The Pillars of a Future-Proof Localization Workflow
Explore what it takes to build a sustainable localization workflow that enables scalability and boosts growth in the long run across global markets.
According to a 2020 McKinsey survey, 66% of business leaders across multiple industries were piloting the automation of at least one business process—an increase of 9% compared to 2018. Whether in developed or emerging countries, companies are finding themselves under more and more pressure to streamline internal workflows to support scalability and growth.
This is especially true if they need to adapt their products to different target markets simultaneously. For successful localization in today’s global marketplace, you need to get rid of spreadsheets, lengthy email threads, and disparate desktop applications before competitors start to weigh in. Learn how to optimize your localization workflow to start getting ahead.
Map out a localization-aware global strategy
As a crucial endeavor for any business seeking international growth, localization often comes to light only when it’s done poorly. For many companies out there, localization still comes as an afterthought and isn’t reflected in their global growth strategies.
Back in the day, succeeding globally was mainly about having a strong brand. Today, we live in the age of the customer, where sustained growth goes hand in hand with customer satisfaction and centricity.
Companies need to address their existing and potential customers like they know them intimately, regardless of their language and culture. This means delivering a range of assets—product information, marketing collateral, technical documentation, customer service interactions, and a long list of etceteras—in a way that feels truly local and authentic.
Localization is all about adaptation
When you decide to launch your product in one or more new markets, even before you set your localization workflow in motion, you need to assess how the product fits the requirements of each market and identify any modifications it might need. What are the linguistic, cultural, legal, political, and business specificities of the target region?
The lack of awareness of linguistic and cultural differences has caused a few expensive marketing blunders over time. The quick translation—or lack thereof—of a product name or a slogan can poorly impact product adoption, and it has happened countless times. For instance, the Ford Pinto launch wasn’t a success in Brazil because ‘pinto’ is Brazilian slang for male genitals.
Fully considering localization right from the start of a project is the best way to avoid Ford’s mistake and ensure that all aspects of the business collaborate effectively towards the same goal.
Optimize brand consistency with translation technology
As your customer base grows in several markets simultaneously, ensuring globally consistent messaging—a crucial factor in driving brand loyalty—while fulfilling local market expectations can get tricky.
The best approach to optimizing brand consistency is adopting translation solutions that will:
- Ensure complete continuity between departments by enabling a workflow that brings all stakeholders together onto the same platform.
- Allow you to group users in teams (e.g. by language) and in projects, so they can work collaboratively throughout the localization process.
- Increase efficiency, output quality, and terminology unification through term bases, translation memories, error detection functionalities, and other features.
Foster team collaboration across functions, time zones, and geographies
Most of the time, the source content that needs localization originates from separate teams. However, approaching the localization workflow in silos puts you at risk of dealing with quality issues (like inconsistencies in translated assets) and an overall lack of efficiency, including in terms of costs.
Centralizing efforts proves more productive. Having just one team keep a helicopter view of all things localization is an excellent way to keep all your departments more in sync on all aspects of the business.
A one-team, one-tool approach will keep your localization workflow streamlined, effective, and timeproof. The result will be no more duplicated efforts or tools, with undisrupted growth.
Awareness is key
Raising the profile of the work of the localization team to internal stakeholders lays the foundation for a future-proof localization workflow.
On top of educating every company department, making everyone aware of your localization process is crucial for the success of a global brand.
Keep the different company divisions (Product, Marketing, Design, Development, Customer Service, etc.) in the loop about every step of the process by having your Program or Localization Manager present your team and work to them. People collaborate better when they understand each other and how they work.
What makes a strong localization team?
Selecting the right professionals is going to be key to the success of your company’s localization effort. A great setup combines management functions and people with specialist skills.
Here are some must-have roles for a strong localization team:
- A strong team-lead visionary: You need someone who can rise above the specifics of a particular situation and see the big picture of the localization strategy and overall team efficiency: think localization manager or program manager, for instance.
- A reliable project team: This is the part of your overall team in charge of the day-to-day work of localization projects. It includes translators, localization engineers, and project managers.
- Dedicated quality assurers: This area typically involves linguists focused on checking the overall quality of the work produced by your translators, as well as testers for products like apps.
Combining in-house power and external support from a language service provider (LSP)
While having an in-house expert team in charge of localization is a must, your workflow might involve outsourcing some tasks to an external provider of localization services. It might either be on an ongoing basis or when there is an uptick in work. Here are a few tips to make that work for you:
- Vet your vendors: When you reach out to an LSP, ask them about their workflow (including quality assurance) and their requirements when hiring their resources (do they ask translators to take a test, or do they check degrees, or both, for instance?). Do this before you have a translation project which needs outsourcing, so you don’t rush through the process.
- If you work with small LSPs, have at least a couple of vendors for each language combination, so you’re not at the mercy of your language provider’s availability.
- Have your team regularly perform linguistic quality assurance (LQA) on the output of your vendor’s work and share the feedback with them. This will be beneficial to the overall quality of your localized assets and the collaboration with your LSP.
LQA is the most effective when performed by a linguist using an automated quality assurance tool, a feature that is a core element of most computer-assisted translation (CAT) tools and translation management systems.
Choose a single platform to centralize, automate, and integrate your localization efforts
In addition to centralizing your localization efforts under one team, you should also centralize them on a single platform. Having everything in one place often makes things more manageable, and localization is no exception. The good news is that a translation management system (TMS) allows you to do just that.
A TMS works with all types of file formats—from Word and PDF for your legal team to HTML and INDD for your designers and web developers. That way, all the work is done and stored in one place.
A TMS also makes the life of your localization project manager easier. They can manage in-house team members (from any department) and vendors alike on one platform, adding users and allocating access rights. They also have easy access to reporting and invoicing capabilities.
Meanwhile, your translators and editors log into the same platform to do their work, accessing linguistic features such as a translation memory and a termbase. Users are able to leave comments and questions and upload documents, which is very useful when translators have questions for the legal team or the web designers—no more information lost in long email chains.
Seamless integration and end-to-end automation
A translation management system is your best ally to achieve a sustainable localization workflow because integration with your digital ecosystem is part of its DNA.
For example, integration between your TMS and Figma or Sketch will allow your designers to easily send their work for translation and retrieve the translated output. Integration with Git or GitHub will let your developers do the same. And if some of your tools are not covered by default by a TMS, you can create your own integration thanks to an API.
Integrate machine translation into your localization workflow
To localize all your digital assets, especially when dealing with multiple target languages at once, you need strong tools that allow you to work at an increased speed. Machine translation (MT) has a lot to offer in that respect. It’s powered by artificial intelligence, doesn’t include human intervention, and can quickly produce a high volume of translations.
Raw machine translation has considerably advanced over the years and is useful for low-visibility or low-traffic content, such as internal documentation, website footers, or repetitive technical content like instruction manuals.
To ensure machine translation quality, your translators can go through the output and do machine translation post-editing for:
- The translated output to be legible and accurate—this is called light post-editing (LPE)
- The translation to be error-free and ensure it takes into account important aspects such as style, tone, and cultural nuances (full post-editing or FPE).
To increase your translators’ productivity and reduce translation costs, you should look for a TMS that comes with the ability to estimate the quality of machine translation output. This way, you’ll be able to channelize post-editing resources where they are most needed.
Equally important, if a modern translation management system comes with fully managed machine translation engines, it should also be able to automatically select the best-performing one for your translation project, based on engine performance data.
Go for agility and continuous delivery
Adopting agility is a great way to avoid expensive localization mistakes. Let’s consider your product development process, for example: In a waterfall approach, localization typically happens at the end of the project lifecycle once your product has been fully designed, developed, and tested.
Under agile product development, embracing a continuous delivery approach for your localization workflow is only logical. When your designers have created a draft version of the app homepage, for instance, the translators can access it and do their part. They can raise any issues they may see (think “the Spanish call to action doesn’t fit in the screen”), allowing the designers to go back to the drawing board before it’s too late.
Break down the localization process into manageable steps
Once you’ve laid the groundwork, it’s time to organize the localization process into clear stages that allow for a clear overview of responsibilities, translation projects, and their progress.
Since it’s an agile localization workflow, the below phases iterate as the content becomes ready for translation, as opposed to a waterfall fashion in which every step is completed before the next begins.
1. Preparing assets for localization
This stage refers to the process of making all assets available and suitable for localization. Typically, it’s your Design and Development teams that will take care of this step by completing two major sub-stages:
- Internationalization: While defining the product’s look and feel, your Design and Development teams can take certain steps to ensure that, regardless of the language a user speaks, they will still have access to all the content. Some internationalization practices include, for example, leaving room for text expansion and supporting right-to-left languages in user interface elements, and allowing data input formats to vary between languages for dates, times, and currencies.
If you’re unsure where to start, check out this checklist that Microsoft, an internationalization pioneer, put together.
- Resource file extraction and organization: After the internationalization of your product, your developers can start extracting from the code all the textual content (translatable text) into resource files. This will allow them to import the translated content back into the source file without touching the code.
During this step, they will also name and store each file appropriately for each language in a content repository. Later on, they will import the translated versions back into an updated version of the same file.
In terms of file formats, they need to be compatible with the localization tool that you’ll be using, although most tools support a wide variety, from Word and PDF documents to more specific extensions such as XLIFF, XML, or PO.
2. Setting up the localization environment
Now that all files are ready to export, your Localization Manager can start the process of setting up the localization tool environment—typically, a translation management system (TMS).
Here, they can create different projects (for example, one for your app and one for your website) and onboard translators and reviewers to work on them. They can also let other stakeholders access it so that they can see progress, provide feedback, answer questions, and more.
Your Localization Manager will also import any resources that will enhance the quality and consistency of translators’ work, such as translation memories, glossaries, style guides, etc. This way, your translators can benefit from translation memory matching and review tools that will continuously check each translation against your existing content, flagging any inconsistencies.
3. Content localization
This phase is where translators start the actual localization work. They’ll usually have to work in a translation editor within the TMS that shows two panes, one for the source language and one for the target language.
If a translator needs additional context to translate a certain word or sentence, they can normally bring it up within the editor by simply clicking on a special icon. This will allow them to search through and access all relevant assets such as glossaries, style guides, translation memory matches, UI screenshots, and more.
As they complete their part, translators will submit their translations to your Localization Manager, who can push them onto the next step: revision.
4. Content revision
Translators’ work may require revision to correct errors in spelling, grammar, and tone. This is why your Localization Manager will often invite reviewers into the process.
The revision stage is usually a lot more efficient when you have a TMS in place because all stakeholders can collaborate in real time, in a single platform. This makes it possible, for instance, for reviewers to leave comments and ask questions to the original content authors or to the translators, and for them to share knowledge with one another until the localization is confirmed as complete.
Once the content has been revised, your Localization Manager can approve it, and developers can import and replace it in their source files, ready for deployment.
5. Localization deployment
During this step of the localization process, through UI testing and code checking, your developers will ensure that the localized version of the content is fully compliant with your product requirements. They will:
- Import the localized files from the TMS back into the code base in a compatible format.
- Test the localized versions against the source files to ensure that they are identical (i.e., that only the translatable text varies between them, but not the code).
- Deploy the localized files across any other platform that supports deployment, like a CMS or the App Store.
- Repeat for all languages into which you’ve localized the source content.
The actual deployment process might be a lot more involved for some assets than for others (e.g., a 100-word customer service email is a lot simpler to deploy than a landing page). However, there is always one constant: You need to test localization functionality to make sure everything works properly.
This way, you can be confident that your customers will receive the same experience in any language.
6. Post-deployment QA and monitoring
The localization process is never finished once you launch your product or service globally: There may always be something new to localize, especially when your development is agile.
Therefore, after the launch of each iteration, it’s important to monitor how well your localization has been received by users. This can include hard data about how many downloads you received or, for a website, how frequently people visit your international landing pages. But it can also include qualitative data, such as user feedback on social media or inquiries from your international offices.
With a TMS in place, you can normally measure all these things and more through its reporting functionality.
Localization sustainability is about readiness
The steps we’ve outlined above are only a general overview of how an efficient localization process can play out. They will likely be specific to your type of product and localization use case. However, they build a sustainable baseline for a localization workflow that can take you far in the future of globalization.
As your business grows, an efficient and scalable localization workflow will allow you to scale up your localization speedily and effortlessly. Plan for growth from the outset by choosing the right tools to future-proof your workflow. This includes choosing a powerful TMS that will be flexible and scalable enough to grow with your business and help it reach its full potential in any market.
Looking for more information about how to set up a strong localization workflow? We suggest these guides: