The 17 insights we learned from Smartling's 2025 events

If 2024 was the year everyone experimented with AI, 2025 was the year expectations caught up. Suddenly, localization leaders were being asked to show impact (not just output), to explain quality in a new way, and ultimately, to turn AI hype into real results.

Across a year of events, workshops, and gatherings around the globe, we had the privilege of sitting with the companies shaping that shift. The best part? These weren’t just panels about theory. Instead, they were honest conversations about what’s working, what’s breaking, and what leaders actually need heading into 2026.

Here’s a glimpse at what we learned together.

What we learned from global companies at Smartling’s Global Ready Conference

The 2025 Global Ready Conference brought together leaders across industries who are navigating AI transformation in real time. And instead of talking about AI in the abstract, they showed us what it actually looks like inside their companies.

MongoDB talked about how global readiness is now inseparable from AI readiness. Their team walked through how LLMs are reshaping documentation, learning, and the way people—especially developers—consume content. One of the biggest takeaways: if your content isn’t clear at the source, AI magnifies that problem. In terms of takeaways, this means that investing in glossary governance and structured content is no longer optional.

IHG shared one of the year’s clearest, most impactful success stories. With more than 6,600 global hotels and 145 million loyalty members, improving their website translation process meant solving for quality, speed, and cost all at once. They shared with attendees how custom machine translation engines, Smartling’s Global Delivery Network website translation tool, automated workflows, and ongoing AI Human Translation created fast, brand-safe translations at massive scale. Their experience made something very clear: website translation is no longer a side project—it’s a critical part of the guest experience.

Meanwhile, ServiceNow brought a product perspective. Their approach to globalization starts early, weaving user research, stakeholder alignment, and AI-assisted insight gathering directly into product development. Their key message was that a successful global product strategy begins far upstream, long before the first string ever gets translated.

Later, Slator added a crucial voice to the conversation. Florian Faes, Co-founder and Managing Director, and Alex Edwards, Senior Analyst, spoke to teams and organizations operating in heavily regulated environments, where quality is a legal requirement rather than a preference. Together, they explored how AI can support compliance efforts, accelerate clinical workflows, and reduce costs, even as it still requires human oversight in certain circumstances. Their conclusion is important for any organization considering how to use AI to improve key outcomes: AI can increase speed and scale, but governance determines whether it’s usable.

Sprout Social’s session helped teams identify where AI adds value and where it falls short. Specifically, they discussed how raw LLM output still misses critical details like tone, terminology, and brand nuance, which is why it’s critical to work with a translation solution that enables functions like glossary term insertion—and that invests in training LLMs to aid in other specialized translation tasks.

Fisher & Paykel and Global10x shifted the conversation to content impact. They showed how culturally relevant content unlocks global revenue and why consistent global brand experience is a growth engine, not a localization checkbox.

GitLab, Automation Anywhere, and MIIS closed the loop with something everyone felt this year: AI is reshaping not just localization programs, but localization careers. Prompting, evaluation, and orchestration are becoming essential skills. Ultimately, their message was encouraging: while the work is changing, the future of localization still needs people.

By the end of the day, one key message stood out. It’s clear that AI isn’t here to replace localization teams, but that instead, it will amplify the impact of those who know how to use it.

What we learned from Zoom and Nimdzi about proving value

Within Smartling’s AI Translation 101 series, two key conversations helped teams cut through the hype and focus on what moves the business forward.

Zoom shared a clear blueprint for tying localization to business goals. They team walked through how they communicate value internally, how they measure ROI across markets, and why earning influence depends on aligning with stakeholders early. It was one of the clearest articulations we heard all year of how localization teams can speak the language of the business.

From Nimdzi Insights, Senior Consultant Laszlo Varga tackled one of the biggest questions teams faced this year: should you build or buy an AI translation solution? His perspective was practical and unvarnished. DIY AI is expensive, resource heavy, and risky, and raw LLM output is, most often, nowhere near ready for production or publication. Instead, modern AI translation requires orchestration, governance, and quality control. The value of this session was in its focus on what is genuinely workable for most teams.

What we learned from Volvo, H&M, and Adyen about quality

At LocWorld 53 this year, we sat down with Adyen, H&M, and Volvo Cars to talk about how to define quality in an AI-powered world.

Adyen emphasized that quality starts at the source. They described how generative AI is helping simplify content, reduce ambiguity, and make translation more accurate. They also shared how they use data to decide where human review matters most, and how geography can change user expectations.

H&M focused on brand voice. In fashion, tone and consistency are non negotiable, which makes AI’s inconsistency a real challenge. Tools like translation memory, style guides, and glossaries are critical solutions to this obstacle. Their conclusion was clear: you cannot maintain brand identity at scale without strong linguistic assets.

Volvo Cars reframed quality as user impact. In other words, quality is about how the customer experiences the content. They described how some markets require precision while others value clarity and speed. With this in mind, their team is shifting toward evaluating what content actually does for the customer, not just whether it is linguistically clean.

Together, these companies help us reframe the idea of quality in the age of AI. Rather than perfection, quality is about whether the content has achieved its purpose.

What we learned from Pinterest about operationalizing AI

Later, at LocWorld 54, Pinterest offered a clear, real-world look at what it takes to use AI translation at scale. Rather than a theoretical conversation, they showed what it actually looks like to operationalize AI for localization across one of the world’s most content-heavy consumer platforms.

First, they walked through how they rebuilt their entire translation engine around intelligent orchestration, instead of single-model dependency. Their approach uses adaptive AI, multiple translation engines, dynamic quality evaluation, and automated routing so every string lands in the right workflow. And when it matters most, human validation closes the loop with context, nuance, and brand protection.

Pinterest’s story was about scale, but it was also about influence. Their localization team didn’t just adopt AI, they became internal partners for product, marketing, and engineering by proving the business impact: faster turnaround, higher output quality, better cross-functional visibility, and automation that actually works in production.

Ultimately, Pinterest demonstrated how AI translation works at true scale, and gave teams a concrete model to learn from.

What we learned from customers across our Customer Advisory Boards

This year’s Customer Advisory Boards (or CABs), held in San Francisco and London, gave us some of the most honest and strategic conversations of 2025. Customers shared what they’re solving, what they’re struggling with, and what they need in order to use AI more confidently across their organizations.

US CAB: Where priorities came into focus

Our San Francisco CAB brought together leaders navigating AI adoption inside global organizations, and the discussion quickly shifted from theory to real operational challenges. Teams compared the pressures they’re facing, including rising expectations from stakeholders, new demands around quality and governance, and the challenge of scaling AI in a way that feels both safe and strategic. It gave people the space to step back, compare realities, and get clarity on what will matter most heading into 2026.

EU CAB: where identity and influence took center stage

Our London CAB brought together localization leaders from across Europe for a day of open conversation, shared challenges, and practical strategy. A big part of the day was simply hearing how other teams approach similar problems. 

One of the biggest themes was the need for deeper guidance around AI adoption, governance, and next-generation workflows. There was also momentum around shaping what comes next. Many teams saw the CAB as a chance to influence Smartling’s direction and ensure their needs were understood. 

Later in the day, Chris Dell, industry advisor and former content leader at Booking.com, joined to lead a workshop on identity, influence, and the strategic role of localization inside global companies. After a morning focused on AI and product direction, his session brought the conversation back to how teams define what they stand for, communicate their impact, and position themselves as partners to the business. 

What we learned from our workshop series in Paris, Copenhagen, and Dublin

Across Paris, Copenhagen, and Dublin, our workshop series quickly opened into practical conversations about how teams are approaching AI in real workflows. Attendees compared notes on what they’re testing, what’s sticking, and where they still need clarity as expectations rise internally.

In every city, the workshops reinforced the same message: people aren’t looking for AI hype. They want clarity, community, and practical ways to build smarter, safer workflows heading into 2026.

Industry dinners around the world

We also kept one of our favorite traditions going this year: bringing the localization community together over great food and even better company. These dinners provided a space for people across the industry to connect, swap stories, and unwind.

A quick shoutout to the cities that hosted us in 2025: New York, San Francisco, London, Dublin, Amsterdam, Berlin, Malmö, Menlo Park, Monterey, Paris, and Copenhagen. These nights ended up being some of the most memorable of the year.

Olhando para o futuro

If 2025 taught us anything, it’s that localization teams (and marketing, product, and dev teams!) are stepping into a new era. This time of flux provides opportunities to create clarity in the noise, drive strategy alongside other teams, and deliver real business outcomes with AI and localization.

Lastly, we are grateful to every customer, partner, and community member who joined us this year. 2026 is already taking shape, and we can’t wait to build it with you.

Join us in 2026

Smartling is going back on the road in 2026. Interested in joining us in one of these cities? Learn more and express your interest, and we’ll follow up with availability and next steps.

Want us to bring a future workshop to your city? Let us know where you’d like to see us next.