According to a recent Nimdzi survey, on-time delivery (84% of respondents) and quality of translation (83%) are the decisive elements when shopping for a translation/localization partner.
It comes as no surprise that quality, in all its aspects, counts towards a company’s success, because growth and expansion of a business are much easier when core products or services remain consistently good.
However, we live in a flawed world, where humans make mistakes and machines break when you least expect it. The secret to delivering constant quality lies in preventing errors as much as possible, avoiding rework and, ultimately, saving time and money.
But, as an LSP or a manager of an in-house localization department, how will you translate this “secret” into everyday practice? In this article, we summarize 5 strategies to help you keep up with the basic principles of quality management and hit the ground running toward high performance.
#1 Consider the source
The quality of a translation (whether done by humans or by machines) is intrinsically linked to the quality of the source content. So, design your content keeping in mind the three pillars of the authoring process: clarity, consistency and conciseness.
Prepare authoring guidelines with information about the target groups and examples of your corporate voice and tone.
Create templates for structured content, for example, technical and educational documents. A well-designed template that is also easy to read will make your content future-proof.
For more guidance about creating and preparing source files, take a look at our previous articles:
Setting up your source files for translation
Writing machine-translatable texts
#2 Maintain your language data spick-and-span
You can also achieve a certain consistency in your company’s communication material by using translation memories (TMs) and terminology databases.Consider using two kinds of TMs:
Project memories that are automatically created and filled as translators type within the various segments.
A master memory that contains all key data regarding a domain, a client and the pre-filtered, and relevant content from the project memories.
Terminology databases will also benefit a wide group made up of:
Translators, who will be able to translate more accurately.
Clients, who save time and money because no rework is needed and the documentation will always be up-to-date and accurate, with less risk of liability issues.
Consumers, who are able to understand the documentation/communication without confusion. This, in turn, will result in less complaints or support requests.
In Wordbee, all translations are saved to your project memory and consolidation into the master memory can be automated. You can easily view your TMs as well as terminology databases in the resources section:
To dig deeper into language data, have a look at our previous articles:
Three Ms: Master Memory Maintenance
Strategies To KonMari Your Translation Memories
Terminology Management: Getting Started
#3 To each content typology its own workflow
Blog posts, infographics, white papers, product specifications, e-mails, e-books… The list of content typologies is endless. You can find many of them within one single project, each requiring different localization criteria and, therefore, a customized workflow. For example, a minimalist stream for perishable content (e.g. machine translation and one round of light post-editing), a more traditional TEP (translation/editing/proofreading) workflow for content with a longer shelf life, or a more detailed workflow with back translation & reconciliation for medical texts.
In Wordbee, project managers can customize workflow templates to accommodate specific needs and manage complex projects. Here is an example of what such a custom workflow could look like.
#4 Automate QA Checks
Within a project’s workflow you have the option to automatically run a Quality Assurance (QA) check after the completion of a job. QA profiles with certain rules can be saved in advance and selected to be used to run the QA check in order to guarantee that the right set of rules is followed, based on the client’s requirements and expected quality level.
Another option is to have your vendors perform a mandatory QA check before delivery to fix nuisance errors, like dates and numbers, formatting tags, forbidden terms, spaces, punctuation, etc.
In Wordbee, you can choose between mandatory & automated QA checks in the workflow settings:
In addition to automated and mandatory QA checks, a real-time Quick QA feature as well as the Spellchecker help translators avoid errors while they are translating or proofreading within Wordbee.
Here are more resources about Quality Assurance and Quality Control:
A Holistic Approach To Quality Management
Optimize your LQA with Automatic QA Checks In Wordbee
#5 Measure, analyze, improve, repeat
This is the motto at the base of every total quality management (TQM) methodology.
If chosen wisely, Key Performance Indicators (KPIs) can shed a light on your vendor and project management performance. For example, you could create reports to:
measure delivery, rework, or full cycle time in order to improve process efficiency;
vet, select and rate vendors to find the perfect fit for each project;
monitor capacity utilization rate, latency time and uptime rate to keep track of your technological investment.
This is what the vendor rating statistics could look like in Wordbee:
More information on Business Analytics and KPIs in Wordbee:
Transforming Translation Data into Business Analytics
KPIs for the Translation Business
根据Nimdzi最近的一项调查,准时交付(84%的受访者)和翻译质量(83%)是选购翻译或本地化合作伙伴时的决定性因素。
这一点儿也不出人意料,从各个层面而言,质量都是公司成功的关键——因为当核心产品或服务始终保持良好时,业务的增长和扩张就容易多了。
然而,我们生活的世界并不完美。人类会犯错误,机器会猝不及防地掉链子。提供恒定质量的秘诀在于尽可能地避免错误,避免返工,才能最终节省时间与金钱。
但是,作为语言服务提供商或全职本地化部门经理,你将如何将这个“秘密”转化为日常实践呢?在这篇文章中,我们总结了5个策略,帮助你跟上质量管理的基本原则,并切切实实地奔向高绩效。
#1考虑来源
翻译的质量(不管是人工翻译还是机器翻译)与源内容的质量有着内在的联系。因此,在设计内容时要牢记创作过程的三大支柱:清晰、一致和简洁。
准备创作指南,其中包含有关目标群体的信息以及企业意向与态度示例。
为结构化内容(例如,技术和教育文档)创建模板。一个设计良好且易于阅读的模板能让你的内容不过时。
有关创建和准备源文件的更多指导,请参阅我们以前的文章:
设置用于翻译的源文件
编写机器可译文本
#2快速维护语言数据
通过使用翻译记忆和术语数据库,你也可以在公司的沟通材料中达到一定的一致性。你可要考虑使用两种翻译记忆:
项目记忆,在不同的段自动创建和填充作为翻译类型。
主翻译记忆,包含关于域、客户端和预过滤的所有关键数据,以及来自项目记忆的相关内容。
术语数据库还将使由下列人员组成的广泛群体受益:
译者将能够更准确地翻译。
客户能节省时间和金钱——因为无需返工,文件将始终是最新的和准确的,责任问题的风险较小。
消费者能够理解文档或沟通,而不会混淆。这反过来会减少投诉或支持请求。
在Wordbee中,所有的翻译是可以自动化保存到你的项目记忆中,并且整合到主记忆中。你可以在参考资料部分轻松地查看TMs和术语数据库:
想要更深地挖掘语言数据,看看我们之前的文章吧:
三个MS:主记忆维护
KonMari翻译记忆策略
术语管理:入门
#3每个内容类型都有自己的工作流
博客文章、信息图表、白皮书、产品说明书、电子邮件、电子书……内容类型的列表是无穷无尽的。你可以在一个项目中找到许多这样的工具,每个都需要不同的本地化标准,因此需要一个定制的工作流。例如,对易逝内容的极简流程(例如,机器翻译和一轮轻量级的后期编辑),对内容的更传统的TEP(翻译/编辑/校对)工作流,以及对医学文本的更详细的反向翻译和校正工作流。
在Wordbee中,项目经理可以自定义工作流模板,以适应特定的需求和管理复杂的项目。下面是这样一个自定义工作流的示例。
#4自动化质量保证检查
在项目的工作流中,你可以选择在任务完成后自动运行质量保证(QA)检查。具有某些规则的QA配置文件可以预先保存,并选择用于运行QA检查,以保证根据客户的要求和预期质量水平遵循正确的规则。
另一种选择是让你的供应商在交付前进行强制性的QA检查,以修复一些令人讨厌的错误,比如日期和数字、格式标签、禁止使用的术语、空格、标点符号等。
在Wordbee中,你可以在工作流设置中选择强制和自动QA检查:
除了自动和强制的QA检查之外,实时快速QA功能以及拼写检查器可以帮助翻译人员在WordBee中翻译或校对时避免错误。
以下是更多关于质量保证和质量控制的资源:
质量管理的整体方法
在Wordbee中使用自动QA检查优化语言质量保证
#5测量、分析、改进、重复
这是每一种全面质量管理(TQM)方法学基础上的座右铭。
如果选择得当,关键绩效指标(Key Performance Indicators,KPI)可以帮助你了解供应商和项目管理的绩效。例如,你可以为以下对象创建报告:
测量交付、返工或全周期时间,以提高流程效率;
对供应商进行审核、选择和评级,以找到适合每个项目的最佳供应商;
监控容量利用率、延迟时间和正常运行时间率,以跟踪你的技术投资。
以下是Wordbee中的供应商评级统计信息:
有关Wordbee中的业务分析和KPI的更多信息:
将翻译数据转换为业务分析
翻译业务的KPI
以上中文文本为机器翻译,存在不同程度偏差和错误,请理解并参考英文原文阅读。
阅读原文