By Marcos Chiquetto
Translate a text is to read something in one language and write it in another language. What kind of technology would be required?
Let´s take the example of our Brazilian translation agency, based in São Paulo, a small-scale organization considering the global translation industry. We have two directors, six project managers, and a group of about40 active translators and reviewers, including in-house personnel and freelancers. We work with three languages: English, Portuguese, and Spanish.
And what kind of volume a company of our size handle? Well, let´s consider our volume of work in 2019. In that year alone, we carried out a total of 5230 projects.
That is right. Our small company executed five thousand, two hundred and thirty projects in 2019. This works out to an average of 20 new projects per work day.
And what type of projects are we talking about? The following table shows examples of our large, medium, and small projects:
These are figures for a small Latin American translation agency. What kind of volume would a multilingual global agency handle? Let´s assume the company works with 20 languages, and has, on average for each language, a quarter of our company´s volume. This company would process something around 25 thousand translations jobs per year, or about one hundred jobs per workday.
To continuously cope with all those jobs, a series of actions are necessary:
- Each job must be assigned to a project manager, who will choose translators and reviewers, estate the tools to be used, and lay down a schedule. All these decisions are registered in a job database, whose control software is the backbone of the company.
- Once a job is set up, the actual translation must be done. Even though the translators and reviewers may not be the same of previous jobs for that client, the style and terminology are to be the same. Moreover, text pieces of previous jobs must be reused, to assure consistency and reduce the price for the client. For all those purposes, a translation memory software is mandatory (see our article on translation memory).
- To be competitive, the agency may also decide to use machine translation as the first step of the job, increasing the productivity of its human translators. For that purpose, a machine translation engine must be available.
- Once finished the work of translators and reviewers, a final quality assurance procedure is performed by means of a QA tool.
- Then, all service providers should be paid, and receivables from all clients must be controlled, what demands a software module to handle thousands of small and large financial operations.
- Last but not least: all previous activities are only possible because the agency has clients. For that purpose, a permanent marketing effort is necessary, involving some software tool to handle databases of contacts, prospective clients, marketing pieces, etc.
The above summary alone shows at list six different kinds of software tools that are necessary to run a translation agency.
And even having all necessary tools, your translation will only be as good as the human translators you have in your team. If you have poor translators working in your translation jobs, no software tool will produce high-quality work.
Definitely, it is not a trivial task.