Translation Workflow Reference Tables: Setting Job Expectations the Easy Way

Easy-to-use tables make setting expectations for translation, revision, and edit validation a breeze!

Setting expectations before starting translation-related tasks is a critical step in the project lifecycle. By clarifying work requirements in advance, linguists and other language services providers can set reasonable rates, workflows, and schedules that deliver exactly what clients need and want. However, defining work scope on complex tasks sometimes involves lengthy email exchanges and multipage instructions. This process can be time consuming and prone to misunderstandings. Fortunately, even projects that appear complicated can be simplified with the right tools.

I would like to share three tables¹ I often send to clients before confirming, quoting, or starting projects. These tables cover three major tasks of my everyday work (translation, revision, and edit validation, or proofing review) and serve as a point of reference in the expectation-setting process. When I provide pre-set workflow options (referred to in this article as “approaches”) for my client to choose from or customize to their needs, we can quickly reach a common understanding on the work to be done.

Each table presented here is specific to an individual type of task. For example, if I’m asked to revise another translator’s work, I email the project manager a screenshot of the Revision Approaches table (Table 2) and ask which of the six approaches listed in the table best describes the workflow I should follow. The project manager then identifies the approach that matches the project requirements, or else explains how my work needs to vary from one of the listed approaches. Of course, I could also email a Word or Excel file containing the table (or even provide a link to a Dropbox or Google Drive downloadable file) and let the project manager type his or her requirements into the file and email it back to me with instructions. Regardless of the format, once I receive the client’s feedback and understand the scope of the work involved, we can then agree on a price and schedule.

It’s important to point out that the tables are convenient tools for reaching an agreement before a project has begun, but are not intended for ongoing discussion once the work is underway. There are several reasons for this.

1. I have to know what the client expects before I can set my price or decide if I can meet the budget or schedule.
2. Changing the work scope after the job has begun (or worse, after it’s finished) so that I have to rework previously completed sections results in wasted time, money, and effort.
3. I find that having client expectations in mind throughout my work helps me maintain consistency, efficiency, and quality.
4. Having a written record of the project scope provides an objective basis on which the client can compare my delivery with the specifications we agreed to at the beginning. This helps avoid conflicting opinions about the outcome.

Once we’ve reached an agreement on project terms, the client may include the applicable table with other instructions in a purchase order or even in a more formal project contract, which I can then sign or accept electronically. Of course, a formal contract is a best practice. But if a job is small enough, I might even take the client’s emailed selection of an approach and my emailed acceptance of the instructions as the basis for defining project scope. Thus, these tables can be flexibly incorporated into the workflow in whatever way is best suited to the circumstances.

Each of the three tables introduced in this article uses a two-column structure. The left column lists the names I’ve given to several pre-defined approaches for the respective task and the right column provides an additional description of what each approach covers. The Translation Approaches options (Table 1 below) addresses the varying degrees to which a translation might adhere to the source. The Revision Approaches options (Table 2) stipulate the types and extent of edits made to a translation during the revision process. For the Edit Validation Approaches (Table 3), each choice relates to the criteria used when deciding whether or not to accept reviewer edits to a previously translated document.

The methods I suggest in these tables correspond to how I work, but they aren’t unalterable. If none of the given approaches is a perfect fit for the project requirements, a client can pick the one that’s closest to what they have in mind and then specify what they want changed. This still takes a lot less effort than explaining everything from scratch.

Also, even though this article describes a process where I (a freelancer) clarify the scope of the project and the amount of work desired by emailing a table to a client (in practice, usually a project manager at a translation agency), the tables can be used in other arrangements as well. For example, a project manager assigning a job to a freelance translator might include the appropriate table in the initial project instructions and designate which approach the translator should take. Agencies can also send tables to their direct clients, as appropriate, when defining the scope of the project at the originating project level.

Choosing a Translation Approach

Table 1 below provides a framework for choosing a translation approach and discussing adjustments to that approach depending on the client’s specific needs. In this table, the approaches are listed according to how closely the translation matches the source text in terms of meaning, terminology, and how the client wants the final text to read. Approach #1 at the top (Strict and Literal Translation) requires that the translator match the translation to the source as closely as possible. Moving down the table, each subsequent approach targets an outcome that is less tied to the source and gives the linguist more discretion in composing the translation.

Table 1: Translation Approaches—Degree of Adherence to the Original

Here are a few examples of how this works. A client requesting a back-translation to check a forward translation would likely select Approach #1 (Strict and Literal Translation).³ In litigation settings and when translating official certificates and user manuals, having a translation that is faithful, complete, and precise is a top priority, but a natural writing style is also important. In this case, Approach #2 (Faithful but Natural Translation) would be the approach to communicate correctly what the source says, but not in such a literal way that the text sounds stilted or loses much of the nuance of the source. On the other hand, for materials that are meant to inform in a general way (such as survey responses) or that will be published for the general public (such as newsletters and brochures), I’ve found that clients prefer a final output that doesn’t sound like a translation. In this case, they might opt for Approach #3 (Translation that Reads as Naturally as the Source).

Of course, each client and project is different, which is why it’s important to ask about expectations before starting to ensure everyone is on the same page.

Revising/Editing and Proofreading

Revising/editing can be painstaking work, especially when the goal is to deliver a perfect final product. The effort increases even more when the original translation isn’t as good as it should be. However, if the client doesn’t value stylistic improvements as much as just getting the text free of objectionable errors, then the effort and budget may be adjusted accordingly. With the exception of Approach #6 (Essential Edit), which only focuses on mistranslations, the approaches in Table 2 below are listed in order from least to most effort required to improve a translation that has already been completed.

Table 2: Revision Approaches—Types and Extent of Edits Made to a Translation

By clearly defining each term and workflow, Table 2 is also helpful for achieving mutual understanding when individual professionals use different words for the same task (e.g., proofreading versus editing versus revising). I consider any workflow that involves comparing the source against the target a revising/editing job, and I don’t make a distinction between revising and editing. On the other hand, proofreading refers to a unilingual read-through. Notably, neither of the two proofreading approaches listed in Table 2 assumes a source-to-target comparison unless the proofreader finds something specific to check, such as unclear or apparently incorrect wording. On the other hand, depending on the quality (or lack thereof) of the original translation, Approach #5 (Intense Revision/Editing) could take just as long (or even longer) than retranslating from scratch. Also, proofreading and revising/editing don’t entail a significant formatting or terminology effort. If formatting or terminology work is needed, this should be billed separately or even handled by someone else.

I don’t try to persuade clients that one approach is better than another, but clients do need to be aware of the price/quality tradeoff and what degree of improvement they can expect. Since the approaches require different investments of time and effort, the rates charged for each should also vary. Unfortunately, even if the client communicates their expectations about the final product clearly, this still doesn’t provide enough information for a linguist to precisely assess in advance the effort required for a revising/editing or proofreading job. This is because the effort to reach a predefined final output quality also depends on the quality of the original translation.

Ideally, the cost for revising/editing and proofreading should be billed in two parts. The first component is a fixed per-word rate. Then a charge should be added based on a metric, such as edit distance, which provides a quantitative measure of the number of edits made to a document.⁴ Thus, the variable component of the charge rises as the extent of the edits increases. This formula captures the basic effort needed to understand the text and the additional work to make changes. It also aligns the client’s interest (value for cost) with the linguist’s interest (compensation for time and effort). Changes can then be compared against the pre-agreed approach to validate that the work was performed to specification. Unlike hourly billing, this hybrid rate structure provides an objective basis on which to evaluate the value delivered without bringing up trust-killing questions such as how long the work took or should have taken. This approach also provides an incentive for the linguist to invest in tools and skills that improve his or her quality and efficiency.

Validation of Reviewer Edits

Client-side reviews (sometimes also called “in-country reviews”) and other independent proofreading and edit workflows of translations delivered can result in anything from a few light edits to a complete rewrite. Additionally, what the agency client wants from the linguist when validating those reviews also varies on a case-by-case basis, and what the end client wants on a validation doesn’t always align with best practices. Sending the information in Table 3 to an agency client for guidance allows the linguist to set a budget and determine an outcome that will ultimately achieve the objectives of the task.

Table 3: Edit Validation Approaches—Criteria for Deciding Whether to Accept Reviewer Edits to a Previously Translated Document

For example, Approach #1 (Assess-and-Defer) might be appropriate when no budget is available for the review and the translator is just being asked for a general opinion on changes. On the other hand, end clients who send back edits to be finalized usually want their changes incorporated into the final version as long as the edits don’t introduce errors or otherwise make the translation worse. In such cases, one of the two Reviewer Deferred approaches would be appropriate since these options respect the reviewer’s work but also ensure the reviewed version isn’t worse than the original translation. When agency clients have a translation independently edited before delivery to an end client and ask the translator to review the edits and finalize, Approach #3 (Even-Handed) is usually preferred. Occasionally, even an extreme approach such as a Defensive review (#5) might be in order.

Clarify Expectations Now to Save Time Later

The tables in this article are designed for reference in specific workflows that I follow daily. I’ve found them useful for setting and meeting expectations with clients on specific tasks, and I encourage you to use them in your work. However, there seem to be as many workflows as there are linguists and language providers, and it’s likely you’ll want to adapt these to your own unique situation, and even create table for other tasks. Fortunately, this concept of standardizing common routines into clearly described options can be applied to a wide range of work circumstances. I designed these tables to avoid composing a new email to clarify details for each and every job. However, before creating the tables, I had to think carefully about the available approaches and how to describe them in ways that would be easy to understand at the beginning of a project and to compare against performance at the end of a project. The key requirements for a useful table are: 1) the table must apply to a frequent type of task; and 2) the approaches must be clearly and objectively described. Don’t hesitate to contact me with questions or feedback. I’ll be particularly interested to hear how you have adapted these tables to your own unique situations and workflows.

Notes

1. Visit http://xl8.link/tables to download these tables in PDF format for easy use in your projects.
2. Transcreation refers to an approach to translation used commonly in marketing and literary work that emphasizes the cultural and linguistic context of the target language. Thus, in transcreation, the translator may adapt a message to a target audience with translated elements that aren’t necessarily included in the source text. For more information on transcreation, refer to http://bit.ly/Wiki-Transcreation.
3. Clients commission a back-translation to identify possible errors in a forward translation. Then they validate the forward translation by comparing the back-translation to the source (both of which are in the source language). Therefore, the closer the back-translation matches the forward translation, the easier it is for the client to identify potential issues during this comparison review. This is the reason clients generally choose a Strict and Literal Translation approach for back translations. For more information, refer to http://bit.ly/back-trans.
4. According to Wikipedia, “In computational linguistics and computer science, edit distance is a way of quantifying how dissimilar two strings (e.g., words) are to one another by counting the minimum number of operations required to transform one string into the other.” The reason for suggesting edit distance as a factor in billing for revising/editing and proofreading work is that the number of operations required to improve a text is proportionate to the effort required to make those improvements. Therefore, edit distance is a useful way of measuring effort and value delivered. For more information on edit distance, refer to http://bit.ly/edit-distance. For a detailed description of my two-part revision model, refer to http://bit.ly/revising-post-ed.

---

Steven S. Bammel is the president of Korean Consulting & Translation Service, Inc. A Korean>English translator specializing in financial and business documents, he works with a technical team in Korea for English>Korean translation, but translates all Korean>English jobs himself. He has degrees in economics and management strategy and is currently a PhD candidate at Hanyang University in Korea. Contact: sbammel@koreanconsulting.com.