[24]7.ai Conversations
Here’s how I partnered cross functionally—with product managers, developers, business executives, conversation designers, client managers, and more—to shape the product strategy, processes, and end-to-end experiences that went into creating the industry-leading platform for designing, deploying, and managing omnichannel, AI-powered chatbots: [24]7 Conversations.
Users and Clients Say...
- “[24]7.ai Conversations is efficient! It will save us time in communication, modeling, debugging, designing, and building.”
- “Single source of truth and integrated! It connects all of the tools together, allowing collaboration in the same place, and preventing duplicate work.”
- “I love the aesthetics! It’s sleek and clean. The visual flow helps to see contextual paths and is easy to use.”
This is a testament to the rigorous process requiring manpower, method, and purpose, that transformed our complex operations of AI chatbot creation into a simple one.
Here's how I did it.
Align Chatbot Builder Function with Company Mission
From the start, I made sure our product KPIs connected to the company's mission. This instilled purpose in my efforts, drove the vision, aligned my thinking, and gave me measurable goals. It unified our business, tech, and UX organizations into one team with one common mission.
Outline High-level Client Workflow to Include Chatbot Optimization
I conducted user interviews to determine the high-level workflow of our clients’ operations—from consulting their business requirements all the way to optimizing their deployed chatbot.
Map the Previous Chatbot Building Operations
I thoroughly examined (interviewing practitioners, observing collaborative interactions and tools, etc.) how [24]7.ai previously executed the chatbot platform building process. I produced a user journey map that highlighted the steps, tools, and various types of expertise required. The laborious, manual, and time-consuming former process combined [24]7.ai products, processes, and people with numerous dependencies, gating procedures, and dispersed tools.
Not surprisingly, this caused deployment delays and appeared to our clients as a slow process that failed to service timely business and customer needs.
Measure the Velocity of Chatbot Operations
I measured the velocities of each task, workflow, tools, and expertise. I analyzed real app deployments and interviewed practitioners and client managers to quantify process times. Then I simplified the operations into basic phases.
Set Chatbot Building KPI goals
Post-UX explorations, technology assessments, and other predetermined factors helped me project our KPI goals. I've placed it here to compare with our old operations velocity above. You can see the improvements I was aiming to achieve. Formerly months-long processes would now be completed in days or hours. Although this was ambitious, it gave me motivation to challenge myself and shoot for the stars.
Determine Chatbot Platform Operational Problems and Needs
After analyzing and interpreting our current operations, identifying pain points, gaps, roadblocks, and dependencies, I defined our problems and user needs.
Building a Chatbot Builder: Problems
- There is no easy way to build an app with the features we need in a self-serve, independent, and quick way.
- We must rely on other teams to build a chatbot across time zones, which is an expensive and time-consuming process.
- There is no easy way for us to determine if the application we build is working well.
- There is no easy way to integrate with other applications since everything is in a silo, making it difficult to manage.
Building a Chatbot Builder: Needs
- Enable chatbot builders and conversational designers to independently create, test, and deploy chatbots on their own to reduce effort and time.
- Enable chatbot builders and conversational designers to easily preview, detect issues, and optimize.
- Enable chatbot builders and conversational designers to design conversation flows, dialogs, logic, and models in one place.
Define Chatbot Builder User Segmentations
I analyzed our user segmentations to determine which ones highly impacted our KPIs. I also examined our client organizations to determine which segments would use our products and services. I realized the conversation design process was meaningfully extensive, prompting me to optimize for this practitioner. Through our client user research, we also found that customer service experts and generalists were required to fulfill all necessary chatbot building tasks.
Our Vision for Our Client’s Chatbot Creation Process
-
A self-serve, integrated conversational flow diagram design tool that enables users to create conversational AI chatbots efficiently on their own.
- Design a conversational flow with abilities to sequence, stitch, and build dialogs with rich content and interactions.
- Automate the process of creating natural language and intent classification models when users create conversations.
- Instant and automated app logic creation with capabilities to preview, test, and deploy the chatbot.
- Complete self-serve, self-managed, all access, with no dependency on other expertise, or waiting for other associates to complete their work.
- Single source of truth, one interface that manages all components of a project, with multiple user access and collaboration.
Unify Business, Technology, and Design
I brought together different types of expertise from various practices, so we collectively understood all the problems in creating a chatbot development platform, as well as the potential solutions. I conducted two Agile design sprints within two years of each other, leading to knowledge sharing, product alignment, and design prototypes.
Rapid Prototypes
I created rapid prototypes as the output from design sprints, ideation workshops, and hackathons. I used the prototypes to guide our product strategy and to build a real product in the Agile sprint development process.
Anatomize the Conversational Chatbot Design Workflow
I made it a mandate to directly connect to the hip and heart of our primary user base: Chatbot conversation designers. Every design effort required user interviews, consultations, concept testing, and usability and beta testing on everything from high-impact features to minuscule functional improvements.
I analyzed our chatbot conversation designers’ Jobs-To-Be-Done (JTBD), the tools they used, and the workflows for designing a conversational AI chatbot.
Define Principles of the Chatbot Designers’ Experience
-
Testing analysis from the design sprint prototype, and the insights gained from our users, proved to be key product experiences that ensured acquisition, adoption, and retention.
- Spatial relationship and readability of flow: relativeness, interconnections, sequence, visual organization, and hierarchy.
- Visual signification of node types and rich content.
- Intuitive interactions to navigate around, follow flow paths, and expand journey branches.
- Creating in the moment, in context, helps determine flow, dialog, and grammar considerations.
Examine Competitive Chatbot Landscape
I examined our competitor service and tools. All vary in user base and focused on technical practitioners: developer-centric functionalities and workflows, with no graphical interfaces. I wanted to be more simple, provide clear separation between experts, visualization, multi-user/multi-environment, enterprise-grade lifecycle management, and integrated multi-modal and channels.
Create Chatbot-Building Information Architecture
I created flow diagrams, user journey maps, user stories, and wireframes to illustrate the workflows, motivations, tasks, high-level flows, site maps, and features. This helped me align the technical and business requirements with the product stakeholders. I focused on holistic product strategy, core functionality, and kept it high level.
Execute a Phased Agile Approach to Reality
In collaborations with my PM, together we defined the MVP through a series of phases and product releases that ultimately led to our full-fledged mature product vision. We prioritized our feature roadmap based on scope, client & user needs, business opportunities, and feasibility of the chatbot building capabilities, interface design, and technology constraints.
Detailed Designs Along The Way
Through the Agile development process, I began detailed design: micro-interactions, workflow & feature user interviews, concept testing, technical consultations, implementation, QA, and beta testing. To see how the design has evolved through releases, refer to the article: The Paradigm Shifts of AI Chatbot Designing.
Where We Are Today
AI chatbot development
Conversation flow designer
Chatbot response editor
Calculate and Measure ROIs and KPIs
The manual development processes for building a complex chatbot app's SCXML, QA, APIs, model creation, architecture, and project management took roughly 315 hours (an estimate quantified by previous chatbot deployments). The newly designed tool automated and streamlined these processes through new architecture and interfaces, dramatically reducing the development time to 48 hours (measured by a real client deployment).
This is an 85 percent decrease in time to build.
The previous deployment process for generating, testing, and then publishing a fully interactive chatbot app to the client's website initially took four weeks. The newly designed tool automated and streamlined these processes through new architecture and interfaces, reducing the deployment time to 15 minutes at the most.
This is a 99.9 percent decrease in time to deploy.
Once a chatbot is deployed and containment rate is analyzed, a designer needs to enhance the conversation, which previously took eight weeks to increase the containment rate by 8 percent. With faster build and deploy times, a designer can reach the same containment rate increase in one week.
This is an 87 percent decrease in time to boost containment rate.
I calculated client monthly spending on professional services, which provided internal practitioners to build, design, and deploy a chatbot for them. The migration and adoption of [24]7.ai Conversations mitigated the need for professional services as the tool automated most of these processes and workflows. This contributed to a 50 percent cost reduction in client spending, amounting to tens of thousands of dollars in savings.
Calculate Our Potential Improvements
Users still faced significant obstacles and delays including ramp-up and training, app performance bugs, and workflow workarounds requiring manual processes. This led to other features and fixes to alleviate these issues, some in the backlog, and a few that were newly identified. Backlog features have increased in priority, and I’ve created tickets and prioritized the newly identified ones. I estimate it cost an additional 16 hours of our users’ time to build and deploy.
If all issues are fixed, it will reduce our time to build and deploy by 34 percent.
Product Outcome Success
- Successful porting of existing client chatbot app deployments
- Client acquisition and adoption.
- 50% monthly cost reduction for clients.
- Decreased time to build by 85%.
- Decreased time to deploy by 99.9%.
- Decrease time to containment rate increase by 87%.
Conclusion
An extensive, meticulous, and strategic service design process helped me understand our users, the problems I needed to solve, the jobs to be done, business goals, and technical opportunities, all of which led to a great product vision and user experience. Keeping things simple, efficient, and optimal for users is a key competitive advantage and differentiator. Enabling a self-serviceable, quickly accessed, and independent product is key for our clients to meet the needs of their customers.
Other resources
