We need to talk about words; terminology management at Google

We need to talk about words; terminology management at Google

Around the world in 7 localization interviews 

Hello reader! This interview is part of the series around the world in 7 localization interviews, in which I’ll talk to the world’s top figures in localization and try to understand their businesses.

If you would like to get notified when new interviews are published, or get the final ebook compiling their advice, you can sign up here. 

(Alternatively, we help businesses deliver on incredible localization strategies by providing the insight they need in 190 markets. If you're struggling with finding real users to test your software from either a qualitative or a functional standpoint around the world, we can leverage our crowd to put you in touch with the right person.) 

Natalia Garcia Martinez, Terminology & Content Program Manager at Google 

Although localization is not translation, it's hard to get away from words. Words are the smallest individual units in software to be localized, and together, the most significant. A business like Google translates hundreds of thousands – maybe millions – of words.

So I supposed it made sense for a business like Google to hire a Terminology Manager, a job title I confess I hadn't heard before meeting Natalia Garcia Martinez, the Terminology & Content Program Manager at Google. 

Natalia works in Google to manage, optimize, and refine the wordset they put out to international audiences. It was so interesting to hear how things worked, and what other businesses can take away from her experience working there.

What is Terminology Management?

‘Terminology management is a dedicated team or operations infrastructure designed to manage terminology at a global level,’  Natalia explained. ‘When a product is going to launch globally, or is global, part of the “naming” strategy within marketing is moved to the international market decision level.’

‘From an operations perspective, we’re focused on managing and maintaining the intellectual property of the company’s imagery through language in international markets.’

So practically speaking, that means ‘we’d identify the content types that we’re going to translate – whether that’s a helpcenter component UI video, or any number of other things. Based on those tests, we’d extract our list of keywords from a source language. And then we would localize those terms following a number of best practices, like knowing the descriptions, context, etcetera. What are the keywords that build the brand of your product? Or the image of your product?’

That’s best practices, glossaries, and a terminology infrastructure: thousands of perfectly selected words arranged to the IP infrastructure for Google’s enormous suite of products and services.

This is a helpful exercise for most businesses, Natalia argued.  After all, there’s the general operational timesaving, ‘because we have everything pre-translated.’ There’s an iterative benefit against customer metrics: ‘once you see how users react to that content, there’s an ongoing practice there to identify whether terms need updating.’

And finally, there’s the consistency ‘across all products, content types, and the brand.’ 

An unspecified brand value lurking behind “consistency” there. But it might also apply to language consistency between product navigation stages – drive desired behaviour, like using the checkout. This is the kind of thing that Global App Testing can run QA for.

When do you need terminology management? 

‘As soon the need for localization arises – that’s when you should start having terminology management. It doesn’t need to be a big team. It probably starts as part of the role of localization program manager.  Any translation company would also support, or sometimes require, the provision of the list of glossaries.

But it’s an investment that pays greater dividends when made earlier. ‘But you need terminology management early on.’ 

Want to drive your global growth in local markets?

We can help you drive global growth, better accessibility and better product quality at every level. 

Chat to an expert

Terminology vs Scale: the challenges working at Google

Guidelines, glossaries, and best practices create a problem. 

Google is a company of over 150,000 employees. In a world of thousands of competing internal initiatives, how do you guarantee anybody listens? I had wanted to ask Natalia about stakeholder management – absolutely crucial in a cross-functional team, or a team the size of Google.

‘The volume at a scale like Google is really really massive,’ said Natalia.

To some extent the work of the terminology team is hardwired into the translation process. ‘The tools guarantee the linguists see [our work],’ said Natalia.

But there is still a great deal of thought required to ensure that terminology practices get implemented. ‘The most important thing at Google is work at scale. We put several approaches on the table at the same time. 

‘The first is self-service. We have some processes which are automated. We have some processes as part of the overall localization playbook – where stakeholders can go, and know that…”I find the information here, these are the processes I need to follow, this is what I need to submit” This is the most transactional part.’

‘The second approach is partnering with the teams that have a direct influence over the main decision makers, and that's usually marketing. Marketing have their own playbooks – so embedding localization there is an incredible way to reach wider audiences. 

‘The third approach is getting the support with, and of, any project owner. So we start seeing what our key products are based on. What's the priority for the organization? We identify a main point of contact. Those are going to be our sponsors.’

The reality is that it’s difficult to influence people. 

‘I feel localization in general is perceived as a commodity’ said Natalia. ‘Not at Google’, she added. But for most companies, ‘I think it’s very difficult for people to think of global terminology first. When content creators think of what they’re going to produce, they do it in their own language.’ 

When nobody engages, argues Natalia, that’s when the penny drops that ‘“there’s something we should have done earlier” – and that’s when the best practices for terminology come into place.’ 

‘Usually you find a sponsor there that really understands because of the implications and and all the challenges to rethink how our content is translated. So it’s usually about raising a lot of awareness, presenting a lot of case studies.” 

I asked Natalia about AI, but I was surprised to hear she didn’t agree that was a major commodifying force – at least not for terminology. ‘Machine translation is benefitting a lot from the work done on terminology management to train the engineers and to improve the results – to customize those engines to specific types of content.’ 

‘I think terminology now has a bigger spot in the overall [translation] process that it did before – we invest in it precisely because of the volume and scale and pace we need to get content done.’  Natalia pointed out that AI will mean that you can apply this work into multimedia as well very shortly. It’s now possible to dub video in the voice of the original actor in a new language in such a way that their lips move to the dubbed script.

‘I think it’s going ahead with the expected curve for technology integration,’ argued Natalia. 

Translation mistakes you’ll make 

In a previous life, Natalia worked for a language services provider TransPerfect where she had worked as a Director of Quality. My final set of questions was about the way that clients engage with LSPs.

What mistakes do clients make? 

‘The one thing I would see recurrently is the assumption that a localization services provider is going to know everything without actually getting any information from the client. There’s an expectation… that a linguist would know what they’re going to encounter, how to translate it.’ 

‘Run a discovery phase, says Natalia, when you work with a third party. It will come out in the quality of the translation. 

If you want to save money? Invest in the original materials. ‘Style guide, glossary, identification of content types, formats, context, audience type,’ listed Natalia. ‘That’s usually one of the biggest mistakes I’d encourage everyone to invest time on.’ 

But the biggest difference between clients is the level of localization sophistication. I was surprised to hear how big the difference was.

‘It could be very difficult to have some conversations’ said Natalia, ‘because you get some teams focused on the most traditional side of things.  You get teams coming and saying “we need to launch in market X. We understand brand and tone of voice, etcetera – we want decent translators.” They understand tone of voice – but not products, development and release cycles, or anything else from a best practices standpoint.’

‘Even things that I would consider … simple considerations, such as forecasting. It’s not the same thing to localize a product with string management, or to translate a brochure where you can’t see what it looks like.’ 

Had Natalia’s view on LSPs changed since moving client-side? Natalia laughed. ‘Yes. I’m aware now there’s reasons a client might not get the material to you.’ 

I think’ said Natalia, and with the exception of Google, LSP use is ‘an absolute need. If localization is not your core business it’s too hard.' 

‘I think that they have a really really good advantage in that sense. in terms of providing technology in terms of providing expertise… and having a big perspective on how different clients work and operate.’

‘If you really want to produce large amounts of content volume in all the languages, it’s going to be very difficult to sustain that internally. So you need to really understand what your resources are. And to make the most out of those resources, which might not be having [an LSP] localizing content, but thinking of how the shape of that market is going to look in a language, or how we’re going to drive more localization adoption, best practices adoption. 

‘My biggest challenge today is scale. Even when I think I’ve figured something out, I know who to talk to. We have incredible processes that run smoothly – but when something new comes up, you realize the scale has increased since the last time that you put that process in place. 

‘Google is constantly innovating and evolving direct system products and creating new products. And there are new teams coming up all the time. There is a big challenge there. And also keeping in line with priorities of the company as a whole.’

Enjoy the interview?

Try these:

👉  L10n pipeline optimization with Pleo 

👉  i18n teambuilding with Shopify 

👉 L10n leadership with Deliveroo 

👉 L10n strategy with Pinterest 

P.S. We help you execute faster localized releases and generate smarter localization intelligence with access to a crowd of testers and panellists around the world