Cisco • Shipped 2024
AI support for enterprise IT customers
Cisco’s CX Cloud (customer experience platform) is an asset management tool for customers. I created AI-powered self-help and assisted help experiences to improve the efficiency and efficacy of CX Cloud support services. Confidential information in design assets are blurred out.
Role
Product designer
Team
3 product designers
1 content designer
1 ux researcher
2 software engineers
Timeline
Jun - Sep 2024
Tools
Figma
Miro
Qualtrics
THE PROBLEM
IT support in CX Cloud is time-consuming for customers & costly for Cisco
Users lack tools & resources for self-help
When users encounter an IT issue, they have to open a support case
Opening a case is a manual process
Users have to complete a multi-step workflow to open a case
Users have to wait for assisted help
Demand for support always outpaces supply, meaning users have to wait in a queue
THE SOLUTION
An end-to-end AI- powered support experience
50%
reduction in cases opened
$10M+
in costs saved
USER RESEARCH
Key finding: Customers need to troubleshoot issues immediately, whenever there is technology downtime
I synthesized previous research and conducted new user research through a survey, finding that customers prioritize support services in CX Cloud
Customers rank rapid support high in importance & low in satisfaction
9 out of 14 customers surveyed indicated Rapid Support as an underserved value proposition
Majority of top tasks performed in CX Cloud involve case support
11 out of 14 customers surveyed marked case support as one of their top 3 most performed tasks
Competitive analysis shows widespread integration of AI tools in IT support & case management
I surveyed the competitive landscape of IT support platforms, finding that the majority had AI embedded in the workflow. I conducted SWOT analyses for Cisco’s main competitors and created an inspiration board.
JOURNEY MAPPING
Key finding: Self-help has the biggest gap between current capabilities & customer expectations
To better understand customer pain points along the support journey and identify opportunities for integrating AI, I mapped out the support experience starting from the moment a customer is made aware of an issue.
After discussing with my team, we decided to focus on creating ux solutions for self-help and transitioning from self-help to assisted help, since this area presented the biggest customer need and business opportunity.
IDEATION + CONCEPTING
How might we leverage AI to enable customers to articulate their issue & get actionable recommendations?
Customers experience delayed time to resolution for their product issues because they lack the tools and guidance to self-troubleshoot. My team collaborated in several design workshops to come up with ideas for a self-help experience in CX Cloud.
Reframing the problem
We prioritized user needs, turning pain points into HMW statements.
Creative brainstorming
We created an idea board through several rounds of rapid brainstorming
Idea clustering
We clustered similar ideas and conducted silent voting, identifying top ideas to further develop.
Design requirements
I identified several key design requirements to guide my explorations for a guided self-help support solution.
1 / Avoid chatbot designs (too much back and forth)
2 / Use guided questions & prompt recommendations
3 / Provide sources for solution recommendations
4 / Personalize recommendations based on user’s profile
Design concepts
I hosted a design sprint where my team rapidly sketched out solution concepts.
ITERATION THROUGH FEEDBACK
Key finding: Users prefer iterating their inputs & receiving immediate AI recommendations
Wireframes
I created medium fidelity wireframes to visualize key screens in the self-help user flow.
User feedback
I showed my designs to 5 internal IT support users whose needs closely aligned with those of our customers.
1 / Users want to quickly edit and the update their inputs
2 / Users prefer full-page troubleshooting interfaces
3 / Users want to see support documentation & related cases
4 / Users want to easily transition from self-help to assisted help
The final solution achieves two main goals:
1 / Empower customers to resolve their own issues with guided self-help
24/7 customer support from AI Assistant
Immediate troubleshooting recommendations
Customers can update issue description to improve response
2 / Automate case creation for accelerated access to assisted help
AI auto-generates case details using issue description & system data
At-a-glance view of all case information with full edit control
Streamlined case creation reducing workflow from 6 steps to 1 step
Learnings
1 / Designing for B2B products requires an understanding users & customers
This was my first time working on a B2B product. The complex B2B landscape taught me to keep learning about product technicalities throughout the design process, and that it’s ok to test my assumptions and pivot along the way. I learned to differentiate between users (IT personnel) and customers (enterprise payers), as well as how to design for the needs of both groups.
2 / Your team is your biggest resource
I’m so grateful to have worked with such a collaborative team of designers, researchers and engineers who really pushed my design craft to new levels. Working so closely with people across different roles and levels of seniority taught me to leverage each team member’s strengths and ask context-specific questions to get the most valuable insights informing my designs.