Traditionally, teams shipped in one language first and only thought about localization once a feature was live. It worked for a while, but it also meant designs broke in translation, non-primary markets had to wait for “later phases”, and teams spent weeks fixing issues that could have been caught on day one.
Design-stage localization changes when localization happens, and that makes all the difference.
In this article, we’ll walk you through what design-stage localization looks like with Lokalise + Figma, how AI-powered translation and quality scoring keep things fast without sacrificing control, and how product and marketing teams can reuse the same workflow to ship multilingual screens, emails, and campaigns in minutes.
🧠 Good to know
Incorporating the localization process early in the design stage can help avoid costly rework down the line, ensuring that translations don’t impact your design negatively and keeping the overall process streamlined.
Design-stage localization means you plan and test localized content while designs are still in tools like Figma and Figma Buzz, instead of waiting until development or campaign production is finished.
Everyone works from the same source of truth early on, so you can see how real copy behaves in different languages before anything goes live.
For teams using Figma, that looks like:
The result is a shared workflow where product, design, marketing, and localization teams catch issues early, avoid expensive post-release fixes, and ship multilingual products and campaigns much faster.
In the traditional dev-led workflow, localization only kicks in once a feature is “done” in one language. Designers create the UI, copywriters finalize the English, development builds it, QA signs off, and the feature ships in a single market.
After that, a second loop starts: the copy is sent for translation, developers pull new language files into the product, and only then do issues surface.
This sequence almost always leads to the same problems:
Taken together, this means localization is always a step behind the product. Teams spend time fixing avoidable issues, global rollouts are hard to coordinate, and users outside the primary market get a noticeably weaker experience. And that’s exactly the opposite of what a growing product needs.
Can you get away with the traditional workflow, though? The answer is yes, but this is far from an optimal solution. Let’s see what are the three most common issues teams run into.
Different languages require different amounts of space in design. For example, translating from the English language into German can result in a 20-35% text expansion. Translating from English to Swedish can result in a 20-35% text contraction.
slack.com vs slack.com/intl/de-de/
Plus, if you want to include Korean, Japanese, Chinese, and other non-latin script languages, this can also result in vertical text expansion.
Furthermore, it’s important to remember that some languages (such as Hebrew and Arabic) are written from right to left. As you can imagine, this can, and most likely will, distort or break your design and lead to bugs.
Even more so, localization is not just about translating words, your design might not be suitable for other markets for cultural reasons, which can lead to low adoption.
Your translators likely won't be going through your UI and trying to figure out where each content key is supposed to be placed, what its function is, and how it will look after translation.
This becomes even more cumbersome if translators are not provided with any contextual information in relation to the translation and its intended purpose (like screenshots).
If this is not kept in mind, it will lead to translation and localization errors and a confusing experience for your multilingual customers.
The first two issues above are common localization challenges, and will always be present when designing and developing multilingual software products.
However, if you work with translations only after development, you'll encounter two big challenges that could have been avoided or significantly reduced:
This adds up and localization becomes unnecessarily expensive and time-consuming. In turn, this also slows down your releases to all markets resulting in lost opportunities (revenue from all your markets), as your team will be busy fixing existing errors and spending more time on LQA (Language quality assurance).
We calculated the associated costs of l10n (localization) bug fixes and the numbers aren’t pretty at all. For the average mobile app, in terms of complexity, the total costs of fixing those bugs can reach $52,800 / year.
The biggest challenges with translating content after development are the later releases of our projects, or releases with less languages (due to tighter deadlines). Maintaining changes in development and keeping current keys in sync is also more difficult after development.
Črt Mrzlikar, Project Manager, CNJ
As you can see, the traditional localization workflow is just not as efficient as it could be.
Design-stage localization brings all decisions into Figma and Figma Buzz, where product UIs and campaigns are actually created. Instead of guessing how German, Japanese, or Arabic will behave, teams see real copy in real layouts and can adjust before anything ships.
For product and marketing teams using Lokalise + Figma, that means:
💡 Did you know?
Teams using Lokalise AI orchestration see 80%+ of translations accepted without post-editing, which can drive translation cost savings of up to 80% compared to human-only workflows.
Here’s what an optimized localization workflow looks like when you start from the design stage:
In the design-led model, the whole flow is rearranged so localization isn’t a second loop after release. Here, localization is baked into the way features are designed and shipped.
Designers, writers, translators, and developers still touch the work, but they do it in a single pass instead of in two separate tracks.
The diagram you can see above shows three big shifts:
In practice, this means fewer surprise issues at the end, less duplicated work across teams, and a smoother path to “day-one parity” between your primary market and every other market you care about.
Thanks to the new ‘duplicate’ feature in the Lokalise plugin for Figma, we went from 10 hours to 1 hour of task management time spent on localizing a new feature. That’s a 90% faster localized feature rollout.
Romain Dahan, Product Manager @ Withings
With Lokalise, you can now add designers (and copywriters) into the mix early on.
The same Lokalise plugin powers both Figma Design and Figma Buzz, so your team can use one familiar workflow to localize product UIs, marketing visuals, and campaign assets in every language.
Once you connect your Buzz project to a Lokalise marketing project, Buzz can automatically pull approved translations into your branded templates so you can swap languages directly in Figma and see how your copy fits in real time.
In a design-stage workflow, the slowest part is usually getting good translations back into your designs. That’s where Lokalise AI orchestration does the heavy lifting.
Designers can stay focused on creating new concepts instead of resizing text boxes, while marketers own the day-to-day localization of banners, social assets, ads, and email visuals using pre-made brand templates.
Instead of you choosing a single MT engine or copying strings into a generic AI tool, Lokalise routes each string to the best AI model for that language pair and content type, applies your translation memories, glossaries, and style guides, and then scores the result for quality.
As soon as you push the copy from your designs into Lokalise, AI can generate high-quality first drafts for all target languages in minutes. Because the system understands your preferred terminology and tone, those drafts tend to read like something a human on your team would write.
The scoring layer is what ties this neatly into design-stage localization. Every AI translation gets a quality score based on an MQM-style evaluation, so you can instantly see which strings are safe to drop back into Figma for visual QA and which ones deserve a human look:
This helps teams cut post-editing effort by as much as 80% without losing control over critical content.
At the beginning, for a marketing translation, it took us on average about four weeks from request received to localization delivered. Today, we are down to less than 12 days — that's an increase in speed of 57%.
Rosario Messina
Senior Product Operations Manager at Awin
When localization runs through Lokalise and connects directly to Figma and Figma Buzz, it stops being a last-minute hurdle and becomes a way to launch better experiences, faster, in every market.
Lokalise gives your team one workflow, one source of truth for copy, and one AI-powered engine to handle the heavy lifting. You’ll spend less time fixing issues and more time shipping work you’re proud of.
If you’re ready to see how this works in practice:
The next time you ship a feature or launch a campaign, your primary language and every other market can go live on the same day, without a “we’ll localize it later” backlog.
Author
Alex is Product lead at Lokalise, a tech enthusiast, and co-founder of an amateur football league. He has helped many customers to make improvements to their localization workflows and has become Lokalise’s product guru. In his free time, he enjoys travelling, playing football, and basketball.
In Part 1, we covered why design-stage localization is a game-changing solution for agile, multilingual product development. If you read it, you’ve likely thought about what it would take to create a unified localization workflow based around Lokalise, your design tool, and localization team. You might be asking “should I use mac
When you’re expanding into new markets, your content needs to speak multiple languages. But managing translations manually can quickly turn into a logistical nightmare. This is where a Translation Management System (TMS) comes in. In this comprehensive guide, you’ll learn everything you need to know about translation management systems, starting with understanding what is tr
Behind the scenes of localization with one of Europe’s leading digital health providers
Read more Case studies
Localization workflow for your web and mobile apps, games and digital content.
©2017-2026
All Rights Reserved.
