Help Center

An AI-powered, unified support experience

Project Details

As lead product designer I drove the product vison, research, testing, and design of a new AI-powered help center.

Responsibilities

Research
Concepting
Prototyping
Usability testing
Design

Tools

Dovetail
Adobe Analytics
Zoom
Figma
Fullstory

Timeline

5 months

Status

Beta testing

The Team
Lead product designer (myself)
3 engineers
Product manager
Design manager
Research consultant
Senior product manager
Collaboration Partners
AI team
Support team
Knowledge Base team
Communies team
Director of Strategy & Operations
Director of Global Support

Impact

310% increase in engagement
34% reduction
in bounce rate
27% increase
in session length
14% reduction
in support tickets

Background

VMware Support Resources

VMware provides self-service support resources via several microsites:

Unification

In an attempt to simplify the self-service journey, VMware leadership folded these microsites into a single customer portal, Customer Connect. Theoretically, customers could now access all support resources from a single place.

Missteps

Unification was implemented by shoehorning the resource sites under a shared navigation and global search. The search tool was lacking: it had difficulty properly indexing the thousands of VMware content and had a tendency to show repetitive results.

The resource sites essentially still operated as separate entities, leading to a clunky and fragmented experience.

The customer portal unification did not have a positive impact on any support metrics.

Self-service engagement remained low
Support tickets continues to increase

Seeking a new direction, the Director of Strategy and Operations and Senior Director of Global Support approached our team for assistance.

Identifying

The Business Ask

Today, customers can freely file a support ticket for technical issues. Tickets are resolved by technical support engineers, making them more expensive than self-service support content. Leadership wanted to reduce costs by decreasing support tickets, asking for following:

Introduce friction in the support ticket journey to deflect support tickets

The Problem

The ask, while beneficial to the business, was hostile to our customers. It also ignored the root problem:

Findings self-service support is a complex and disjointed experience, hindering our customer's ability and will to resolve technical issues on their own.

This results in customers filing support tickets, leading to longer time to resolution, further frustration, and increased business costs.

As a designer, I wanted to empower our customers to solve their issues – not obstruct and frustrate them. How might we make self-service better, instead of making ticket filing worse?

Starting Out

Planning

With a less-than-desirable outcome already in the books, leadership wanted an accelerated turnaround for experience concepts. We chose to go with a lean approach, planning a 4 week process from concepting and MVP to shareout with leadership.

Workshopping

2 days

Concepting

1 week

Testing

1 week

Iterating

1 week
Workshopping
2 days
Concepting
1 week
Testing
1 week
Iterating
1 week
Sharing
1 week

Workshop Insights

Customer behaviors:
  • Try resolving technical issues themselves first
  • Start their support journey with a Google search
  • Commit to filing support tickets once they start filling one out
Business needs:
  • Increase in self-service engagement
  • Decrease in support tickets and operational costs
  • Scalable solution to adapt to customer segmenting based on support contract tiers

Concepting

Concept A: Early Intervention Troubleshooting

For this concept, I proposed an experience that harnessed the power our internal AI tools.

The VMware AI team had built an ML-trained text box which analyses customer issue descriptions and returns technical documentation, live agent chat channels, and AI guided troubleshooting flows.

Despite these capabilities, the ML text box was only being utilized on the support ticket page, far too late into the customer's support journey.

ML text box in the support ticket form
All support services in one place

Reading resources, AI services, and live support engineer options

Quality over quantity

ML optimized search returns resources meeting a threshold of 80%+ relevancy

Instant path to a support ticket

Ability to directly file a support request for critical scenarios

Seamless integration of support services

Easily switch between support resource types without leaving the page

Concept B: Hybrid self-service tool and support ticket

This concept combined discovering self-service and filling out a support ticket into a single journey.

Customers are offered self-service resources after providing some technical details and an issue description. They can choose to engage with these resources, or continue completing the rest of the form after a short waiting period has elapsed.

Familiar experience

Customers have seen the support ticket form before and are comfortable using it

Saves time and effort

Providing issue details for support resources simultaneously builds out a new support ticket

Delayed, not denied, access to the support team

Customers are given the chance to review support resources before filing their ticket for any severity level issue

Product accuracy

Customers can select a specific product from their entitled products list to ensure accurate support resources and ticket routing

Testing

While designing our prototypes, I reached out to VMware beta users to find research participants. We scheduled an intense week-long period to interview 15 users, synthesize research, and collect key findings.

Half of our participants tested Concept A followed by Concept B, with the other half testing Concept B followed by Concept A. We presented them with a broad task:

You are unable to access an ESXi host on vCenter. Using this prototype, how would you resolve this issue?

What did our users prefer?

Concept A rating:

8.06 / 10

Concept B rating:

6.33 / 10

Users strongly preferred concept A as an overall support tool. They liked that it accommodated both high and low criticality situations, and appreciated its streamlined approach to support resources. They found the click-through interaction of concept A and loading button in concept B frustrating, preferring to see resources all at once with no deterrence towards opening a ticket.

Iterations

Delivered Design

You made it!

Got any questions?

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.