Know Your Customers (and Keep Them): Building Better AI Products Better

Company: WorkFusion
Role: Vice President, UX Design and Documentation

Previous Next

Goal

Customer problems are defined and clearly articulated so the product team knows precisely what needs to be solved and is confident in the delivery of meaningful solutions.

Process diagram of how a customer feedback loop allowed us to make smarter and faster product decisions.
Building a culture of user-centricity can be a daunting challenge unless you include a wide range of stakeholders in the feedback-gathering sessions, allowing them to feel invested in solving the problems that get defined as a result.

Opportunity

Redesign the product planning and prioritization process to be more end-user-focused and responsive to customer needs to:

Example of the output from internal design thinking sessions focused on reimaging our products through the lens of customer-centricity.
Leveraging design thinking techniques and a series of cross-functional workshops, I refocused the team on solving end-user problems while teaching valuable new creative and analytical skills in the process.

Flow diagram showing how moving from internal to external focused decision-making would happen across multiple releases.
A sound proposal for transitioning to a more customer-focused planning process must consider factors such as team capacity, backlog complexity, and the need to continually upskill the entire team on research methodologies.

Solution

Faced with an organizational culture of low design maturity, I had to solve several problems simultaneously. First, I had to build support and respect for the design team, which I started by mandating UX feature and release sign-offs as well as equal status for all designers at the scrum team level. Second, I became the de facto voice of the customer in all planning and prioritization discussions, moving the conversation from what was simply feasible to what was desirable for users. Third, I introduced customer feedback and product instrumentation as critical data points for all product decisions. Finally, I worked to correct the lack of understanding of design’s role in a mature product organization through a series of workshops and educational sessions.

Screenshots of the first AI-powered feature we included in a product release.
We wanted to be very clear with how we introduced advanced AI functionality in our product so as not to confuse users as to what the intelligent agent was doing and to be as transparent as possible by allowing them to see and adjust any and all work carried out by the agent before pushing it live.

Screenshots of the new data visualization engine we designed and how it manifested in customized user-centric dashboards.
Users were telling us that our off-the-shelf dashboards were not fit for purpose, so we spent a lot of time with customers to understand what their specific needs were and delivered a new, more customizable visualization engine that embedded AI functionality within to help them derive actionable insights from the data in real-time.

Outcomes

Flow diagram of the internally focused product planning process in place when I joined.
The inherited waterfall approach to product and feature planning had no allowances for external research or an iterative design approach to improving the product before putting it in customers’ hands.

Flow diagram of the new research-driven product planning process that I implemented.
Introducing user research, behavioral analytics, and iterative feature development allowed the team to make data-driven decisions that reduced overall churn at the scrum level and increased customer satisfaction with our new features and product release output.

Accomplishments

What I learned