Helping customers find the perfect gift through their Google Home.
CLIENT: Kmart | PROJECT: Voice Assistant
FIRST needed to do more than just simply hook up with a product database. We wanted to create an engaging and, more importantly, useful and assistive experience for customers.
Moving a leading retailer into the world of assistants
- Kmart wanted more than just an Amazon “buy product” experience.
- Kmart wanted to take the pressure off customer service phone enquiries during the busy Christmas period.
- Kmart still wanted to showcase the breadth of its products in a subtle way.
“This is an exciting new platform for us as a bricks-and-mortar retailer. We enjoyed the collaborative process we went through with the FIRST team. The first version is out in the wild and we’re learning more everyday through this new channel of interaction with our customers.”
– JOACHIM HOLTE
What FIRST proposed
As a leader in Conversational Commerce design, FIRST had to come up with an idea that solved customers’ problems. Christmas is a time for gift-giving but everyone has at least one person for whom selecting a gift is tough.
FIRST worked with Kmart to determine the best core use case for the Action and came up with a gift selection engine based on budget and recipient characteristics — adult or child and age of child.
We worked with the development team at Kmart to integrate the entire shopping feed into the Assistant. The explicit invocations created allowed implicit invocations to be surfaced in Google such as “Hi Google, I’m looking for gifts for a kid”. Over time Kmart, not Google, will become the answer to this generic question.
FIRST also included information about Kmart’s returns policies and hooked in store locations and opening hours, successfully reducing customer enquiries to the store at the busiest time of the year.
Finally we combined the two actions and were able to determine not only the appropriate gift but also whether it was in stock at the nearest store to the customer’s location.
Built on the Actions on Google platform, the solution utilises the Natural Language Processing (NLP) platform DialogFlow to provide a rich conversational experience with customers. This was powered by Google’s Machine Learning and AI to understand the intent of the user during the conversation.
An integration layer was developed by FIRST to allow several internal systems within the Kmart infrastructure to be utilised by the Google Assistant platform to allow the conversation to include useful relevant information.
A complete Google Analytics integration was implemented to allow detailed analysis of conversations with customers and allow the tracking of conversations both within the Assistant and subsequent activity on the website when the customer transitioned online.
Outcomes – what have we built
FIRST succeeded in all of Kmart’s goals. Kmart has a genuinely assistive gift finder that sits over the top of its entire product database. This is more than a buy product experience. They have voice-enabled product availability, customer assistance answers and location services fully accessible via a new medium. The fact it was delivered at Christmas time was merely icing on the cake.