Evaluating the True Cost of Translating Internally
In today's global economy, companies no longer have the luxury of offering monolingual products. In order to survive, let alone thrive, they must produce national language version products if they want to sell outside of their national borders. Some organizations have opted to perform this task internally, one of the primary reasons being cost.
Many of the costs associated with in-house translation are direct, up-front, and definable. Other costs are less so and are easily overlooked. Nonetheless, they need to be explored and where possible calculated within each organization's own unique environment.
Another significant factor that may contribute to the cost of translation is timely product delivery.
This article briefly explores the reasons why certain organizations translate internally, as well as what some of the associated expenses are. A sample translation project is dissected in order to scope it out in terms of true cost and delivery, highlighting some issues which need to be considered. This article also discusses the hidden costs to organizations that have translation work done with in-country staff on the premise that it is "free".
The objective of this article is not to spell out and quantify all of the possible factors contributing to the cost of translating internally, but rather to act as a catalyst to suggest to the reader what may be considered. Decision-makers within organizations need to continuously evaluate the efficiency of their operational models in order to ensure resource optimization and ultimately commercial survival.
Much like death and taxes, there is one issue that organizations that wish to succeed in the global marketplace can count on dealing with: translation. Sooner or later, target market requirements (be they driven by clients, prospects, competitors, legalities, or otherwise) will force the question of how to "do translation".
Why and How Do Companies Translate the Way They Do?
Most organizations that need to deliver national language versions (NLV) of products on a continual basis will have made the decision to dedicate in-house personnel to the translation effort. The level of in-house resources varies depending on several factors including:
- corporate culture
- financial capability
- number and type of languages to be translated
- complexity of the task at hand
Corporate culture very often dictates the approach to translation. For some of the companies I have worked for, control over every aspect of the process was paramount from scoping the project, dictating the tools and methodology to be used, and controlling delivery dates. The level of in-house resources responsible for NLV products was quite sophisticated and included dedicated management, engineering, translation, lexicography, and quality assurance (QA) staff.
Organizations that have adopted the culture of outsourcing some elements of project work may have a dedicated project manager along with linguistic QA resources who liaise with outside entities such as translation/localization agencies. Still others may outsource the entire process - including overall management.
Creating and maintaining an in-house translation department is not an insignificant proposition. Companies that insist on or prefer doing the work with in-house resources are likely to be mid to large multinationals with a robust product line and services offering. They have a continuous volume of work which justifies hiring dedicated staff, in addition to the international revenues generated from multilingual product offerings.
The number and types of languages may also contribute to the translation effort organizations decide to handle internally. It's one thing to support five languages with internal resources, but quite another to produce and support 10 or more languages on an on-going basis. At my current company, we service one client who requires content to be translated into over sixteen languages, and they are adding more languages as market growth allows. Another client that we worked with was committed to making its product available in every language that Microsoft translates Windows components into (currently 44 and counting). And whereas language groups such as FIGS (French, Italian, German and Spanish) are fairly easy to work with, many organizations quickly shy away when faced with a long-term commitment to translating ideographic languages which often involve not only natural language translation but quite possibly double-byte enablement, and the acquisition of specialized software for complex desktop publishing.
Finally, the nature of what is being translated is also a key influencer in how and where translations are done. Documentation translation for the most part is a fairly straightforward process; localizing and validating application software for multiple platforms gets a bit more complex.
Irrespective of the organization's level of internal resources responsible for translation, each translation project incurs both direct and indirect costs. Very often, it is the indirect costs that get overlooked when determining the true cost of producing translated content.
More than any service that I have ever been involved with, foreign-language translation continues to raise eyebrows (at least in the U.S.) and solicits the invariable exclamation of "I had no idea that it would cost so much to get this translated into ." And this is despite the notable progress that U.S. companies have made since I joined the industry in 1982 in 1) recognizing the need to translate and 2) investing in human and non-human resources necessary to get translation done.
What's Does It Cost and When Can I Have It?
Once a program, website, document, training program, or other content has been identified as a candidate for translation, the race is on to scope out the project in terms of price and delivery. The sheer volume of content that gets translated on a daily basis around the world has forced companies to become savvy purchasers of translation services. Those with internal resources have the luxury of opting to have the work done themselves. However, depending on the scope of work to be done, the required delivery, or a lack of technical expertise, these companies may still request quotations from professional agencies or translators with the objective of getting the most for their money. They then compare the cost of outsourcing the work to the cost of doing it themselves.
Unfortunately, many companies do not calculate their hidden costs when comparing internal costs to an agency's flat $.xx per word rate. And that is often the yardstick used in determining which is the most cost-effective way to get the job done, the perception being that many of the services used internally are free.
Let's explore what is involved in translating a product internally in order to try to determine what the real costs are, and I will be the first to admit that this is very difficult to measure. There are so many variables that need to be plugged into the equations, and the equations themselves vary greatly from organization to organization. I think the reader will understand the points being made in spite of assumptions that I am making, such as:
- All resources involved are salaried employees with full benefits
- Our sample organization's dedicated translation team consists of a manager, an experienced translator, a technical resource, a desktop publisher, and an in-country reviewer (ICR) located in the target geography
- Salary and loaded cost assumptions are listed in Table A below
- The product to be translated is a User's Guide developed in an authoring tool designed for this purpose (we will assume Adobe Framemaker for purposes of this example, which we find to be a very common authoring tool for manuals).
- Translation will be done using Trados Workbench which is probably the most widely used translation toolset used by professional translators.
- Project scope is detailed in Table B
- Quality Assurance is executed by the company's in-country staff
- Other factors used in the exercise are based on personal and historical productivity data
Overhead Allocation Rate
Loaded Hourly Rate
Table A: Salary Assumptions
Pages to be formatted
Number of source files
Table B: Sample Project Scope
* Leveraged words are those within phrases which have either been previously translated or are repeated throughout the text.
Because very few translators work directly in Framemaker, source files need to be processed in order to allow the translators to work with RTF files, allowing them to work with Trados Workbench. Assuming that no problems are encountered, this phase of the project will take approximately 15 minutes at a cost of $9.24 ($36.95 x .25) based on a productivity rate of approximately 50 files per hour.
Translation and Editing
For the translation/editing phase of the project and assuming a daily productivity rate of 2,500 unique words per day and 3,800 leveraged words per day, we can estimate the time and cost as follows:
78,000 / 2,500 = 31.2 days x 8 = 249.6 hours x $31.85 = $7,949.76
18,000 / 3,800 = 4.74 days x 8 = 37.92 hours x $31.85 = $1,207.75
Once the translation and editing have been completed, the translated content needs to be proofread. In our example, this work needs to be done by the ICR, and we use a standard proofreading rate of 5,000 words per day:
96,000 / 5,000 = 19.2 days x 8 = 153.4 hours x $31.85 = $4,892.16
The translated files need to be converted back to Framemaker with Trados S-Tagger and then formatted. Based on a rate of formatting 25 pages per hour:
590 / 200 = 2.95 days x 8 = 23.6 hours x $36.95 = $872.02
Once the files have been reformatted in Framemaker in order to accommodate language expansion/contraction, the translator needs to review the document to verify that the layout of the translated document is identical to that of the source language document, words that were hyphenated follow the target language's hyphenation rules, indexes and tables of contents are verified, and any miscellaneous text (call-outs for graphics, headers, footers, etc.) is properly translated. Changes are made via hardcopy or as comments in the PDF file, and are then passed back to the desktop publisher:
590 / 216 pp/day = 2.73 days x 8 = 21.84 hours x $31.85 = $695.60
Any corrections or suggestions made by the translator are made to the translated Framemaker documents, indexes and tables of contents are regenerated, as are final PDFs. Historically, this process takes 15% of the time of the initial DTP effort.
23.6 hours x 15% = 3.54 hours x $36.95 = $130.80
Time needs to be estimated for the management of the project. For our sample single-language project, Project Management involvement wouldn't be nearly as much as say for a multiple-language software localization project.
490 project hours x 5.275% = 25.85 hours x $63.70 = $1,646.65
If we breakdown the accumulated costs for this project, we would have:
Translation, editing and proofreading:
61 days (12 weeks)
If we break down the costs to unit costs, much as a professional agency would do, we would see that for this job:
the per word cost of translation is 96,000 words / $14,049.67 =
the per page cost of desktop publishing is $1,707.66 / 590 pages =
project management =
Obviously the cost for this same project would vary from organization to organization depending on internal costs, productivity rates, etc. However, our sample gives us a fairly good idea of how long a project of this scope should take and approximately what the costs and production schedules would be.
While the project cost may be acceptable, the delivery time may not be, especially if it is a critical document that a client may be waiting for, or that is holding up an international product release. Although not possible to quantify in general terms, there may be a significant cost to the company of not producing the product in a shorter period of time. In the companies I previously worked for, it was not uncommon for sales contracts to call for delivery of translated versions by a specified date. Severe late penalties were incorporated into contracts and countless thousands of dollars in late penalties were paid time and again because deliveries were not be made when promised. Of course, the sales person never bothered to consult with the Translation Department Manager prior to making the commitment, but that's a subject we can leave for another article!
Let's Send It Overseas, It Won't Cost Us Anything!
An assumption that I made in estimating product delivery for our sample project concerning the ICR merits mentioning at this point: most of the companies I have worked with in the past and continue to work with in my present position do not have dedicated ICR staff just sitting around waiting for work to be handed to them from the Translation Department. ICRs that I have worked with have included sales and pre-sales staff, helpline support representatives, development staff, administrative personnel and in one instance, the Country Manager of one of our overseas offices (and his rate was a lot higher than $31.85 per hour).
Unless they are a dedicated resource for the Translation Department, using in-country staff poses issues that influence both cost and delivery.
First of all, these projects come in addition to their primary job tasks, thus adding to their responsibilities (without any additional pay). Next, this extracurricular translation work may very well not be a priority item on their agenda - their initial job responsibilities come first. This, of course, makes it close to impossible to design a project plan with firm delivery dates, as the time that is spent on translation is usually whatever time remains after having executed their primary responsibilities. And with pre-sales, sales, and support staff, this "available time" is extremely unpredictable.
Other than being native speakers of the target language, many in-country resources are not qualified to be critiquing translation work. Most are not trained as translators, and very often not fluent enough in English to evaluate what an accurate rendering of the source content is. Also, do not assume that because someone is fluent in a foreign language that they have good writing skills any more than you would make the same assumption for a fluent English speaker. Their writing skills, even in their own language, may be inadequate.
In my experience working with in-country staff, the final product is rarely consistent: whoever had the extra time was assigned to do the review work, resulting in several different people working on the translated content. And very often, the target language text was rewritten to accommodate the local country requirements, whether or not this represented an accurate translation of the source content.
Other intangible consequences of using non-dedicated in-country staff for translation tasks add time and cost. Projects have to be centrally managed by the Translation Manager, and it is not an easy task to manage resources who usually sit half a world away and who report to someone else. And while the Translation Department may implement translation tools and specific platforms for executing their work, the overseas offices usually don't invest in these tools, since translation is not their primary role. The centralized translation team frequently absorbs additional work, time, and cost in order to account for the discrepancy between tools and platforms.
Finally, consideration should be given to the hidden costs involved in having sales and support staff doing translation work. Was a potential sale lost because the appropriate resource was not available for a key pre-sales presentation? Was an existing client lost because they did not receive timely support concerning a critical product issue? It may seem that these situations are too unlikely to merit serious consideration. However, both have occurred during my tenure with previous employers.
The global reality today is such that most organizations that wish to succeed need to be active in the global marketplace. And to succeed in that space, they face the challenges of making their products available in the language of their target market, and doing so in the most cost-effective and timely manner.
Organizations that have developed dedicated in-house translation capabilities should continuously revisit the costs associated with this approach and analyze the true costs inherent in their methodology. What they find may validate their original thinking, or may prompt them to modify their operational model in order to maximize the cost and delivery benefits to their organization.
Advances in translation automation technology in particular and IT in general, the Internet, along with the knowledge and experience that organizations have accumulated in the recent past in developing multilingual products give them the opportunity to:
- work cost-effectively
- launch and support products more rapidly
- increase market share
- maximize revenues and profitability
- maintain a satisfied client base
Those who understand, monitor, and evolve their in-house model will strengthen their position in the international market. In today's global economy, this is no longer a luxury, but rather essential for their survival.
Interpro Translation Solutions
Interpro Translation Solutions provides world-class translation, localization, desktop publishing and project management solutions which enable clients to deliver multilingual products to their global audiences.
Our mission is to work in partnership with our clients in order to communicate successful ideas globally, to deliver confidence by understanding and addressing their foreign language translation needs, and to provide a personal commitment to delivering uncompromising quality and service.
Ralph Michael Strozza
Ralph Strozza is the President and Chief Executive Officer of Interpro Translation Solutions Inc., which he co-founded in 1995.
Ralph has worked in the translation and localization business for over 20 years in Europe and North America. He established and managed in-house translation departments for TAO International in Toulouse, France, Intergraph Corporation (Huntsville, Alabama), and System Software Associates (Chicago, Illinois) prior to co-founding Interpro.
Ralph received Bachelor of Science degrees in International Marketing and French, with a minor in Spanish from Northern Illinois University in 1981. He received his Master's Degree in French and Italian from Northwestern University in 1983. He is a native speaker of English, and is fluent in French, Italian, and Spanish.