When and Why to Automate


2020-09-15 05:00 RWS Moravia Insights


阅读模式 切换至中文

Automation is on the rise. Taking the human element out of certain processes makes a lot of sense. You can increase productivity, reduce errors and have a more profitable business model. But you need to have a great deal of faith in complete end-to-end automation. Are you ready to get in your car and have it drive you to your destination while you sleep? Or would you feel more comfortable setting your speed and engaging the cruise control, but still being responsible for the steering and braking just in case? These kinds of considerations arise when automation is employed in the world of localization, too. On the surface, expanding into new territories with new languages may seem suited to automation such as machine translation to handle high volumes and speed up the work. For example, open Google Translate, choose your destination language, type in your content and it comes out localized. The results can be pretty accurate, but it’s not always this easy, and sometimes what comes out needs a lot of fine-tuning to really engage with the local culture. Why automate There are many factors that are making automation increasingly appealing: Return on investment – Because there are less variables in the process, costs can be easily mapped out and shouldn’t vary much over time. Speed – Technology can move more quickly and accurately than humans can. Productivity – Operating times can be around-the-clock, seven-days-a-week. Efficiency – Human error is greatly reduced or eliminated. Visibility – Processes are transparent and exactly the same every time. When to automate So, will automation help your business expand into new global territories? The answer varies on a business by business basis. There is no magic, one-size-fits-all answer. Keeping a human element in a process allows you to deal with nuances that inevitably appear when localizing your brand’s products or services. For example, for every opportunity where machine translation can help, there is creative content that needs someone to really understand it so its intended meaning isn’t lost in translation. Because of this, you’ll want to stay away from a ‘closed’ solution where there is only automation. What to automate There is, however, little doubt that some level of automation can be advantageous to your localization strategy. It can provide a consistency for core content that you can then adjust to individual markets. Automation can also provide a standard framework for local marketing teams to work within. Take a content management system (CMS), for example. When localizing, a strong CMS that allows your project teams to easily access and modify content is essential, and it might operate across a number of different platforms with a single interface. Often there are multiple users working on different stages of a single project, and a CMS automates the project workflow, timeline, stakeholder communication, approval processes and internal/external resourcing. When content has to be localized for a specific country, a human needs to make sure the source content is correct and that the digital elements work with each other. Then, once the variables are in place, the CMS’s automation can be triggered and operate at scale, delivering value to the business. The advantages of a translation management system (TMS) Also, the growing use of applications and mobile technology puts an increased dependence on translation management systems in order to rapidly grow an app-dependent business on a global level. A TMS helps organize the localization workflow, track the progress of translation projects and reduce manual tasks via automation. It’s important to consider each of the elements that are going to be automated. They’re not all standard and many CMSs and TMSs will not talk to each other. Once again, it is us humans that have to step in and sort out the finer details. What about quality? Quality assurance (QA) and testing are crucial to any localization process. Even the most straightforward translation process—create content, give it to a translation team, translate it—needs a QA step before it’s rolled out into the market. The QA process can vary depending on requirements. It could be as simple as a proofread by a translator with local knowledge and as complex as a functional and linguistic testing program. Even if the QA requires several steps that involve human input, automating as much as possible (such as file handoffs, quality checks and screenshot captures) will typically have time and cost savings that justify building the automations. In addition, QA should be factored into the overall process during the planning stage. This way, the decision on where and when to employ automation is factored into the project early on and its development and implementation costs can be included in the budget. This is where designing a modular approach from the start is the best route forward. A process can be broken down into components, some of which will suit total automation and some of which will need human intervention. As we move into the future, the ratio of these two elements may change, but they are both likely to be needed for the long term. What does the future hold? In the years to come, automation will evolve as different industries continue to embrace technology. Robotic processes are becoming more widely accepted and implemented in many fields, not just translation and localization. What the future holds, ultimately, is anyone’s guess, but there are some exciting developments in bringing the global marketplace together. First, the demand for real-time localization keeps growing. “Real-time” is contextual, but we understand it as meaning instantaneous. It’s the chatbot that replies to you milliseconds after you ask it a question. Or the technology that provides immediate translation, allowing us to quickly communicate in another language. This is where the market is heading, and it all relies on having the right technology and suppliers in place behind the scenes to make the magic happen. Immediately. Second, advances in digital automation are helping brands move around the world. As those brands gain success in new markets, it’s inevitable that more resources will be put into this international digital landscape. As with a number of industries, the future of localization is moving in this digital direction because people want faster and easier access to the most up-to-date information available. Building automations for your business If you’re thinking of bringing automation into your business, just make sure you ask yourself these questions: Is it the best solution? What level of automation do I need? What is needed to do the job and facilitate faster work with fewer errors and less cost? It’s rare that a process will be suited to end-to-end automation. Instead, targeting specific activities and automating them, with manual intervention where needed, is probably the best option. Inevitably, the answer is going to be different for everyone. There is no single solution here, just a process of evaluation. Steps include: Communication – Who are the process owner and project stakeholders? How is everyone going to communicate? Goal setting – Why automate? Is the goal to save time? Save money? Have better visibility? Is there another reason? Historical process and adaptation – How was the decision made to automate? Who was involved in the process? Is there any further research that needs to be done? Current workflow and data collection – Keeping things as simple as possible, what are the steps in the process, including time frames, and who is the resource at each step? Automated workflow – How is this an improvement of the current workflow? What is the cost to implement it? What test comparisons have been run? What are the limitations and how will they be addressed? Return on investment (ROI) – What has been learned from the proof of concept? How does this validate automation and show ROI? Going live – How will the project go live? How will it be monitored? What will the ongoing quality assurance process be? And if automation is the answer? Go for it, and work with an experienced partner that can help you roll it out in your current and future markets.
自动化正兴起。 从某些过程中排除人为因素非常有意义。您可以提高生产力、减少错误并拥有一个更有利可图的商业模式。但是您需要对完全的端到端自动化抱有很大的信心。准备好钻进车里,让车载带您到达目的地吗? 或者想图个舒服,设置速度和使用巡航控制系统,您仍负责方向盘和刹车,以防万一呢? 当自动化应用于本地化领域,也会出现这些考虑因素。从表面上看,使用新语言扩展到新领域似乎适合自动化,例如机器翻译,以处理高容量和加快工作速度。例如,打开谷歌翻译,选择目标语言,输入内容就会出现本地化。结果可能相当准确,但并非总是那么容易,有时结果需要大量调整才能真正融入当地文化。 为什么要进行自动化 促使自动化越来越有吸引力的几大要素: 投资回报--因为流程中变数较少,所以成本容易规划出来,并且不会随着时间的推移而变化太多。 速度--技术比人类移动得更快、更准确。 生产率----操作时间可以是每周7天不间断运行。 效率--人为错误大大减少或消除。 可视性--流程是透明的,并且每次都完全相同。 何时实现自动化 那么,自动化会帮助您的业务拓展到新的全球领域吗?答案因企业而异。没有放之四海而皆准的万能答案。通过在流程中保留人为因素可以让您处理本地化品牌产品或服务时不可避免出现的细微差别。 例如,对于机器翻译可以帮助帮助的机会,都有创造性内容需要有人去真正理解,以便其原意在翻译中不会丢失。因此,您将希望远离只有自动化的“封闭式”解决方案。 什么需要进行自动化 然而,毫无疑问的是某种程度的自动化有助于您的本地化策略。自动化提供核心内容的一致性,然后您可以根据各个市场进行调整。自动化还可以为本地营销团队提供一个标准框架。以内容管理系统(CMS)为例。在进行本地化时,使您的项目团队轻松访问和修改内容强大的内容管理系统是必不可少的,并且内容管理系统可以使用单个界面跨多个不同的平台操作。通常有多个用户在单个项目的不同阶段工作,内容管理系统可以进行自动化项目工作流程、时间表、涉众沟通、批准流程和内部/外部资源配置。 当必须将内容本地化到特定国家/地区时,人们需要确保源内容正确并且数字元素可以相互配合。 然后,一旦变量到位,触发内容管理系统的自动化并进行大规模操作,从而为业务创造价值。 翻译管理系统(TMS)的优点 同样,应用程序和移动技术的日益普及也增加了对翻译管理系统的依赖性,以便在全球范围内快速发展与应用程序相关的业务。 翻译管理系统帮助组织本地化工作流程,跟踪翻译项目的进度并通过自动化减少人工任务。 重要的是考虑即将进行自动化的每个元素。 这些元素并不都是标准的,许多内容管理系统和翻译管理系统互不通信。我们人类必须再一次介入并理清更精细的细节。 关于质量 质量保证(QA)和测试对于任何本地化过程都是至关重要的。即使是最简单的翻译过程--创建内容、将其交给翻译团队、进行翻译--也需要进行质量检查,然后才能推广到市场。质量保证可能会根据要求而有所不同。这一过程可以简单到由具有本地知识的翻译人员校对,也可以复杂到功能和语言测试程序。即使质量保证需要几个涉及人工输入的步骤,但尽可能地自动化(例如文件切换、质量检查和截图捕获)通常会节省时间和成本,从而证明了构建自动化的合理性。 此外,质量保证应在计划阶段纳入整个流程。通过这种方式,就可以在项目早期就决定何时何地采用自动化的决定,并且可以将其开发和实施成本纳入预算。 从一开始就设计模块化方法是最佳途径。可以将一个流程分解为多个组件,其中一些将适合完全自动化,而另一些则需要人工干预。 随着我们迈向未来,这两种元素的比例可能会发生变化,但从长远来看,我们都需要这两种元素。 未来该何去何从? 在未来的几年里,自动化将随着不同行业继续采用技术而发展。机器人过程不仅在翻译和本地化方面,而且在许多领域也越来越广泛地被接受和实现。 未来该何去何从,归根结底谁也说不准,但是在将全球市场整合到一起方面,存在一些振奋人心的发展。 第一,实时本地化需求不断增长。“实时”与上下文相互关联,但我们将其理解为瞬时的。 在您提出问题后,聊天机器人会在几毫秒内回复您。 或者提供即时翻译的技术,使得我们快速地用另一种语言交流。这就是市场的发展方向,而这一切都取决于幕后有合适的技术和供应商来实现。立即使用本地化。 其次,数字自动化的进步正在帮助品牌走向世界。随着这些品牌在新市场取得成功,不可避免的是将更多的资源投入到国际数字领域。正如许多行业一样,本地化的未来正朝着数字化方向发展,因为人们希望更快更容易地获取最新的可用信息。 为您的企业构建自动化 如果您想将自动化引入您的业务,请确保您问自己以下问题: 自动化是最佳解决方案吗? 我需要什么水平的自动化? 需要什么来完成这项工作并促进更快的工作、更少的错误和更少的成本? 很少有流程适合于端到端自动化。相反,针对具体活动并将其自动化,在必要时进行人工干预,可能是最好的选择。 不可避免地,每个人的答案有所不同。这里没有单一的解决方案,只是一个评估过程。步骤包括: 沟通–谁是流程所有者和项目利益相关者? 大家如何沟通? 目标设定--为什么要自动化?目标是节省时间吗?减少成本吗?可视性更好?还有其他因素吗? 历史进程和适应性--如何做出自动化决策? 谁参与了该流程? 是否需要做进一步的研究? 当前的工作流程和数据收集--使事情尽可能简单,流程中的哪些步骤(包括时间表)以及每个步骤的资源是什么? 自动化工作流程--这是对当前工作流的改进吗?实施自动化工作流程的成本是多少?进行了哪些测试比较?有哪些局限性,将如何解决? 投资回报率(ROI)--从概念验证中学到了什么?这如何验证自动化并显示投资回报率? 上线-项目如何上线?将如何监控?正在进行的质量保证流程是什么? 如果自动化是答案呢?坚持下去,并与经验丰富的合作伙伴合作,可以帮助您在当前和未来的市场中推广自动化。