Web Localization for Marketers: Automatic vs Machine vs People

Cross border e-commerce is booming. Even back in 2016 a global Nielsen report suggested that 57% of shoppers had purchased from an overseas retailer in the previous 6 months. In recent months the global COVID-19 has had a huge impact on retail across the globe. Bricks and mortar shopping has dropped significantly in the US and UK, with the decline of the total retail market in the US this year is expected to be double that experienced in the 2009 financial crisis. At the same time, we have seen a huge surge in cross-border e-commerce.

Now, RetailX is suggesting that cross-border e-commerce in the EU grew by 30% this year. In the US, data from Global-e found that international commerce had grown 42% by May this year.

Wherever your retail brand is based international sales could be a lifeline. It’s no surprise that marketers worldwide are looking to capture this ever-growing segment of new business. However, to effectively capture cross-border consumers marketers need to go beyond merely providing site translation once a visitor lands on their site. They need to look at localization. This means taking into account elements like native language SEO, providing images that are appropriate for a local market – if you’re a European retailer trying to sell to the Asian market, exclusively using euro-centric images on your site is going to disengage your potential customer. Web localization is making sure your site is taking into account all the cultural nuances of the regions you’re trying to sell in.

This can seem like an impossible task. Many retail sites have hundreds of regularly updated pages and employing professional translators would be prohibitively expensive.

At the same, time many might consider machine translation and localization to be sketchy and too inaccurate to rely upon. But as anyone that uses machine translation software knows, technology is improving all the time. Technology can be an incredibly valuable tool for web localization, and when partnered with real people, it can reach dizzying heights.

Automatic vs Machine

A common misconception is that automatic translation is the same thing as machine translation. According to the Globalization and Localization Authority (GALA), machine translation, by definition, refers to “fully automated software that can translate source content into target languages”. So, machine translation technologies include providers like Google and Yandex Translate, Microsoft Translator, DeepL etc. But these machine translation providers applied to a website will usually only overlay native languages once the visitor is on the site.

Translation solutions take a different approach to using machine translation. Automatic translation encompasses this process of translation but goes beyond that. Using a translation solution not only deals with the translation of your content but also the managing and editing of the content, the SEO of every translated page and then handles the publishing of that content automatically, potentially live without you having to lift a finger. For some retailers, the output from this application of technology will be enough to start boosting international sales and will be incredibly cost-effective.

People vs the Machine

Of course, one of the main drawbacks of using machine translation in localization is accuracy. Many marketers will feel that full human translation is the only reliable way forward. Of course, the cost of that is huge and prohibitive for many retailers – not to mention it doesn’t take care of how that translated content will actually be displayed.

Machine translation can save you a lot of time. Of course, the accuracy is dependent on the language pair chosen and how developed and proficient that translation tools are for that specific pair. But say, as a ballpark estimate that the translation is good 80% of the time, all you need to do is get a professional translator to verify and edit the translations accordingly. By getting the first layer of machine translation you’re accelerating the process towards making your website multilingual.

Furthermore, from a financial perspective, this choice is a huge consideration to make. If you’re hiring a professional translator to start from scratch and work on copious amounts of web pages, the bill you’ll rack up will likely be astronomical. Whereas if you start with a first layer of machine translation and then bring in human translators to make adjustments where necessary (or maybe your team speaks multiple languages) both their workload and your bill will be significantly reduced.

Website localization can seem like a daunting project, but handled correctly with a combination of technology and people power it’s not as big a job as you think. But cross-border e-commerce needs to be a strategy for marketers moving forward. In 2016 Nielsen was reporting that 70% of retailers that had forayed into cross-border e-commerce had been profitable with their efforts. So, any foray into localization should be profitable if done effectively with technology and the limits of technology in mind.