Peter Kovacs is the Director of Global Expansion at Kinsta, a managed WordPress hosting platform operating at the top end of the market, providing fast and stable website and e-commerce platform hosting for over 24,400 companies across the world, with guaranteed uptime.
In this blog, Peter talks about building a global footprint from scratch, being a localization outsider, and how, for Kinsta, it’s tech-enabled humans that rule all the way.
How to Become a Localization Manager
The localization world is used to translator-turned-localization-managers, but there can’t be very many tour guides like Peter who built a full localization workflow from scratch.
“When I started with Kinsta back in 2016, I had no experience in hosting or in IT as I came from the tourism sector. Because of my background, I spoke Spanish, and that happened to be the first language Kinsta needed to translate their website. For the first couple of years, I kept both jobs, tourist guide during the day, and manager of translation projects when I got back to my hotel room. All of this, until I was needed full time at Kinsta only,” Peter laughs.
What I’ve built, together with my team, gives us the freedom to find solutions and resolve issues in ways that are totally tailored to our own needs.
Together with a colleague and friend, Peter built Kinsta’s localization solutions from the ground up, first with Spanish, and then expanding to ten languages within a year, moving to Crowdin for all of the company’s localization streams.
Peter describes himself as a localization outsider. “Personally, I’m not a fan of industry events. My focus has always been — and still is today — to build something quite unique at Kinsta, and I can’t see the same approach working for other companies. What I’ve built, together with my team, gives us the freedom to find solutions and resolve issues in ways that are totally tailored to our own needs.”
We Translate 100% Of Our Content
At Kinsta, content comes first and there are no exceptions with languages – it’s total localization and no exceptions.
“We translate 100% of our content, because we’ve learned that the minute you don’t translate something, it will come up missing later, just as you are about to launch a new feature or send out a newsletter. And if we start letting things pile on, it’d be harder to manage our workflow efficiently,” explains Peter.
Kinsta’s localization workflow includes multiple projects in Crowdin depending on the usage scenario. Alongside Kinsta’s global website and marketing collaterals in one Crowdin project, Peter and his team manage content for three distinct user groups, each with their own dashboards and tools:
- Global resellers and Kinsta affiliates.
- In-house developers.
- Customers around the world who manage their websites.
Across the board, Kinsta produces authentic, authored content and knowledge base articles which benefit not just its clients but the wider global audience. Unless hyper-specialized to a specific market, all articles are localized. Content has become one of Kinsta’s key assets, consistently topping analytics as the most popular pieces both in English and in the localized versions.
Localize your content with Crowdin
Translation Automation: If It’s Been Approved, It’s Out the Door
During the early years at Kinsta, localization was mostly a manual task, taking a huge amount of time and effort. Today, the workflow is highly automated in Crowdin, with strings added by developers going through directly to the translators and proofreaders, and back through approval, in the blink of an eye.
Many of our freelance translators have been with us since their language was first localized. We 100% trust them, as they are now experts in our products and solutions.
And translators are very much part of the team at Kinsta, Peter says.
“Many of our freelance translators have been with us since their language was first localized. We 100% trust them, as they are now experts in our products and solutions. Their work goes beyond simple translations, and becomes “true” editing to allow them to tweak and localize the copy for each specific market the best,” Peter explains.
On average, five minutes after the approval, translations are already available on the website.
On a continuous localization loop, once content has been approved, it moves into automated deployment and is out the door.
“On average, five minutes after the approval, translations are already available on the website. We have built a workflow that allows us to be really fast,” Peter adds.
The Localization Business Case
Today, Kinsta’s nearly 300 employees are spread all over the world, and the headquarters has moved from the UK to the US.
A global workforce reflects Kinsta’s presence across the globe. “Traditionally, most of our clients came from the US or other English-speaking countries, but we have now achieved a healthy 50/50 split between English and non-English speaking markets,” says Peter.
As the Director of Global Expansion, Peter enables growth through reaching local markets in their own language.
Cutting their teeth with the early Spanish trials and errors, Peter and his colleagues have tailored a unique localization solution for Kinsta. Today, the suite of languages includes English and Spanish, but also Portuguese, French, Italian, German, Dutch, Danish, Swedish, and Japanese.
Sometimes, there is just not enough return on investment, and you have to sacrifice a language or two.
But localization also always needs a business case, and sometimes that just isn’t there. Along the way, Finnish and Norwegian were dropped from the localization program.
“Sometimes, there is just not enough return on investment, and you have to sacrifice a language or two. Finnish and Norwegian were those languages for us. Instead, we find that English works really well in these two markets, so we still have a good presence there,” explains Peter.
Tech-enabled Human Translation All the Way
For Peter and his team, localization is all about tech-enabled human translation.
In our future, I see more translators and SEO experts joining our team to help us grow even more by allowing us to take advantage of more advanced tools.
In Kinsta’s Crowdin, machine translation is enabled for translators to make use of MT suggestions where it makes sense to them, with the needs of the content leading their choices.
Some content is also pre-translated with DeepL’s MT engines, but even then, always post-edited and proofread by native translators.
“In our future, I see more translators and SEO experts joining our team to help us grow even more by allowing us to take advantage of more advanced tools. But even when we use tools, the last word will always still be with the translator – they are the ones we trust, the ones we know, and the ones who understand our solutions,” Peter summarizes.
At the end of the day, it’s all about the brand, the solutions, and the company’s credibility. For Kinsta, that’s achieved with top-of-the-range technology with Crowdin and machine translation, but led by expert linguists.