There is no easy answer to this question because it depends on who asks. To make it more complicated, it also depends on who pays.
The easy answer: The best CAT tool is the tool that does not exist yet.
In this article we will explore the different stakeholders in the translation ecosystem and how CAT tool companies monetize their product offerings. We will look into how the flow of money eventually influences how CAT tools are built today.
CAT stands for Computer Assisted Translation or Computer Aided Translation and refers to a scenario where a human translator gets help from a software application to increase output and quality. CAT tools are sometimes also called TMS, translation management systems.
A translator is someone who translates documents, software, online help content and so forth. While any piece of content is unique – built to order so to say – there are similarities between seemingly disconnected content. CAT tools attempt to automate the translation of these commonalities by using a database called Translation Memory (TM), which is connected to the translation process. Essentially, the translation memory offers the translators previously translated content and thus increases productivity.
Translation Memories, along with terminology databases, were the most important features of CAT tools in the beginning. Today, software manufacturers are incorporating new features that cater to the needs of other user groups, for instance project managers at Language Services Providers (LSPs).
The first company to offer translation tools was Trados, who started operations in 1994. According to Nimdzi Language Technology Atlas 2021, today there are dozens of CAT tools at this moment.
Dozens of tools sound a lot, but imagine what photographers need to go through when selecting a camera from the thousands of models available on the market.
Fun fact: Photography is similar to translation in that different tools do not necessarily produce different outputs. The tools just make the creation process easier or more complicated.
Follow the money
Companies want to keep their customers happy, because after all, unhappy customers do not make repeat purchases and do not recommend a product or service to others.
So, if we want to know why a product has certain features, we need to ask who actually pays for the product.
Regarding cameras it is clear: the photographer pays.
Regarding software products it is not so clear, because software products quite often are made for different stakeholders. In the translation industry these stakeholders are translation companies, also known as agencies, and translators.
If translation companies pay for the software, they will expect features such as workflows, dashboards and analytics. Eventually, their project managers will become more efficient because they can take on more work. As a result, the agency will be able to streamline operations and become more profitable.
Now, if translators pay for the software, they will expect features such as customizable quality assurance, the ability to add additional translation memories and glossaries as well as the ability to “hack”. Hacking in this context means to type really fast.
This last point does not seem important, however, some tools are so clunky that they invite translators to take multiple forced coffee breaks during the day, thus driving down the amount of earnings per hour. This is particularly true for online tools where the translator types the text into a web based application.
From what we have seen in the tools landscape, Trados Studio is the only tool where freelance translators have to pay for a license. Most other CAT tools are lent to freelance translators on a per-project basis, or they are free tools anyway.
Is it possible to keep all stakeholders happy? Let´s take a look.
Rather than giving recommendations for features, this section looks at the daily challenges of the two main stakeholder groups: translators and project managers.
We will look at this through the lens of job descriptions, in other words, tools need to enable stakeholders to do their job.
- “I need to monitor deadlines.” Project Managers need to track the progress of translators in real time. This is among the most compelling use cases of online translation tools. If translators say they are nearly finished but the Project Manager can see that the tool only indicates 37% progress, there is a problem.
On the other hand, PMs cannot check the progress if translators work in an offline tool such as Trados Studio. The only way to find out about the actual translation progress is to ask the translator to deliver whatever is ready at that point.
- “I need to send progress reports to my customers.” PMs need to be able at any time to tell their customers how a project is going. PMs need to be able to pre-empt problems, work on a solution and seamlessly communicate with their customers.
- “I need to manage translators, many of them.” It is not uncommon as a project manager to be outsourcing to 50 translators and vendor companies at the same time. No matter how proficient you are in planning and sending e-mails, you may get lost at a certain stage.
- “I need to manage KPIs – I want my bonus.” Project Managers are evaluated by customer satisfaction, project margins and deadline faithfulness. All these numbers need to come out of a reliable system.
- “I need to monitor and maintain client assets” Client assets are essentially translation memories and glossaries. PMs need to make sure they are properly updated and maintained and stored in such a way that others can use them immediately. Like walking into a library and finding the book you need based on certain meta information.
- “I need to provide insights to the folks in Sales.” My colleagues in Sales and other departments may ask me what’s going on in Production. Perhaps we need to find more translators in order to support that production peak in summer. Or maybe we need to find a way to create quotations with a higher win rate.
- “I need to be profitable.” Translators are service providers. This means that their revenues come from trading their time against money, rather than providing specific goods. Other examples for service providers are lawyers and business consultants. They all sell their time to their customers. Whether they charge an hourly rate, or a rate based on the words to be translated doesn´t matter.
At the end of the day, the faster they work, the more money they make. And that´s exactly the business case for efficient tools with an intuitive interface. Inefficient translation tools will slow translators down – they make less money – efficient tools will help them to finish earlier and give them more time to either take on more work, learn a new skill, pursue business opportunities and other things.
- “I need to keep my deadlines” This point ties in with point 1: There is nothing more annoying than a tool that slows down work and forces a translator to decide between delivering on time and compromising on quality.
- “I need to meet the established quality standards” Same here: Translators need to decide – or check with their client – what is more important for a given project, meeting quality standards or meeting the deadline.
- “I need to be able to provide feedback” Translators frequently act as reviewers. In that capacity there needs to be a way to give the translator feedback. The most accurate way to do this is to enable tracked changes in the CAT tool and then provide this version with tracked changes to the translators.
Another option would be to export the translated version to a Word file, subsequently export the reviewed version to a Word file and then use Word or a third-party tool to run a comparison. Some CAT tools do allow neither of this.
CAT tools have a high “stickiness”: Once implemented, companies are rather reluctant to change the established way of working, retrain all stakeholders involved and to admit they made the wrong choice in the first place. Switching to another tool is expensive and it puts the manager on the spot who decided to implement the incumbent tool.
If CAT tools manufacturers want to be successful and indeed create the best CAT tool, they should do at least two things:
- Based on numbers, frame the switching costs into an investment.
- Engage with all stakeholders – not just those who pay – and add product features that people actually want and are user friendly.
For translators it is difficult to have their voice heard, however, they can still go to rating portals such as G2 and Capterra and write a review. If the feedback is constructive and engaging, these portals sometimes even pay a small reward.
A totally different question is whether CAT tools will become obsolete anyway because machine translation is eating them for breakfast. After all, CAT tools are meant to assist the human translator, however, if MT does the job, what need is there for special translation tools and translation memory databases?