Forbrugerrådet Tænk - 40% Higher Conversion Without New Content or More Traffic

What was the actual problem?

Users’ expectations for fast and precise answers increased significantly as AI became a natural part of everyday life.

Our website worked well, but our search solution could no longer keep up with these new expectations.

Users were asking specific questions but were met with a traditional search experience returning many results.

This made it difficult to find the right answer.

At the same time, our member support team was handling a large number of repeated questions that had already been answered in our existing content.

Why was this a problem now?

Because user behavior had changed.

More and more people now expect to receive direct, fast, and accurate answers rather than having to search through multiple pages themselves.

Our existing solution could not meet this expectation.

As a result, users didn’t always find what they were looking for, and in some cases left the site without getting an answer.

What was the specific challenge in your setup?

Our search was keyword-based.

This resulted in:

• Too many search results
• Lack of relevance in results
• A confusing and overwhelming user experience
  

At the same time, the solution required ongoing manual optimization.

This took time, and therefore wasn’t always done to the extent we wanted.

We were dealing with a solution that both created friction for users and required internal resources.

What had you tried before?

We had previously explored AI solutions, but at the time they were too expensive, and the project was put on hold.

What were you looking for?

We needed a modern solution that could:

• Deliver fast and accurate answers
• Work across our platforms
• Be trained directly on our own content
  

At the same time, it was important that implementation was smooth and that the solution could continuously improve based on real user behavior.

Additionally, we wanted a solution rooted in Denmark (EU-based).

Why did you choose Raffle?

Raffle.ai met the requirements we had.

We also experienced:

• Strong support
• A stable implementation
• An AI model that could be fine-tuned to our content
  

It was also crucial for us that the solution could handle large volumes of content.

We have many products, tests, and articles, and this is where hybrid semantic search is particularly strong.

It enables understanding across content, connecting related knowledge, and helping users find what they’re looking for.

What has changed after implementation?

After implementing both the AI search engine and the chatbot, we’ve seen:

• More precise search results
• Faster, more actionable AI summaries
• A chatbot handling many of the most common questions
• Better insight into what users are searching for
• Close and proactive collaboration with Raffle
  

At the same time, our member support team has experienced a reduction in repetitive inquiries.

What is the impact?

Initial results show a 40% increase in conversion rate.