Chatbot

Evolving Chat/AI trends have changed the way banks show up for their clients.

Project Goal

    1. SBU Workshops to uncover use cases, current solutions and pain points

    2. Interviews with LOB partners (chatbot POs) to understands potentials areas of improvements

    3. User interviews for internal chatbot users

    4. synthesized arch integration diagrams with the selected vendors

    5. Market research for UX best practices for Chatbots

    6. Design recommendations to better chatbot use

    7. Competitor research on capabilities and solutions that provide good experience

    1. Architecture and Technology diagram (data flow, data pipeline, tech components)

    2. capability matrix for vendor comparison

    3. Chatbot Training

    4. Current vendor capability map

    5. Dashboard & Reporting capabilities

    6. Chatbot programming

    7. Will vendor solutions work with our current tech stack

    1. A well-researched and articulated POV on the current and future state of chat bots; including the following considerations

    2. Features and Capability Map

    3. KPIs & Data Metrics

    4. Providing the landscape of the chatbots stakeholders at the bank and bridging gaps to reduce silos

    5. Chatbot management and maintenance practices

    6. Current cost and saving for different Chatbots

    7. Chatbot use cases

Provide a clear understanding of CIBC’s current chatbot solutions, including chatbot capabilities, technology and pain points.

Design and demonstrate art of the possible enterprise solution for CIBC

Primary Research

- Common frustrations

Chatbot doesn’t understand the questions or has very limited access to answers (data). This makes for a terrible bot experience for users. Make sure to set up your bot in a way that it can answer questions.

- Common frustrations

Your chatbot has no personality – it answers dully and inflexibly like a machine. That’s neither the best customer experience nor the best impression for your brand.

- Common frustrations

There is no handover protocol. For some apps, like WhatsApp, it’s required to have a handover procedure from bot to human agent. However, it always makes sense to set up a handover process that leads customers to human agents that can help them tackle more complex issues.

- Common frustrations

The dialogue only follows rigid patterns, but cannot do anything with differently asked questions or comments, and is then stuck in an endless loop. Find happy flows that will smoothly allow your bot to ask for more information in different ways.

UX Best Practices

  • Reduce redirection: ability to answer maximum user queries within the chat interface of the chat window instead of redirecting to external webpage

  • The tone of voice: Chatbot Dialogue aligned with the Target audience. Anticipate User Behaviour and needs Predict and pre-empt user behaviour

  • Topic-led conversations: encouraging customer reflection, in contrast to task-led conversations, aiming for efficient goal completion

  • Free text interaction, users interact mainly using their own words, rather than button interaction, where users mainly interact

  • Focus on Client Experience Chatbots need to be researched, designed and developed with the aim of strengthening user experience.

  • Data-driven framework for customer service chatbot that utilizes the past customer behaviour data from clickstreams

  • Easy to understand Chatbot UI the font end design - how the user navigates through the page. The data is stored once the conversation has ended.

  • Make interaction human-like by increasing wait time between messages. Provide smart suggestions with the help of buttons or quick replies

AI - Capabilities that are Required

Sentiment analysis​
Determining the sentiment behind a phrase​

Summarization​
Expressing the most important facts of a text​

Semantic search​
Applying user intent the meaning to find the right content ​

Question answering​
Answer questions posed in a natural language​

Speech recognition​
IA machine is able to process speech audio​

Machine translation​
The translation of one language to another by a machine​

Insights

According to our research, companies that offer the greatest chatbot experiences have the following characteristics that set them apart from the competition: they are simple to use and navigate, with obvious clickable buttons. An interactive experience is crucial for keeping users engaged, which benefits the organization. Having the ability to comprehend various language structures, detect information while being capable of analyzing duplicates and other user expressions, and provide users with findings that are both particular and logical.

Frustrations

When the chatbot can't comprehend what the user wants or is attempting to say, the user becomes frustrated by the time it gets to the agent. It can be annoying for the user when they have to keep looping in order to reach a human agent and frequently the chat system shuts down in the process. Becoming stuck in an unending cycle while speaking to a dull robot.

Good user experience for the chatbot
Focusing on providing the client with their best experience in terms of navigating to the chatbot and engaging the chatbot - this in turn benefits in retaining users in a way that they would want to return.
Users don't enjoy the chatbot's ability to anticipate their wants by using a combination of wait time in answering the user, open-ended words and button interactions that lead to different pages and websites. hence decreasing the need for it and making the user's engagement with the AI feel human.

How can chatbots be implemented well?
Instead of diverting them to human agents, uncover genuine value for the user by using them to fill in the gaps left by the user.
To leverage the demands of the business, chatbots must be employed in a certain way.  Successful chatbot implementations are correlated with using chatbots to transform business use cases rather than using chatbots to just utilize AI in general.

What clients require and how we can provide those needs should be the main points of focus.
If a consumer has a problem, they will call customer support. If they are in a place on an app or website where they need to contact customer support and the majority of problems are encountered, implement chatbots to make it seamless so they won't need to be frustrated when they meet the customer care representative providing contextual help: Instead of limiting a user's ability to contact customer service, chatbots could offer the needs.

LoBs In Focus

  • Currently have a chatbot powered by 24/7 AI

    Current chatbot is not a good experience and isn’t bringing enough value

    This could potentially be a high-value use case for implementing a better chatbot/scaling Watson

  • In 2021, there was a proposal (done by IBM) for a chatbot for Contact Centre employees for mortgages. Not Approved.

    NSL chatbot work could potentially be leveraged

    There’s an opportunity for client-facing chatbots since there’s no way for clients to access help outside of business hours

  • Knowledge Central is a Sharepoint site where contact centre employees get quick answers to serve clients

    Mark Parker is working on NSL chatbot for KC

    While smart search has improved the experience, there’s an opportunity for integrating a chatbot

Opportunities for AI capabilities

Previous

Brand TLD