Localization Tips & Product Updates

When most companies think about localization, they picture a translation team tucked away in a corner, handling strings after the real product work is done. Gideon Hod sees it completely differently. At Turo, the peer-to-peer car-sharing platform operating across five markets and three languages (with more on the way), localization is a product operations competency that keeps engineering moving fast and users happy worldwide.

The web was built page by page. A homepage, a product page, a blog post — each designed, written, and published individually to one primary channel: the website. And for a long time, that worked. But the digital world has changed: it's faster, it’s multi-channel, and now it’s AI-search driven. Content doesn't live in one place anymore. It's distributed across websites, apps, social media, AI search engines, and more. The brands showing up well across those channels aren't just publishing more — they're publishing smarter with content structured so any system, human, or AI agent can understand exactly what it is and what it's for.

The headline updates are custom instructions and additional context for AI pre-translation, AI Pipeline presets to help you deploy complex workflows in seconds and major enhancements to our Crowdin plugin for Figma. Beyond this, we’re excited to announce that Crowdin Enterprise is now available on the AWS Marketplace. We’ve also added category-based proofreading rates for better budget control and a highly requested auto-save feature in the Editor. Here’s a closer look at everything we shipped in April.

Expanding technical documentation into new markets sounds straightforward. In practice, it breaks down fast when technical documentation teams treat localization as an afterthought, tacking it onto the end of a content workflow. When content is written, approved, and then handed off for translation, it creates duplicate effort, inconsistent terminology, and increases translation costs. The more technical content you produce, the harder it is to keep content aligned across languages. If you are managing content for highly regulated markets, the challenges compound. Localization isn't a task you throw in at the end. It's a content strategy decision that belongs in the content lifecycle from day one. When it isn't built into the content lifecycle upfront, even well-run documentation teams struggle.

Translation for video is a special case: the audience doesn’t just read the words – they read them while the image moves. Titles, supers, on-screen text, end cards, trailer lines, in-game cinematics, marketing spots – if it appears in the frame, in time, the rules change. That’s a different job than localizing a website. In a video, even a good translation can still be wrong on screen – because the problem isn’t just the wording. It’s keeping language locked to the edit as motion changes, timing shifts, and last-minute copy arrives.

Snov.io is a sales automation platform used by over 3 million users to find, verify, and engage with leads. Snov.io currently supports 9 languages on its website and 5 languages within its platform, with some high-traffic pages translated into as many as 14 languages. Behind this massive operation is Olena Holovach, Snov.io’s Localization Manager. Starting 6 years ago as a localization pioneer, Olena has transformed the company’s approach to localization into a highly structured, AI-driven workflow.

For as long as most of us can remember, the localization industry has run on a single, stubborn unit of measurement: the words. You count them, you price them, you invoice them. Simple. But what happens when a large language model can draft thousands of those words in seconds, and a post-editor can process them by the hour? That’s the uncomfortable question at the heart of a recent conversation on The Agile Localization Podcast, where host Stefan Huyghe sat down with Diego Cresceri, Founder and CEO of Creative Words, to talk about what comes after the per-word era.