Discover more of Wales with personalised recommendations

The Challenge: Inspiring Every Type of Visitor

When people visit a site like Visit Wales, they’re often looking for inspiration: where to stay, things to do, and how to make the most of their trip. However, the problem is that everyone is different. Families might want activities which are child-friendly. Solo travellers may be seeking adventure or looking to stay in a quiet countryside location. If the site can’t respond to those unique needs, users lose interest fast.

Research backs this up. Most people now expect a personalised experience online, and they get frustrated when they don’t get one.

So, the big question for Visit Wales was: how do we take the rich data we already collect and turn it into really valuable recommendations for visitors?

Data as the Foundation for Personalisation

The raw data was there, but it needed to be unlocked. Using Drupal as the platform, behavioural information from Google Tag Manager and Google Analytics was channelled into BigQuery, giving us a clearer picture of how people moved through the site.

Patterns quickly emerged: families often explored accommodation and activities together, event seekers followed distinct paths from inspiration to booking, and new visitors behaved differently from returning ones. These insights became the foundation for building personalised journeys.

Building Personalised Journeys with Machine Learning

Working with iCrossing, we developed a machine learning model designed to recommend the “next best step” for each visitor.

Here’s how it works:

• If you’re browsing family-friendly accommodation, the system also suggests family-friendly activities nearby.
• If you’re new to the site, recommendations are guided by the journeys of similar past visitors.

These recommendations feed directly into Drupal, appearing at the bottom of product pages. Instead of reaching a dead end, visitors are encouraged to keep exploring — whether that means discovering events, places to stay, or activities to try.

We’re also planning to extend this approach to article content, so users will be guided to their next inspiring story or travel guide.

Seamless Integration with Drupal

Because the recommendation engine is surfaced through Drupal components, it integrates seamlessly with the site design and editorial workflow. Content editors keep control, while visitors enjoy a more tailored, engaging experience that feels natural within the platform.

The Results: More Views, More Engagement

The difference was immediate. In August 2025, Event pages recorded nearly 6,000 additional visits compared with the same month the year before — a 30% year-on-year increase.

Visitors weren’t just browsing more pages; they were finding relevant, inspiring content that kept them engaged and connected to Wales.

Long-Term Value: Growing Loyalty and Return Visits

Personalisation also pays off in loyalty. Google’s DAU/MAU metric (which measures how many monthly users also return daily) rose from 20.7% in August 2024 to 21.6% in August 2025.

It may look like a modest increase, but in real terms it represents thousands more people choosing to come back. That shift shows that when content feels personal, visitors are more likely to return — not just once, but again and again.

This project is just the start of a new wave of personalised features we’ll continue developing with the Welsh Government.

What’s Next: Expanding Personalisation Across Content

The next phase is to bring the same level of tailored guidance to article content, helping users discover more stories, travel tips, and inspiration. This will make every part of Visit Wales more personal, ensuring no visitor journey ends prematurely.