Skip to main content
To the homepage of Knowit

CASE

ARLA

How Arla Uses AI to Scale Global Food Inspiration – Without Losing the Local Flavour

RETAIL & SERVICES

Arla inspires consumers around the world with thousands of recipes developed by their own chefs and editors. But when you run a global food universe across six continents, translation becomes much more than words. It’s about flavour, tone and culture. Together with Knowit and Google, Arla developed an AI Translation Service that enables recipes to be scaled globally – quickly, efficiently and with full respect for local traditions.

    AI
    Business systems

CHALLENGE

From global inspiration to local flavours

How do you translate “koldskål med kammerjunker” into English without ending up with something like “cold bowl with cookies”? 

For years, Arla translated recipes manually. The process was time-consuming, expensive and required local editorial resources in each market. At the same time, the need to publish more content faster continued to grow without compromising quality or cultural relevance. 

For a global brand like Arla, the challenge was never just language. It was about preserving the flavour and soul of their content whether it was read in Denmark, France or the Philippines. 

SOLUTION

When AI and editors work side by side

Together with Arla and Google, Knowit developed an AI Translation Service, integrated directly into Arla’s own Editor Interface. 

When an editor wants to publish a recipe in a new market, the AI can translate the entire recipe – title, introduction, instructions, SEO text and ingredients based on brand parameters and local guidelines. 

Ingredient lists are always pulled from Arla’s internal ingredient database, ensuring consistency. A tomato will always remain a tomato. 

After the translation is generated, local editors review the content, refine the tone and adjust cultural references. Their corrections are then used to continuously train the model, allowing the system to improve over time. 

Recipes cannot be published without local editorial adjustments. For Arla, it is essential that AI never dictates the content it supports it. 

The result is a collaboration between humans and technology: the technology handles complexity, while editors preserve creativity and cultural nuance. 

TECHNOLOGY & PROCESS

Technology & process

  • Technology: Google Cloud Services, Machine Learning, Natural Language Processing
  • Integration: Directly integrated into Arla’s Editor Interface
  • Security: Data handled within Arla’s existing cloud infrastructure
  • Sustainability: Reduced reliance on external translation agencies and manual workflows
  • Future-proofing: The system can scale and improve continuously across markets 

RESULTS

Global scale, local flavour

The AI solution enables Arla to: 

  • Spend more time on local quality and cultural relevance 
  • Publish recipes faster and more efficiently across markets 
  • Reduce costs related to translation and content production 
  • Maintain a consistent tone of voice globally 
  • Improve organic visibility through better SEO optimisation 
  • Build a stronger digital foundation for future technologies 

Internal measurements and operational insights show significant improvements in the daily workflow: 

  • Time-to-publish reduced from days to minutes 
  • Localization costs reduced by up to 60% 
  • Editorial time spent on translation reduced by more than 70% 

The AI Translation Service has made it possible to share global food inspiration at an entirely new pace – without losing the local flavour. 

In partnership with Knowit and Google, we have turned a complex business challenge into a scalable, AI-driven solution that transforms how we translate and adapt recipes across markets. By combining technology, expertise and creativity, we have built a system that respects cultural nuance, SEO performance and brand voice — reducing time and cost while improving quality and consistency. This collaboration empowers our local editors to focus on what they do best: adding the final local flavour and bringing food inspiration to life through high-quality content.

Thomas Jeppesen

Global Head of Owned Media, The Barn

THE FUTURE

Humans and technology working together

Arla is now expanding the solution to additional markets and exploring new opportunities for AI-supported content production. 

The goal is not to replace editors – but to equip them with better tools that allow them to work faster, more creatively and with deeper insights into the audiences they inspire. 

AI can do a lot. 

But it still can’t taste koldskål. 

Contact us

Would you like to learn how AI can help you scale global content without compromising quality or culture? 

Anne Fink

Client & Business Unit Director

Interim