Looking into the future of language services
by Max Zalewski
Max Zalewski is equal parts adventurer and logophile. He has been translating Arabic, Spanish and Portuguese into English for the last 7 years while living in Denver, Damascus, Barcelona, Aleppo, Madison, Cairo and Granada. Contact him at firstname.lastname@example.org.
Is the language service industry teetering on the precipice of obsoleteness or will it be at the forefront of an increasingly globalized world? On September 20th, 2014, Hélène Pielmeier, Director of Industry Provider Services for the market research firm, Common Sense Advisory, addressed the direction of the trade at the annual conference of the Midwest Association of Translators and Interpreters. The presentation featured an abundance of data about the language service industry with which Pielmeier created a looking glass to peer into the future and predict trends in the market.
Pielmeier commenced the presentation by introducing Common Sense Advisory and discussing the methodology it utilizes to collect data. CSA is the only firm dedicated to market research of the language service industry. One way CSA acquires its data is by surveying and conducting in-depth interviews with LSP’s, universities, and buyers of language services. There are two types of membership with CSA: buyer and LSP; at present, it does not cater to freelancers. CSA performs research from the perspective of both types of members. CSA defines language service providers (LSP’s) as companies that provide language services and have two or more employees. For LSP’s, they analyze management, strategic planning, profitability and growthundefinedspecifically, they observe the three pillars of growth: sales, account management and marketing. CSA also researches specialty services (transcreation, interpreting, etc.) and technology (translation management systems, machine translation) marketed by LSP’s. On behalf of the buyers, CSA monitors what they want, satisfaction, and their perception of the price versus quality ratio, among other variables.
In addition to surveys, CSA conducts consumer panels, briefings, and feature reviews. The market research firm is a pioneer in the landscape of the language services industry. Despite being a $45 billion dollar per year industry, CSA is the only entity dedicated purely to researching it. In addition to collecting its own data, CSA also monitors public data and filings, participates in online communities, and attends conferences. Each year, CSA produces reports, webinars, and longitudinal studies about the general trends of the language service industry as a whole.
At its essence, the language service industry aims to solve problems created by language. From the 7 billion people across 195 countries, CSA breaks them down into 687 locales, which it defines as geographic locations with “the minimum of unique combinations of economy, politics, culture, and languageundefinednot counting minority languages, individual states, etc.” Twenty-six of these locales are located within the top ten trading nations.
One way to monitor trade is through online commerce. CSA analyzed 2,400 websites of major corporations and found that 12 languages reach 80% of the online population. Furthermore, 90% of the most economically active people online can be reached by just 13 languages. Ordered from most to least economic activity online, these languages are: English, Japanese, German, Spanish, French, Chinese, Italian, Portuguese, Dutch, Korean, Arabic, Russian and Swedish. Interestingly, CSA does not expect the status quo to persist. On the contrary, it predicts that as a result of the overall online economy growing, more languages will be incorporated into websites. CSA forecasts that 20 languages will be required to cover 80% of online activity in the future. This represents an auspicious bit of news for the language service industry, both LSP’s and freelancers alike.
Common Sense Advisory’s survey of consumers of translated products produced some expected and unexpected results. It is well known that consumers will not buy what they cannot read, but where do people start to make decisions based on their willingness to use their second language? According to the international average, 55% of online consumers prefer to make a purchase in their local language. However, this figure jumps to 74% when consumers were asked if they preferred to have post-sales support in their local language. CSA’s data shows that even if post-sales care is available in English, and consumers have a good command of English as their second language, they strongly prefer to have post-sales support in their local language. Generally post-sales support is for when a problem has arisen. The discrepancy in language preference correlates to the change in the consumer’s mindset.
The demand for post-sales service in local languages is a microcosm of the overall trend that CSA foresees in the language service industry. CSA compiled data based on 831 responses to its annual survey that showed the industry generated $23.5 billion in revenue in 2009, and $33.05 billion in 2013. CSA anticipates the language services market will expand to $37.19 billion by 2018. Take a second to consider the magnanimity of those numbers. That figure means that the revenue generated in the language services industry is larger than the GDP of 40% of the world’s countries.
Based on these figures, the translation industry is growing at a rate of 6.23%, which is considerable, but below the past averages of double-digit growth. Pielmeier says that there are external and internal threats counteracting growth. External forces include economic recession, globalization of the work force, professional purchasing (big companies cracking down on rates). Internal forces consist of translation automation, changing nature of translation, and commoditization.
As expected, the regional concentrations of translation buyers are not spread evenly across the world. The breakdown for percentage of market share by continent is as follows: Europe, 51.09%; North America, 37.81%; Asia, 9.96%; Latin America and Caribbean, 0.48%; Oceania, 0.41%; Africa, 0.24%. CSA calculates this data based on where the LSP is headquartered, not necessarily where the translation is produced.
In addition to geographical location, sizes of the LSP’s vary as well. Sixty percent of LSP’s have only 2 to 5 employees. The final forty percent is divided as follows: 6-10 employees, 17%; 11-20 employees, 9%; 21-50 employees 7%; 51-100 employees, 3%; 101-500 employees, 2%; 501 or more employees, 1%. These numbers are based on 18,000 companies sampled. CSA used a larger sample size in the past but have since become stricter about separating freelancers from LSP’s. These figures demonstrate that the market is very fragmented and there are very few companies making large profits from translation.
The services provided by LSP’s are dominated by translation and on-site interpreting, but there are also growing niche markets like video interpreting, phone interpreting, mobile app localization, machine translation post-editing, and transcreation. CSA performed a study in which they broke down language services into 18 different categories. Translation is by far the most widely sold service at 34% of the market share, followed by on-site interpreting at 10%, and software localization at 7%. Translation is sold by 83% of LSP’s; however, very few LSP’s are invested in smaller rising niche markets like interpreting technology and localization. These services represent a great opportunity for emerging LSP’s as well as freelancers. An LSP or freelancer that provides localization and translation has less competition and a larger market share. Interestingly, translation represents three quarters of revenue for the aforementioned 83% of LSP’s who sell translation; however, many of the fastest growing LSP’s in the world revolve around interpretation.
Within the translation market, there is a discord between supply and demand in that the demand consists of many small projects, but suppliers want large projects. According to Pielmeier, smart companies are adapting to the small project market by creating retail portals for clients, thereby streamlining the overall process and foregoing quotes, signing contracts and explanations of the service. Clients can simply purchase translations online with a credit card and poof, it will magically appear on-time as promised. This industry is exploding thanks to the phenomena of convenient online purchasing. Language service professionals who want to fit into this changing world must use what Pielmeier calls the “agile adaptation methodology,” in which they get several small jobs throughout the day. On the side of LSP’s, technology is imperative in this market, demonstrated by the fact that companies that have adapted to technology have a much higher growth rate.
Technology is changing every field. Even if language service professionals are unconcerned about losing their jobs to a machine, they might want to think about how to adapt technology into their business practices. Among other technological applications in the language services industry, machine translation is the most discussed. Some fields of translation are more susceptible to being replaced by machines. Literary translation, for instance, is unlikely to be affected by increasingly accurate machine translations, whereas engineering and legalese, which are more patterned forms of writing, are already translated using a synthesis of machine translation and human editing.
What exactly is post-edited machine translation? In PEMT, the source text is analyzed by a translation memory, which has gathered patterns of past-translated material in both languages and creates a translation of it into the target language. Next, two types of post-editing can be applied: light editing and full editing. The light editing process only looks for lexical errors and syntax errors and mainly differs from the full edit in that it neglects the style of the texts and does not correct punctuation errors. The full editing process comes closest to human qualityundefinedideally equaling it. Pielmeier says that PEMT matching the quality of human translation is feasible, but it takes expert linguists trained to look for less obvious errors, match an appropriate style and still convey fluency.
According to forecasts from CSA, both machine translation as a managed service and in-house machine translation service will increase nearly double from 2013 to 2016. With regard to the increased demand in machine translation, Pielmeier offered advice to freelancers: “You are not at risk of being replaced by machines. However, you are at risk of being replaced people who are willing to work with files that have gone through machine translation.” She also noted that on the LSP side, it is hard to find translators who are willing to work with machine-translated files and that “it is a fabulous opportunity for newcomers to the industry.” Pielmeier joked that given the mystery inherent in the nebulous nature of technology, any problems can be blamed on the machine, right?
As mentioned earlier, machine translation is not likely to be used for all kinds of content, however online content in general is likely to be one of ways it is most used. Pielmeier stated that there simply aren’t enough human translators to account for the cornucopia of content uploaded to the Internet. The increase in the number of languages being used online will also lead to greater usage of machine translation because content can be translated into multiple languages at once. Lastly, another reason machine translation will be used more in the future is because the expansion of online content does not coincide with an increase in budgets. Some companies only budget enough to generate more and more content for their online profiles but do not necessarily increase their budgets enough compensate enough to pay human translators. The result is that they are willing to sacrifice style and opt for a faster and cheaper machine translation.
The willingness to negotiate the style or accuracy of content brings up an interesting question about the relationship between price and quality. In a survey of 839 buyers of translation services, LSP’s and freelancers were asked whether or not there is a direct relationship between the price and the quality of a translation. All three agreed that there was a direct relationship. However, freelancers believe price and quality are directly related more than LSP’s, who both feel that price correlates with quality more than buyers. Quality is not necessarily the only value that consumers consider when making a purchase, and as a result the traditional process of translation, editing, and finally proofreading has been reduced to a less costly one or two step process.
Pielmeier concluded her presentation with recommendations about how to best use this information. She suggested that no matter what your connection is to language services, you can use this information to understand how you want to fit into the overall puzzle. It is important to note that none of the data she presented directly represented freelancers; nevertheless, the data is still useful to freelancers in order to better understand those who ultimately purchase their services. As a business strategy for both LSP’s and freelancers, Pielmeier recommended identifying niche areas that will provide long-term work, such as localization. In general, being adaptable to technological advances is fundamental, for the best way to avoid becoming obsolete is to learn how to adjust the offer to what clients want.