Many promises revolve around artificial intelligence, including those that already thank solutions like ChatGTP for their early retirement in the concept of basic income. But be careful, because leaving our company to develop tools too quickly can be counterproductive.
In the field of e-commerce translation and localization, Large Scale Language Models (LLMs) like ChatGPT have their limitations and advantages. Although technology has made significant advances, it is still important to understand when and how these models are implemented into a professional workflow.
Out of Innova Digital Export (Dori Lopez) with the invaluable help of the subject matter expert, Oscar Nogueras (translation)We tell you how YES and why NOT to use ChatGPT in your international ecommerce in three main keys:
- Use ChatGPT in your research processes and Minimum Viable Products. Be careful when translating content, always under supervision.
We can lean on that GhatGPT to help us in two parts of the process, either one first machine translation from which to start, which is checked by a human, or in a final revision of the text. The reason why we don’t delegate the translation of texts that might impact the conversion of our business? Its use can pose risks, such as pure errors that cause real headaches for brands (see the end of this article), political incorrectness or sensitivity issues to the target market, vagueness in the data or even problems arising from their privacy.
at the time of investigate Regulations, catalog analysis or in short, immersion in any type of long and unobtrusive documentation, ChatGPT can be a very useful ally, saving us time in the process, offering us a flat and quickly translated viewTo.
Another app? If you’re leaning towards agile methods and are testing a new product in a specific market that you haven’t invested too many resources in yet, ChatGPT can help you with that Start a fast pilot.
- Do not use ChatGPT in your translation processes
In total, Use AI on translations that aren’t an important variable in your consumer’s purchasing decision. If so, it is still advisable to rely on meat staff.
All translation services such as product description, SEO, transcreation, marketing, etc. where the goal is to connect with your target audience, Ideally, they will be conducted, or at least supervised, by someone with knowledge of the target market.
ChatGPT is not yet responsive to cultural differences or barriers of understanding and understanding that exist between markets and cultures.
- Businesses that regret not following the advice in this article.
Cases are known in which a company has gotten into a serious reputational or even legal crisis due to AI. Let’s look at some examples:
- Mango, the well-known textile company, translated a set of jewelry from the word “slave” into literal French, which not only provoked negative reactions from the French public, but even a whole movement led by two well-known actresses of the country, which demanded signatures against the lack of morality of the textile giant. (Shortcut).
- Maybe some reputation crises are desired? Maybe a brand wants to be talked about at all costs…? Because it is a great coincidence that something suspiciously similar will happen to his competitor Zara (Link).
- But what if an error due to machine translation does not remain a mere anecdote or reputation crisis, but an obstacle to entering a new market? Because that’s exactly what happened to Amazon in northern European markets like Sweden.
We hope we were able to convince you that we are still a long way from using AI or machine translation for our digital internationalization processes, although they can be an excellent support tool.
Remember that properly adapting content to international markets can be key to a project’s success. At Innova Digital Export and Ontranslation we are at your disposal to support you on this exciting journey.
Dori Lopez & Oscar Nogueras