Localization in the era of disintermediation

As more industries are affected by a move toward a shared economy model, let us explore what impact this model could have on the localization industry.

Disintermediation is the elimination of the traditional middleman by online sources — the deletion of the intermediary between the seller and the buyer (such as an agent, broker or reseller, or between the source and the recipient of information (such as an agency, official or gatekeeper).

There was a recent article on Fortune.com that explained the impact that Airbnb is having on the hospitality industry. And while some analysts in the articles were predicting dire consequences for the hospitality industry, leaders in that sector seemed to indicate that they had not seen any impact on their revenues. There could be, of course, an explanation for the latter phenomenon — business was generally good in the United States in 2014.

Another explanation put forward by the industry is that Airbnb is actually catering to a different market. But the key quote in the article comes from real estate investor Thomas Barrack, “Everything is being disintermediated. Do we really need more office buildings?”

Certainly, as a commercial real estate developer, Barrack is directly concerned. While what he is saying is not entirely true, it’s clear that disintermediation is having an effect on many industries, such as travel, computer hardware and software, books, stock brokerage and more. Others, like the localization industry, are only beginning to see its effect. Take, for example, Uber, and the impact it is having on the transportation business. As mentioned above, Airbnb is a good example for hospitality, but we also have Zillow in real estate. Apps and websites now exist to share parking spaces, tools, cars and so on. And let us not forget one of the oldest examples — eBay.

What is the impetus behind this movement? There are many factors at play.

Financial. The possibility of earning extra income from existing assets is one of the factors. We all have possessions that we might be using only intermittently. If it were possible to share their use and earn some money doing it, many people would take advantage of the opportunity.

A need to use the resources that we already have. Not all of us feel the need to save the world through recycling. However, for most of us, there is a realization that in many cases, we have assets that are not being put to use. Sharing them could help reduce the waste created by oversupply. This is the basis for the shared economy model.

The need for more personable or customized solutions to our problems. Just as we all have idle assets in our houses or driveways, some are looking for these same assets on a temporary basis. Existing industries already provide some solutions. For example, planning a visit to a different town could involve booking a hotel for a couple of nights. Nevertheless, sometimes you need more than a hotel room. You need more space, you need the freedom that having a full apartment gives you or you need the social interaction that sharing a house with someone will provide. Or you need a quality of service that you do not feel is possible for the price you want to pay. An example of that is Uber. You might need more than a taxi, but you do not want the expense of a limo!

Certainly, until now, some of these assets were already available on the market. You could rent an apartment before Airbnb appeared. But that generally involved an intermediary such as a real estate agent. And you didn’t have the flexibility of a rapid solution to a temporary problem, which Airbnb now offers. What has made all this possible is the greater availability of internet access everywhere, on multiple devices, and the possibility of gathering, centralizing, organizing and displaying the information needed to support all the transactions.  The ease of making any of your idle assets available online on specialized websites facilitates the exchange. Plus, having direct contact with potential customers also lets you provide the personable solution they desire.

The advent of electronic payments has also had a positive impact on some industries. Being able to make payments directly from your phone means not having to reach for your wallet when you ride with Uber.

So, what impact will disintermediation have on the localization industry? Even though labor in localization has historically been provided for the most part by independent contractors, the industry itself has been largely intermediated. Most of the localization services sold today are through a localization service provider (LSP). Furthermore, many translation tasks are assigned by the very largest LSPs to single language vendors (SLV) which adds to the level of intermediation. The translation tasks not provided by LSPs (or SLVs) are mostly done internally by large organizations that have their own translation departments.


Factors behind the

current situation

This process will affect both buyers and LSPs. Interestingly enough, some of these factors affect the potential buyers of translation services as well as the smaller language vendors that could be beneficiaries of the process of disintermediation.

Lack of understanding. The biggest factor here is probably the lack of general understanding of what the localization process entails. For many potential users, that process is just an afterthought for any of the work that they perform, and they generally are not interested in finding out more.

Buyers of localization services might in many cases only be involved in the file transfer process — sending files in the source language and receiving the corresponding translations. What happens in between remains somewhat mysterious.  In addition, the process usually involves having many participants working either in parallel or serially. This requires a lot of coordination, which, once again, goes far beyond what the buyers might be willing to undertake.

Translator selection. Most large localization vendors already have a vendor management department whose sole task is to recruit translators, editors and other possible contractors. It is up to this department to negotiate prices, check on availability and evaluate the vendor performance. There are many translator communities where some of that information might be available, but that is generally not enough to put together a team to respond to specific translation requirements. And these communities do not offer the tools needed to manage vendors and the localization resources. Unless the translation buyer is willing to spend the time needed to put together translation teams for all the languages needed, localization vendors will have a large advantage that is hard to match by the buyer. This is also a factor impeding the potential growth of many SLVs. Many of them lack the resources and tools to put together the translation teams that are required for the processing of more complex projects. And they certainly do not have the resources to maintain and monitor that larger pool of translators.

Translation tools. Another roadblock so far is the possible complexity and cost of the technology required in the translation process. Translation tools are generally complicated to use without the appropriate training and the licensing schemes make adoption potentially expensive, especially when dealing with intermittent projects. In the case of all-encompassing translation workflow systems, the cost and IT infrastructure needed for the implementation are prohibitive for most smaller localization vendors that are therefore locked out of many potential projects easily achievable with the appropriate tools. Localization buyers will generally shy away from making large investments in translation technology.

IT infrastructure. Translation project management requires planning an IT infrastructure. No matter what technology choices the user makes, changes will need to be made to the existing hardware. In some cases, such as when implementing a hosted translation workflow system, these changes can be extensive. New servers, database software and other applications will be required. The user will also need to plan for some centralization of data. The cost effectiveness of many projects today depends greatly on having a central repository for all the translation data in order to maximize reuse and availability for consistency of output. Similar to the case for translation tools, these costs are generally too high except for the largest localization vendors and buyers, leaving the rest of the industry to deal with less comprehensive solutions and forcing increased intermediation in the process.

These factors are, for the most part, not present in the other businesses moving to a more disintermediated model. For example, renting a place is not a complex process, and there are no special tools needed. Once a payment mechanism is in place, the process is quite simple. The common factor with all these sectors is the IT infrastructure.

For instance, Airbnb provides the website where all the data about rentals is centralized. They can then present that data in a format attractive to potential clients and make the experience enjoyable — from choosing your destination to rating the experience. The same could be said of Uber, which provides the IT infrastructure to allow its drivers to navigate quickly from point A to point B, to facilitate the payment process and the hiring of cars. Other barriers certainly exist, mainly regulations, if we are talking about examples such as Airbnb and Uber.

Airbnb and Uber are facing opposition from existing players in the industry that are very familiar with the regulatory environment and that are lobbying local politicians to make it harder or impossible for the new models to penetrate the market. This is not a big factor in the localization business, though. It is possible that the marketplace model will not be applicable to every segment of the localization market for multiple reasons, but regulations will probably not get in the way of expansion.

The possible roadblocks to intermediation have now been identified. So is there demand for this model in the localization industry?

Firstly, customers are looking for more customized solutions. They are looking for a platform that allows them more control over the process and gives them easier interaction with the people performing the localization tasks. More localization tasks are centered on web content where the customer might be looking for a certain tone in the translated content, and direct interaction with the translators is crucial as a result. Customers also want the freedom to choose the localization model that matches their quality requirement as well as their budget.

Smaller localization vendors are also driving the demand for the tools behind that model. They are looking for new growth opportunities, and they realize that tools on the market now will allow them to offer services usually performed by larger vendors for more attractive prices. Furthermore, these smaller vendors can now bring potential customers closer to the people actually performing the tasks. The access to process automation, resources management, payment processing, community engagement and real-time collaborative features will continue to drive the demand for more customized solutions higher.

And that brings us to the next factor — cost. By removing an intermediary (or several intermediaries, in some cases) from the process, costs are being driven down. The customer can access thousands of vendors offering their services at various price points.

A solution where intermediaries are removed will generally allow for faster processes, especially with a cloud-based multiuser application. The demand is there, but do we have the solutions to overcome the possible roadblocks?

The response to infrastructure roadblocks is getting easier every year. Most companies are moving their document processes to the cloud, so it makes sense to go in that direction with our solution to the localization process. A cloud-based application does solve many problems at once.

Obviously, infrastructure requirements are eliminated. Also, and most importantly, linguistic resources, such as translation memories and terminology databases, can be accessed in real time and shared among users. This ease of access and sharing has a great impact on the consistency of the translation especially in an environment where quick updates are performed on a regular basis. With desktop or server-based applications, translators are usually provided with a localization package that contains only a “relevant” subset of linguistic resources. In some cases, it is not sufficient, as the application only considers segments that are statistically relevant to the new document, not necessarily terminologically relevant. This system is imposed by desktop and server based applications because exchanging entire databases is not practical due to their sizes, and maintaining updated databases at language vendors is next to impossible. In some cases, it is not even allowed.

The cloud-based solution also reduces the complexity of the process. One of the reasons that most customers end up contracting localization vendors for their translation requirements to act as project managers is that most projects involve not only daily communication, but also file preparation and file transfers. With a cloud-based solution, file transfers are almost non-existent. Ideally, the end-user would need very little knowledge of the actual process since the system would take over these complexities.

The last and biggest hurdle is in the selection of vendors to be assigned to the localization tasks. Generally, localization vendors maintain a database of possible candidates for assignments. Once a project is approved, either the vendor manager or the project manager will contact the vendor to check on availability and then proceed with the file transfers. Administrative tasks are grafted into that process to handle payments. Of course, potential customers should be shielded from these tasks whenever possible. This is where marketplace models come into play. By allowing the customer to access thousands of possible freelancers directly within the application, rated and sorted by specialties and languages, the task of selecting vendors becomes much easier. Financial transactions can also be simplified with online payment models.

In this manner the localization process is greatly simplified in most cases, and bypassing the standard localization vendors is now a distinct possibility. With centralized linguistic resources, project management features and the basis of a vendor management solution, a cloud-based system can be the answer to many potential clients.

There are a few possible factors slowing down the generalized adoption of a cloud-based solution, however. If the past is any indication, security will always be an issue when documents are in a cloud environment. Nothing can prevent a determined hacker from cracking open a document repository, no matter how secure it looks. This issue might prevent some sensitive sectors from adopting a cloud-based system. Interestingly enough, the overall safety of the information in the cloud is much better than in a desktop-only environment where the data is only as secure as the latest backup.

There is also the potential of connection issues. Most of us take a broadband connection, in widespread use around the world, for granted. But there are still places where they are at a premium. Those most affected would be the more exotic language pairs, so this problem is not of great concern for most companies.

Integration with existing models is a common problem encountered when changes occur. How will existing processes be affected? Is the interchange of data seamless or lossless? Few easy answers solve this issue. Generally speaking, the translation industry has been adopting interchange format over the years. Therefore, we have solutions in most cases, either for translation memories or for terminology. These solutions are rarely lossless, though. In addition, some tool providers are using proprietary formats for the exchange of information, making it sometimes difficult to integrate a new tool into the process.

Some complex localization projects requiring many localization engineering tasks, such as software engineering, testing and desktop publishing, among others, might not be the perfect environment for a cloud-based system. In no industry can there be a perfect solution to all problems, and the localization industry is no different.


Going forward

As in other industries, the localization process is entering a new era of disintermediation. Because of the perceived complexity of the process and the applications, localization has not been at the forefront of the movement. Historically speaking, however, the transition to this new model could possibly be more painless than in other industries for multiple reasons — the lack of a regulatory environment, for one, but, most importantly, for a workforce already based around a freelance model. We are approaching a time when many potential localization clients will be able to build closer relationships with the people producing the final output. And they can accomplish their goals with the ability to build processes that match more closely their requirements for quality, cost and, especially, timeliness. A strongly targeted message is now more attainable than ever. It is possible that with the reduction in effort and cost achieved with this new model, a larger portion of our daily output will end up being translated. This tendency will result in more work for professional translators as well as more harmonious communication.

Does this mean an end to the current localization process? Probably not. Just as Airbnb will not signify the end of the hotel industry or Uber the end of taxi and limousine services, a new disintermediated localization model will hardly sign the death warrant of the localization process as we know it. There are still many projects that can benefit from the resources and the experience that only very large localization vendors can provide. This new model is hardly a zero-sum game. There will be some losers, but mostly winners. Customers will benefit from lower costs and a more customized experience while smaller localization vendors will be able to expand their offerings to compete against larger companies. Freelancers will also gain new outlets for their services. All in all, disintermediation is a win-win strategy for everyone involved in the localization process.