How Long Does It Take To Manage a Translation Project?

Play
decorative image

 

One would think that with the advancement of technology and the rise of analytics and insights, it would not be very difficult to calculate how long it takes to manage a translation project. But perhaps to your surprise, there is no formal research conducted on this topic.

However, for those who live and breathe translation automation like us, thinking and discussing this problem is essential to achieving greater business efficiency. In this blog post, we break up the most important factors that influence project creation and management.

Project Management: A Definition

What does a project manager’s role include? Each company is a world on its own that requires project managers to perform a unique set of efforts that vary in scope and range. Still, there are tasks that do not belong strictly to the project management sphere, even if many project managers carry them out on a daily basis. Here is some food for thought.

File Preparation

Does file preparation (such as PDF conversion) belong to localization engineers or project managers? We will consider it a separate job. It is the tasks associated with the standard workflow in a CAT tool (creating a translation package or a project in an online system) that we will reckon as raw project management.

Quality Assurance

Running QA tools such as XBench is a task that falls into the linguistic sphere, but that is often carried out by project managers. Like file preparation, we don’t regard this to be the core of project management, so we will consider it a separate job.

Desktop Publishing

While many project managers deal with desktop publishing on a regular basis, we also think it’s fair to say it is a separate job.

Pre-delivery Checking

Is it a PM’s job to open and check files before delivery? How much effort should be assigned to the task (it could be a 2-minute spot check or a read-through on the entire document)? For the purpose of this post, we will assume the project manager trusts the reviewer and consider pre-delivery checking a separate job.

The Project Workflow

The project workflow constitutes the cornerstone of a project management department in every translation company. In order to fully understand how much it takes to manage a project, it is also important to address the following:

  • How does the client order the translation (email, customer portal, etc.)?
  • Is there quoting? Most software platforms don’t provide productivity data regarding quoting processes, which, from our experience, can usually take longer.
  • Is there communication previous to executing the project?
  • Is the customer using a translation management system? If so, do you have to reassign translators, or will you work with packages?
  • Do you import files into your translation management system, forward the files, or simply reassign the job to the translator?
  • Is there a language quality workflow (LQA)? If so, does it qualify as time spent on the project or is it something separate?
  • Are there customer changes that need to be implemented? If so, who does the implementation?
  • What steps does the delivery of the project require?
The Specifics of a Project
1. Project size

Larger projects (speaking of volumes) usually take longer to prepare and manage than small ones. Project managers often have to split them up into multiple chunks and coordinate with vendors regarding linguistic quality. Larger projects also involve long waiting times at each stage of the workflow, whether it's downloading and uploading of files or getting/sending a file analysis. Finally, they take significantly longer to quote.

2. Number of languages

Creating a project with one target language is more time-consuming than adding additional languages to the same project. Adding a target language can take between 15% to 30% of the project creation time.

3. An automated vendor selection system

Companies equipped with systems that send out automatic requests to vendors (freelance translators, in-house translators, translation companies) can add new target languages to projects faster as they streamline all the steps associated with a manual vendor selection process: writing emails, phoning, waiting for confirmations, etc.

4. The client review/LQA step

When there is a client-review step in the translation workflow, the project manager often needs to go back to the project to refresh the translation memories. This usually takes a significant amount of time.

5. The number of extra systems used

Not long ago, it was common for companies to store documents and files in a network drive rather than in a business management system. Nowadays, we know that many project managers do not only register projects in the main platform, but also create copies on another system such as Google Sheets for better traceability — and it turns out to be a very time-consuming task (even if, in some cases, it can be eliminated by introducing certain views in the business management system). Many of these tasks originate as part of customer requests for specific data. That’s why we always recommend talking to the customer and checking whether a data export would suffice.

6. Technology

Most translation companies receive standard projects, which should be easy to create and manage. But handling this kind of project is not always done quickly and fast in their platform of choice — often robust and powerful technology platforms that have been designed for complex projects and workflows. And it is also common that these platforms are not able to cope with a certain task, forcing your project managers to spend valuable time outside the system.

In this regard, your technology platforms will determine your possibilities for workflow creation, how easily you can add and select vendors, set deadlines, and how vendors can work with you.

We recommend calculating the number of systems involved during the management of a project. For instance, a single language vendor may receive projects in a specific vendor portal but perform the work in a different translation management system. In that regard, system integration can help you save time: If a vendor has been selected in your business management system, an integration can instantly reassign the document to that vendor in an online system. Doing the same job manually can take up to 3 minutes (starting the program, opening the project, finding the file, adding the vendor, reassigning it, and setting up the deadline).

7. Client instructions

When a client sends detailed instructions, project managers need to spend time reading through them and sharing these with translators. Some platforms allow project managers to turn instructions quickly into a project template, but that’s not always the case. As a general rule of thumb, projects with lengthy instructions take longer to create and execute.

8. Invoicing

Invoicing doesn’t belong strictly to project managers, but given it’s done on a per-project basis, it usually needs a lot of input from project managers. Each customer has different requirements, and even if invoicing is often simple and straightforward, in some cases, it may end up consuming several minutes per project.

How long is a project in general?

At the moment we don’t have a perfect answer to this question, but we know it varies according to factors such as customer, project type, and whether you work for other translation companies or end customers. On average, creating a translation and review project  – ordered via email – into three target languages consists of the following steps:

  • Confirming the project to the customer via email (2 minutes)
  • Entering the project details for one target language into the business management system and uploading the files (4 minutes)
  • Adding the two other target languages and setting up the start dates and deadlines (1 minute per language = 2 minutes)
  • Creating the job in a CAT tool (3 minutes)
  • Selecting the translator and reviewer with an automatic vendor selection (1 minute per language, 2 minutes)
  • Reassigning the translator and reviewer in the CAT tool (0 minutes with integration)
  • Updating the translation memories (0 minutes with integration)
  • Delivering the translation to the customer (3 minutes)

Without having yet opened the file, the project manager has already spent 16 minutes. In this example, the project manager spent 4 minutes on the additional target languages which represent 18% of the project creation per language (not considering customer communication, thus only counting the second to the seventh steps, 4 minutes / 11 minutes / 2 languages), or 25% with customer communication included.

Conclusions

Language service providers that want to raise efficiency need to clearly define what’s project creation and management in practical terms and when it comes to the use of technology. Tasks that are related to this workflow, but which are not strictly project management should not be counted. 

All in all, managing a single language project should take between 10-30 minutes. Companies that can automatically assign vendors (a feature in many translation management systems) will be close to the 10 minutes, while those who do it manually will naturally spend more time.

It is important to remember that adding language pairs to the same project doesn’t take the same time as creating the first language pair. On average the project manager will need between 15% to 30% of the time taken for the first language pair (with less variation because the vendor selection is probably the same in both cases).

István Lengyel

István Lengyel is the founder of BeLazy Ltd, the company that makes agile localization possible across the entire supply chain. Previously, he was one of the founders of Kilgray, the makers of memoQ, and held various positions in the company. He tried most roles in the industry: he worked as a translator, as Director for Customer Success for On Global Language Marketing, and as a consultant;for;several other companies. He holds several degrees including a PhD in translation studies from the ELTE University of Budapest. He is passionate about technology, and believes that machine learning and artificial intelligence will have a much more far-reaching effect in the society than most of us would ever imagine.