It was clear that a chatbot would give the required interactivity so we built a Facebook Messenger chatbot that included a small number of questions to help us find out all about who the customer was buying for. However, we knew that a bot on its own would not be enough to bring people into the LEGO experience and discover a world of imagination that they’d previously never been able to access. So we didn’t just build a chatbot; we built Ralph, a lovable robot with heaps of personality and one purpose- to help you find the perfect present, making you a gifting champion!

Ralph’s tone of voice was based on LEGO Movie characters that helped him stand out from other chatbots. We used warm, inviting language to make the stressful shopping experience seem carefree, fun, and informed. We further brought Ralph to life using playful GIFs/emojis throughout the conversation.

Implementation

To make sure every gift suggestion was perfect, we developed questions around price range, age, personality and interests. This meant the mammoth task of recategorizing LEGO sets, assigning them to new categories that were specifically designed around our interests and personalities. These categories provided relevant suggestions that suited the customer’s price point, child’s age, interests, and personalities. We developed a custom API that provided real time, localized stock updates from store.LEGO.com, and gave customers a unique free shipping code and gift. When customers were happy with their choice, a single tap would take them through to checkout at store.LEGO.com.

Timeline

The bot was developed over a three week period.

Placement

The bot was promoted across Facebook, with specific targeting to reach all our customer mindsets with relevant messaging that spoke to directly to their needs.

Scale

The bot was rolled out across UK, USA, Canada, Germany, France and Poland

Reach ?

The chatbot reached over 2.69 million people.

Engagement

In total, there were 1.2 million post engagements around the bot, achieving an engagement rate of over 45%. Ralph proceeded to have conversations with over 50,000 potential new customers.

Sales

The bot was responsible for driving 25% of all in-season online sales, providing a 6x return on ad spend within certain markets. In a direct comparison study with other conversion-based ad formats (such as carousels, Canvas and collections), the bot drove a 8.4x higher conversion rate and a cost per purchase that was comparatively 65% lower.

Achievement against business targets

Ralph was an experiment to simplify the purchasing funnel for low affinity customers, guiding them through a complicated landscape in an easy to follow manner. Our comparative study proved Ralph’s ROI was far beyond that of other conversion formats, and we’re delighted that he’s been hired to help LEGO customers full-time.