Why continuous localization sucks?

The answer is simple: developers often forget that translation is carried out by humans. And naturally, you cannot think of humans as data processors if you don’t give them efficient tools to process data.

Let’s imagine an application going into 12 languages. It is translated and proofread, but there is a new version which requires updating two strings which account for 30 words of translation. To get the strings updated on time, the project manager at the translation company (or within the IT company) needs to contact and coordinate the job of 24 people. 

In a very optimistic scenario, the project manager is able to staff a team that is already trained and up to speed on the project. A more pessimistic but realistic perspective involves staffing the team with new people that need to invest hours of preparation to get those translations right. The risk for potential errors in terms of consistency and accuracy becomes a real issue. 

Finding the translators is only part of the job. The project manager also needs to handle queries (query management is key in continuous localization and is better integrated with development or ticketing tools), channel them to the customer and feedback the translators. 

The real challenge in continuous localization

What appeared to be just 30 words of translation, even when equipped with the right tools, ends up costing valuable project management time to most companies in the market. The real challenge in continuous localization is not translation efforts but rather project management, query management, and in-context review that may take more time than translation itself. 

But let’s dig deeper, as it can get even more complicated when translations are not done directly by freelancers. Imagine now the 12 languages are outsourced to two different multilingual vendors (MLV) for translation and review. There is one project manager in each MLV that further outsources to smaller companies that provide translation for one or a few language pairs, the single language vendors. Let’s also imagine that every company uses a different single language vendor for every language pair. So when the job comes, they need to contact at least one translator each — which is a best-case scenario as translators often refuse to take jobs. 

That adds up to: 2 multilingual vendors x (1 project manager per multilingual vendor + 12 project managers per single language vendors + 12 translators/reviewers) = That gives a total of 50 people involved! And don’t forget they all need to perform the job within a number of hours.

Continuous localization usually means delivery deadlines between 24 and 48 hours (in our sample, 73% of the companies reported they work with deadlines longer than 24 hours). But in some cases, 4-hour or 12-hour delivery requirements are also very common. So even though the nature of the project is simple, the amount of tasks is staggering.

A myriad of tools: no standard solution 

Today, there is no single “standard” technology for managing continuous localization efficiently.  Besides the whole range of translation management systems for enterprises and other popular translation tools like memoQ, XTM or Memsource all supporting continuous localization, there is also an entire spectrum of tools for software developers. Developers love these other tools because they follow their working logic, but translation specialists are often not so keen on them — just take a look at the section Localization for developers on Nimdzi’s Language Technology Atlas where, by the way, BeLazy is still missing! 

At the end of the day, the unfortunate translation company that wants to run their operations properly (which means using a business management system to be able to access relevant business reports) still has to spend between 3 to 5 minutes copy-pasting information from one system into another, for every project.

Continuous localization with BeLazy

The heterogeneous nature of the technology for continuous localization is what called BeLazy to life: product development shall become a single process and all stakeholders shall be involved. 

BeLazy provides the infrastructure and the toolset for seamless continuous localization across the supply chain: translation companies (both multilingual and single language vendors) and translators. We are starting to build up the connections that will enable technologies within the business of continuous localization to talk to each other like never before.

Still, today’s main issue is the industry’s lack of toolset to enable fast and standardized collaboration, regardless of the technology for translation or business management being used. For every single project they receive, project managers still need to carry out the following tasks: 

• Copy-paste the project information
• Contact the vendors
• Change the vendor in the translation management system (or send the files) 
• Receive the project
• Change the vendor to the reviewer again
• Wrap up the project in the translation management system
• Deliver it often via a portal as well
• Close the project in their business management system
• Get ready for invoicing. 

So our vision is to change this and transform the role of project managers in the translation industry: instead of spending time with manual copy-pasting, which is boring, uninspiring and error-prone, we want them to focus mainly on delivering the best quality to customers, and only dealing with manual tasks when there is something unusual — this is also known as exception-based or lights-out project management. 

To do so, we are working closely together with business management system vendors such as Plunet and XTRF to improve workflows and eventually, also integrating with translation management systems. But it doesn’t stop there. Because continuous localization does not have a single established system, our goal is to play well together with any technology, and enable efficient continuous localization for everyone.

At the moment, BeLazy works best for single language vendors that receive projects recurrently from vendor portals from multilingual vendors. BeLazy automates what used to be a 3-5 minute project creation to the click of a button or even less (watch our announcement about auto-approval coming later this month!). However, our vision is to become the middleware for all continuous localization-related tasks.