We worked with Unimarc to overcome the major factor hampering the customer experience: having to queue for a 'long time' to pay.After an exploration phase, we decided to create a virtual assistant that emulates a preexisting behavior: sending a friend to the queue while you finish shopping.
Virtual Queue provides customers a virtual assistant who queues while they shop. The assistant interacts with them by WhatsApp or Facebook Messenger to notify them when is their turn.
Inventures helped us to face a widespread problem of traditional retail. Its agile methodology, business focus and rigorous analyses allowed prototyping a novel solution that impacts sales and customer experience
Tomás Durandeau - Corporate Marketing Manager SMU
The design of the solution began with observation. We followed different clients in their shopping process, measuring their times and then interviewing them. The most relevant insight was that the perceived time doubles when clients were idle waiting for their turn to pay. Focusing on perception (over real time) opened a new range of possible solutions.
To inform the roll-out decision, it was necessary to quantify the impact of Virtual Queue on the business. So, we make cross-sectional and panel estimates using difference in differences methodology (DID). This gave us information (parameters and significance) of the real impact of Virtual Queue in NPS, sales per transaction, units per transaction and frequency of purchase.
The main technological challenge was to create an intuitive and friendly interface that most Unimarc customers could use. To achieve this, we developed a chatbot prototype on Facebook Messenger platform (public API) that validated the interest in the service. Then, we increased coverage with a connection to WhatsApp and added features like allowing to ask for extra time to finish up, informing the expected waiting time and communicating special offers.