For modern companies with international expansion on their road maps, the speed of software development and releases is of great importance. It helps them gain a competitive edge, increase market share, and delight customers across the world. This is one reason why localization projects are gradually adopting continuous localization workflows.
Continuous localization enables faster translation turnaround times, and therefore, faster times to market. Moreover, this methodology can positively impact the entire localization ecosystem and make everyone’s job significantly easier.
Although continuous localization still comes with some challenges, it certainly has tangible business value. But let’s start from the beginning.
Understanding continuous localization
Before we dive deeper into the value of adopting continuous localization and explore the industry’s best practices, it’s important to ensure we’re on the same page when it comes to what continuous localization actually is.
The truth is, continuous localization has various definitions and these very much depend on who you ask. Some professionals tend to use the phrase continuous localization interchangeably with Agile localization, but there are subtle differences between the two.
Agile localization implies incorporating localization and translation into the Agile development cycle. This means that the product is localized in multiple iterations, in parallel with each sprint. New software features are localized simultaneously, which differs from other methodologies such as waterfall.
Under the waterfall localization workflow, localization starts post-release or it uses so-called ‘string freeze’ periods. During the string freeze period, localization teams work on translating strings while developers work on fixing bugs. At first glance, this seems productive. Alas, developers still have to go back and invest their time in manually identifying modified strings. Depending on the size of the project, this can be a lengthy process.
So, where does continuous localization fit in? And how does it differ from Agile?
Maybe the most straightforward explanation comes from Miguel Sepulveda, Global Localization Manager at King, the interactive entertainment company famous for creating the Candy Crush Saga game. According to Sepulveda, the main difference is reflected in the following fact: with continuous localization, content is always ready for release; with Agile localization, you need to wait until the sprint is completed.
You could say that with continuous localization, you have one never-ending sprint. Within this continuous sprint, localization teams and software developers work together toward releasing new product versions in multiple markets in the shortest time possible.
Continuous localization in product development
If we are talking about continuous localization from a development perspective, it essentially means integrating localization processes seamlessly into the CI/CD pipeline (Continuous Integration/Continuous Delivery). That way, localization does not stand in the way of fast software releases. Rather, it supports reaching customer groups in each target market — all at once.
Continuous Integration/Continuous Delivery is a vital part of modern development practices that focus on making preparation for a release easier. Quality products are released frequently and end customers know what they can expect.
With continuous integration, developers frequently merge changes to the main branch and then use automated tests to check if the integrated changes caused any issues with the product. So, the development team doesn’t wait for release day to merge all the changes. Instead, the team does this continuously,which makes the entire process much more manageable and less risky. With smaller batches, troubleshooting gets a lot easier.
Continuous delivery goes hand in hand with continuous integration. As we mentioned above, with the CI/CD — developers rely on automated tests. However, the team also automates the release process so that they can deploy to production with just a few clicks. This ensures the best experience for the end users.
As you can see, CI/CD creates a productive environment in which the team can focus on rapidly improving smaller parts of the product and therefore have more control over the development process. Understanding the fundamentals of CI/CD and the value it brings helps us to understand the value of continuous localization.
So, what happens when you add continuous localization into the equation?
When you think about it, continuous localization supports the automation of the localization process and it makes it as seamless as possible. It follows the same CI/CD principles as it prioritizes both the quality and the speed of releases. Let’s take a closer look at how this works in real life.
Best practices for continuous localization
Without continuous localization, the localization process would look something like this:
Software developers finish one sprint and move on to the next one
Localization team begins translation in their next free time slot (e.g., in 3 days)
Translations get delivered
Developers need to revisit the code and implement translations
Developers start a new sprint
The process repeats
This is a very simplified version of a possible localization process. It can get even more frustrating if, for instance, the Linguistic Quality Assurance teams don’t get to review the translations in time. This would require developers to go back to the previous sprints and implement any necessary translation changes. Evidently, overlapping sprints can slow down the release cycle even further.
When you paint a picture like this, people tend to agree that the process is inefficient and could be significantly improved. Just think about the usually high hourly rates of developers and the time they waste. The challenge here is to find a sustainable way to implement continuous localization, but it’s not something that is impossible to overcome.
Here’s how to get started with continuous localization:
1. Make sure the localization team and the development team work as one
The first thing you need to do is bridge the gap between the localization team and the development team. If you have great business growth ambitions and are looking for a scalable solution, this is arguably the best way to start.
If you’re outsourcing localization, the biggest challenge here is finding reliable professionals who have enough bandwidth to handle the requirements, plus sufficient experience in cross-team collaboration; and moreover — finding them in a timely manner. The localization team needs to be well versed when it comes to 1) the product itself and 2) the target markets in order to ensure optimal translation quality and accuracy.
The value of continuous localization is reflected in the fact that it encourages you to plan far ahead, right at the very beginning of product planning. By its very nature, it implies a proactive approach.
2. Establish a responsibility matrix
Establishing a responsibility matrix or using any other type of framework to define the responsibilities of each team member is the second step.
You will certainly need someone who would implement the continuous localization workflow (typically a localization engineer), someone who would handle Quality Assurance (QA) and be responsible for the end results in terms of localization (typically a localization QA specialist), and someone who would ensure a healthy project pace, manage workloads, and deadlines (typically a localization manager of some sort).
3. Choose the right tool for continuous localization
When you integrate the localization team into the development team, you are fundamentally restructuring the way you handle localization to prioritize efficiency and faster time to market. Simply wanting to implement continuous localization is not enough, though. You need to choose the right tool that will respond to your business needs and support your CI/CD.
Look for a solution that enables automations (with automated workflow rules, you can ensure the project moves forward and minimize the lag effect)
See which code repository integrations (e.g., Bitbucket, GitLab, GitHub) are available and explore whether language files can be instantly imported via an auto-pull option
Focus on features that automate communication and enable the building of complex delivery chains (e.g., webhooks, powerful APIs)
Assess the quality of storage (cloud is the optimal solution because it enables real-time collaboration and is secure)
Ensure each member has the functionalities they need to do their job properly
The bottom line is, you want a solution that cuts down on manual tasks and allows for optimal flexibility. This way, when the sprint is over, you end up with a complete product that’s ready to be launched in multiple markets.
The best way to approach choosing the right tech solution is to write down your requirements and then ask for support and guidance from the product specialists of each potential vendor. Customer-oriented companies that offer localization software will gladly share specific best practices and help you optimize your workflows if you share details about the localization problems and bottlenecks you wish to solve.
Of course, increasing the speed of development is not the only goal. The trick is to find a way to preserve quality while maintaining speed. Continuous localization provides this much-needed velocity, it helps you optimize the localization process, and it increases productivity rates. Yet, the innovation and potential of this methodology is still to be fully recognized around the globe.