The Pieces of Cake
at the Localization Party
BY Camilla Amici
What is in the pot?
Something is moving in the localization industry, and it has the sound of a loud protest. Today, some fundamentals that once were taken for granted such as optimism, diversification, flexibility, and love for innovation are cracking. People on the bottom line are now standing up, because just following global trends and passively accepting the rules of a deregulated free market no longer seems fair.
We all know that after the pandemic, a global trend emerged: the Great Resignation. Did we need a pandemic to understand life is too short for jobs that make us unhappy and do not recognize our individual value? Whatever the answer, the light is now on. Following the Great Resignation trend early this year, some hot discussions appeared on the social media of localization professionals talking about the so-called “Great Renegotiation.” If we decide not to passively accept what the market offers, the first thing to do in the localization industry is to talk about rates. It is not just talking about money; it is talking about value.
From there, many fascinating debates and articles followed. Much of the freelance translator community now believes the time has come to call out the exploiters’ names. “Exploitation” is a tough term, true, but it sounds quite appropriate for those companies that become richer and richer year upon year while expecting big and regular discounts from their suppliers. Is it a free market? Yes, and this a fortunate occasion indeed, because this means that it can be influenced and oriented.
Not just a single course
Where to start? In my opinion, the first thing to do is analyze how many players are involved in the supply chain and which services they offer. At the present time, there are plenty of different services hidden in a per-word rate, and this simply does not work. If small LSPs have to fight over a few cents to be shared with their freelancers, in the end, nobody will be recognized for their value.
On many occasions, I’ve read comments by some translators who suggest working for direct customers only and avoiding the LSP’s mediation, as they just act as a “paper pusher” who deduct some cents from the people who do the real work. I am pretty sure that this can happen. However, it means that the LSP in question is not a good one and adds no value to the supply chain.
If the LSP is good, the services on offer are much more varied. Freelancers need to feel an added value in working with LSPs, because they have the technical and managerial support that is missing when dealing directly with end customers. Big companies need to see the difference when they contact the linguists directly and when all the recruitment, planning, and quality-check tasks are covered by an LSP. The house is standing only if every stakeholder is recognized and compensated for their value.
Speaking as someone on the small LSP side, I think we should start to differentiate services by asking a rate for project management and a rate for translation. We should negotiate the services separately, because they are different and non-standard in many cases. Already the linguistic work is more than just a “per-word rate,” but if it can at least leave out the agency portion, the translators’ piece of cake becomes bigger.
Two in a kitchen: the human and the robot
We all love innovation, and innovation loves us. How many webinars, conferences, and meetings did we attend to learn from enthusiastic machine-translation (MT) evangelists that the future is now? The claim is that human translators do not have to be scared by the advent of machines, as this generates a “more” effect in the industry: more content to be translated and post-edited, more linguists teaching the machines, more open roles for hybrid professionals who have both IT and linguistic expertise. This is all true, and trying to reject this model probably means shouting out alone after being marooned on a desert island. That said, how is the whole MT story evolving, and what direction is it going?
In the last few years, we have been contacted many times by our clients letting us know that the quality of the MT is highly evolved, and less human effort is needed than in the past to achieve ready-for-publishing quality. The logical result of this improvement is to ask the human suppliers for further discounts, as part of the job is carried out by the machines, and now we have clear evidence for this statement.
The first time I was in a similar conversation, my first thought was: “Wow, you get paid less for having done a great job!” I don’t want to appear naïve: I know how much MT’s raw quality has improved over the years and that, for certain content, the human effort requested is low. However, I also think about all the efforts that post-editors have made to tailor a service that in fact is quite tricky. Ten years ago, the first impulse of a translator looking at an MT raw-output string was to delete everything and re-translate the string from scratch. In fact, they were not able to provide the service for which they were being paid (less). Today, linguists have learned how to apply “surgical interventions” to the automated translated sentences and make the most of what they have. This has been so challenging to reach that I can understand their disappointment when they receive periodic requests for rate reductions. Not considering all the planning issues that a similar system may cause.
The conclusion I have come to is once more: not a single service and not a standard unit. I can agree with the use of edit distance report parameters in defining rates for straightforward workflows: simple content, few instructions, high-quality MT output, and almost no follow-ups after delivery. But what if the process is much more troublesome? For instance, what if the time to read the reference material, comply with hundreds of instructions, and handle implementations and feedback replies to a picky (human!) reviewer is endless? If we can accept to be paid per modified target word or per highly discounted source word, we should likewise expect that all the additional services will be paid separately by considering the time to get started (references, instructions) and follow up (implementations and feedback). At the end of the day, they are other pieces of cake.
Every year, fierce armies of smart students complete their translation studies and are eager to take their first steps in the profession they have so long dreamt about. On one side, they are warmly welcomed by the industry, especially those that aim to offer them additional learning courses or to add the mission of inspiring new generations to their company portfolio. On the other hand, the same companies frequently reject their applications as they have no relevant experience or offer them fees where, again. the term “exploitation” can be called into question. If these young translators accept those rates, as it seems the only way to step in, they can also be accused of ruining the market by lowering already too-competitive rates. So, how to break the deadlock?
Unfortunately, as a vendor manager, I have verified many times that after a university degree or post-graduate degree people are not ready to do the job. What they need at this stage is not additional theory but to translate, translate, and translate and have their translations corrected. This is, of course, a big investment for an LSP, but it is the only way to help people grow. Sharing a nice video of how CAT tools work is helpful as well, but it won’t teach anybody how to translate. Every time I have seen aspiring translators become reliable professionals, it has always been because they have had the chance to have someone to correct their work, sending dozens of reviews, comparisons, and feedback forms to show errors and improvements.
This is what I would like to suggest to anyone setting out on their professional path: Take a breath and be aware that you are in the lion’s den. When trying to contact an LSP for a potential collaboration, figure out the value that they can give. Does the LSP have an internal linguistic department and an ongoing QA program? I know that it is something quite old-fashioned nowadays, but if it exists, it is proof that the LSP cares about quality, and there is someone who looks after reviews. Be suspicious when you work for years for an LSP without ever receiving one single item of feedback, because no news does not always mean good news.
Every young professional is free to head immediately to direct customers and demand high rates. For some, it may work for different reasons that may occur all together: great talent in translating, amazing selling skills, ability to catch the right moment, a bit of luck, and so on. For all the others (and I include myself when I was at the starting blocks), I suggest studying the whole-industry scenario and trying to understand who can give the most value. Also working for a more experienced freelance translator can be a real asset. Someone could call it “subcontracting,” but I honestly think that it is an amazing occasion to develop your own learning curve. Of course, this only applies if almost all the jobs are double-checked and corrections are constantly shared!
The end of the party
Something is moving in the localization industry, and we should not miss this chance. The localization supply chain is very crowded at the moment, and AI is constantly knocking at the door. We should not be scared, but at the same time, we should act for better conditions for everyone. We must demand to be compensated for our own value and recognize the value of people and companies around us. Otherwise the system we have created — one which seemed to be working perfectly — might likely implode. It is not just a piece of cake, but it is worth trying.
Camilla Amici works in translation, vendor and quality management, and marketing and sales at Novilinguists.