According to CX experts, one of the most effective ways for companies to create profitable, sustainable relationships among multilingual and international buyers is to provide localized customer experiences.
At least 75% of global customers prefer to do business in their native languages, and are far less inclined to make purchases on sites they cannot read. More importantly, buyers won't become satisfied customers-much less long-term ones-if they aren't supported, post-purchase, with a localized customer journey.
Translating your organization’s secure customer account-management portal—and other customer-facing portal content such as support assets—is a critical element to successfully:
- Generate customer satisfaction among multilingual customers
- Improve customer retention
- Increase conversions for upsell and cross-sell opportunities
- Proactively communicating with customers
However, how well you succeed largely depends on what kind of translation approach you take with your portal. There are two primary approaches: translations provided by humans, or with software (aka machine translation).
Both have advantages and disadvantages, which can impact factors such as accuracy, conversion rates and costs. 让我们细细研究一番。
机器翻译
Machine translation solutions from companies like Google and Bing have come a long way in the past five years. And breakthroughs in Neural Machine Translation (NMT) software have made machine translation an increasingly viable option for some businesses. This is because:
- NMT models process information similarly to the human brain
- This enables them to essentially “learn” new languages
- NMT software translates content at the sentence level (rather than at a less nuanced word level), delivering more accurate translations
Benefits of Machine Translation
Machine translation offers several advantages, including:
- More affordable than human translation
- Hundreds of webpages can be translated almost instantly
- Using newer NMT solutions, translation quality can improve over time by “teaching” the software with additional content
Challenges of Machine Translation
然而，您会发现，机器翻译与人们在彼此交谈时使用的语言仍然存在较大差距，这也是其劣势所在：
- Provides comparatively simplistic word choice and fluency
- 软件通常无法区分区域性或地方性语言
- 发布之前通常不会有编辑监督
- Hidden costs for reviewing and revising subpar translations
- Improving the accuracy of a NTM solution requires a time- and energy-consuming “teaching” process
人工翻译
Great vendors provide professional linguists who are fluent in their customers’ industries, brands, languages and cultures of international markets. Studies indicate time and again that human translation generates superior engagement, conversions and credibility among multilingual customers.
Benefits of Human Translation
- 一流的翻译准确性
- 内容可根据区域性方言与短语进行轻松定制
- Brand-aligned portal translation that captures your company’s messaging and voice
- 质量通过人工编辑监督与多级审核流程得到保证
Challenges of Human Translation
- 必要的语言、创意和编辑专业经验会造成成本高于机器翻译
- Requires more time to translate text (though great vendors use optimized workflows to reduce turnaround times to 1 business day)
组合方案
Some vendors can combine human and machine translation approaches to mitigate localization costs. 理想的混合翻译法会为神韵和讯息至关重要的高流量页面采用人工翻译。机器翻译可用于品牌敏感性较低的内容，例如措辞选择不太重要的内容。
Benefits of the Hybrid Option
- 人工译员可对具有品牌敏感性和高可见性的内容进行本地化
- 高流量页面上由人工译员准确一致地传达品牌信息
- 比纯人工翻译节约成本的理想选择
Challenges of the Hybrid Option
- Translation quality may vary from human and machine sources, but delivers dramatically higher accuracy than machine translation only
- Human-translated sections are reviewed for accuracy, but machine translations may not
结论
As you explore portal translation solutions, look for a partner that can help you achieve success through a translation process that is, ideally, invisible to the reader.
这意味着必须找到与您的需求同样灵活敏锐的技术和一站式解决方案。
These solutions can provide world-class human and machine translation options for your multilingual portal, and accommodate unique requests—such as using the hybrid approach on specific webpages, or within specific sections of webpages—to deliver superior results while controlling costs.
最后更新日期：2020年3月25日
