26/05/2026

An AI translation can be produced at speed. However, the question of whether it’s appropriate for a specific text or not must be assessed as part of a feasibility analysis. After all, just because everything can technically be fed into AI systems, that doesn’t necessarily mean that it makes sense to use them. The feasibility analysis therefore provides clarity in advance as to whether the content, language and file format are suitable for machine translation and post-editing (MTPE).

Particularly when it comes to parts lists, technical tables or other texts lacking context, AI often cannot provide a basis for reliable results. The language combination is also of importance. What works well in English may require considerably more effort in other target languages. Furthermore, the quality of the text is crucial. If the source material already contains ambiguities, gaps or inconsistent terminology, the MTPE process may quickly start to involve a lot of review work.

If the machine translation output is of insufficient technical or linguistic quality or contains errors, a significant amount of effort must be devoted to human post-editing. This involves reading, evaluating and adapting the machine-generated preliminary translation. In the worst-case scenario, the text may even have to be retranslated in full rather than simply being adapted. This can sometimes involve more work than a direct human translation of the same content. Correcting machine translation output that sounds fluent but is inaccurate in terms of content is particularly challenging. Mistranslations, omissions or additions are not immediately apparent and take up time.

And this is where a feasibility analysis comes into play. It checks in advance whether AI translation and post-editing are really suitable for the type of text, language and subject area. This avoids unnecessary work and ensures that the process remains predictable in terms of both cost and time. At the same time, the feasibility analysis lays the groundwork for an appropriate level of translation quality. Only if the text type, language and source material are suitable for MTPE can a result be achieved that is factually accurate and linguistically coherent. This allows you to assess at an early stage whether MTPE is the right solution or whether a human translation makes more financial sense.

For AI translation and post-editing to work effectively in day-to-day project work, you need more than just gut instinct. It’s important for you to assess four key points: technical, linguistic and content feasibility, as well as references and specifications.

Can the files be processed properly? Can they be integrated into the computer-aided translation tool? Does the machine translation system support the required language combination?

Is the target language suitable for MTPE? Is the machine translation output linguistically consistent and reliable? Are the source texts linguistically complete and correct?

Is the type of text suitable for MTPE? Can the appropriate style be achieved using MTPE? Is the content mainly continuous text or is it highly fragmented?

Do we need to take into account any reference material, GUI conventions, terminology or style guides? Are there any links to other texts? Particularly in a regulated or product-oriented environment, this factor is often crucial to the final quality.

Clarifying these issues in advance provides a clear basis for deciding whether or not to use machine translation. This reduces the risk of starting off on the wrong foot.

A thorough feasibility analysis not only leads to better decisions but also ensures that the project can be planned cost-effectively. It prevents texts from being processed in an MTPE process even though the machine translation output ultimately creates more work than it saves. This is precisely what helps you avoid unnecessary review work, additional costs and wasted time.

It benefits everyone involved. You can keep an eye on costs and avoid unnecessary additional work during post-editing.

Before each project, oneword assesses whether the use of AI translation and post-editing is appropriate. All the aforementioned feasibility factors are taken into account. The result clearly shows whether an MTPE process is viable, whether the texts need to be prepared or whether a human translation is the better option.

This helps to avoid costly surprises as the project progresses. You receive a thorough assessment before any costs are incurred, and can be confident that the process is both technically sound and cost-effective. This gives you a solid basis for your future planning.

Have your texts checked to see if they are suitable for MTPE. The feasibility analysis is free of charge for you. Feel free to contact us about this.