As a pan-European, multi-location business, ecoATM needs to list, manage and optimise a large estate of individual locations across the Google Business Profile platform.
Our aim was to create a virtual agent proof of concept (POC) to communicate with prospective customers across ecoATM's numerous Google Business Profile locations to ascertain the lead generation performance of the channel.
© ecoATM is the market leader in mobile phone recycling and resale, they offer a simple, fast, rewarding, and eco-friendly way to dispose of unwanted phones with a number of smart kiosks deployed across the UK in supermarkets or shopping centres.
Motif partners are a marketing analytics and growth consultancy.
© We were approached by Motif Partners to develop a live functional proof-of-concept for ecoATM on the Google Business Messenger platform.
The virtual agent was to be created on the Google Business Messages platform and Dialogflow CX conversational AI platform to illustrate the potential use case of conversational AI in the digital customer journey.
The agent needed to satisfy the following use-cases:
1. Provide a price offer
2. Answer a range of FAQs
3. Provide information about ecoATM
4. Provide In-depth information about how the ecoATM process works
5. Email signup
We were required to carry out conversation design and create the Google Business Messages agent and supporting Dialogflow agent along with the required business logic middleware to provide the pricing service that the chatbot would consume to calculate a phone price for a user.
The agent was to be deployed onto Google maps on a subset of locations.
Motif partners had a very clear idea of the conversations which the virtual agent needed to support.
We carried out a number of fact-finding sessions with Motif partners and collated all the necessary information for the use cases.
Our conversation designers worked through the rough conversation flows on an interactive conversation design platform: Voiceflow. At this point, we could create a working prototype. This hands-on, realistic version of the final product enabled us to get rapid feedback and iterate quickly on the more complex flows such as "Price your phone" and the faq drill-downs.
After getting client feedback we updated the design and released a new version to ensure the conversational flows were as engaging and efficient as possible.
Once the conversation designs these were signed off we were then able to develop the virtual agent on the Dialogflow CX platform and develop the required middleware to support the Google Business Messages and Dialogflow agents. The solution was then thoroughly tested and client feedback was taken into account and applied in a series of iterations to polish the final result.
The agent was then launched on a number of client locations on Google maps.
We created an intelligent which is able to answer questions about the ecoATM business and provide 'price my phone' functionality where a user is asked a series of questions before being provided with a price for their smartphone.
Agent responses were sent back from Dialogflow to Google Business Messages in the correct format so the platform could provide a rich conversational experience providing rich media carousels, knowledge cards, suggestion chips and buttons.
Google Business Messages has native, secure integrations with Google Cloud services to allow for business logic creation. We created middleware using Google Cloud Functions to support intent fulfilment and to handle webhook calls from the business messages agent with everything secured with Google Identity and Access Management.
The live experience is functional with a high-level of rich functionality (cloud functions, rich media carousels, knowledge cards) that effectively showcases the early potential of the medium. I'm pleased with what was achieved in the time.
Director, Motif Partners
The solution was designed, developed and launched on time and we will be gathering data on usage.
Google Business Messages is a great way to drive customer engagement when organisations have multiple locations.
There are a number of improvements that could be made to the virtual agent down the line to improve the chat experience: including personal context messages, location-specific information and live agent handover.
I like working with small teams and deep subject-matter experts. Working with Adrian was an informative and enjoyable experience. Adrian certainly knows his stuff when it comes to conversational AI. The project did call on Adrian's creative problem-solving on several fronts which I think is a strength of his given several constraints we faced with Google and working around the client's internal systems and processes. I appreciated the creative approach because speed and lean focus are the order of the day for POCs.
|This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Analytics".
|The cookie is set by GDPR cookie consent to record the user consent for the cookies in the category "Functional".
|This cookie is set by GDPR Cookie Consent plugin. The cookies is used to store the user consent for the cookies in the category "Necessary".
|This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Other.
|This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Performance".