just like the “translation-big-data” crap showing only the incommensurable ignorance and stupidity of its advocate.

The usual suspects might keep telling the same old story till exhaustion and one might also buy it, but in doing so they would do nothing but damage the industry and perpetuate the invisibility that has been affecting it for decades.

An Urgency for Innovation

The invisibility is a consequence of the lack of innovation, real and perceived. Every so often, no disruptive innovation is necessary, as the case of curbside pickup tells. Likewise, a disruptive innovation may not be perceived as such, at least immediately, or may not last over time.

The curbside pickup is the perfect example of the difference in pace between the typical change management and innovation following the urgency aroused by a crisis.

If the translation industry’s future is about being either super-specialized or super-sized, with no room for anything in between, players must create a sense of urgency for major innovations and learn to improve their ability to innovate at any time. They might start by setting high-priority, fast, short-term, actionable goals in an innovation effort built from small experiments. All the experiments will hopefully pile up and end in a mission-critical change.

It should be clear, then, that this is not the kind of innovation to which, for example, SLATE, SDL’s self-service, on-demand platform, belongs. Not only is it not a real innovation, since it was expected and rather belated, but mostly because it never seemed participated and convinced. In fact, consumers have long been able to choose between other self-service platforms (like Translated.net, One-Hour Translations, Lingo24, Gengo, tolingo, etc.).

Is there any innovation in CAT tools? In TMSs? In MT? In PEMT? Definitely not per se. Technology is not enough and it is not necessary to innovate. People and processes are. For example, you can implement and run sophisticated technologies to order and bill, not to deliver mugs and bags for curbside pickup, as the real innovation is in the very simple idea behind it. Without any help from any AI.