Project Overview
This project focused on leveraging a multi-modal Smart Assistant AI in the FinTech sector to not only enable personalized savings strategies for users living paycheck-to-paycheck and those interested in expanding their financial literacy, but also to significantly improve key business metrics such as Customer Lifetime Value, increase Average Account Balance, and boost the Conversion Rate for wealth products among two distinct customer segments.
Partnering with our Research team member, we utilized the Google Design Sprint framework and workshop for the Savings Goal Competency; a five-day, all-day workshop series to motivate and inspire rapid ideation + testing.
The Challenge: Overcoming Trust Barriers and Deepening Customer Relationships with AI-Driven Savings Solutions
Through interviews with finance coaches, SEO specialists, branch managers, and research partners, we identified a customer desire for a simple and intuitive saving tool for both enjoyable and necessary goals, such as vacations or debt repayment. Recognizing the challenge individuals face discussing financial matters due to shame or lack of knowledge, what stood out to me was a significant opportunity for an AI Assistant. This tool offers a judgment-free platform that provides personalized, AI-driven financial advice. Our objective evolved to: How can we motivate users to save by identifying their immediate needs and simplifying the savings process with automated, tailored calculations? This approach aims to facilitate small victories that contribute to achieving larger financial objectives.
People reveal more to AI than humans, driven by the belief that AI lacks the capacity to judge. This digital discretion opens new doors for service industries, transforming AI into an unparalleled vessel for sensitive disclosures.
- Tae Woo Kim, Journal of Service Research
Based on the insights from our SME interviews, I was curious how we could replicate the sensitivity and trust of the coaches, while leveraging the anonymity of the AI to mitigate the shame, fear, or discouragement people may have as a barrier to saving. Beyond just setting up the function, I was curious how we could provide successful encouragement to use behavioral science to 'nudge' them back on the path, similar to how gamification encourage streaks.
Consolidated map with How Might We statements of our v1 of the customer journey.
Takeaway 1: Reduce Cognitive Load, Leverage Accessibility Best Practices for Voice Response
In our Design Sprint for the U.S. Bank Smart Assistant, we focused on reducing cognitive load by using inclusive utterances, multiple input methods, and maintaining default iOS settings. Collaborating with developers, we leveraged the AT voice for audio responses, creating a one-to-one experience. Working with other designers, we developed a comprehensive Accessibility section in our Voice Design System, including annotations for color contrast, font size, layout, and cognitive load. These changes reduced rework and increased usability for all users, leading to a 20% reduction in error rates, a 30% increase in task completion speed, and a 25% boost in user satisfaction.
Takeaway 2: Small Steps, Big Gains with Drip Savings Strategy
My concept leveraged BJ Fogg's strategy on micro changes and habit formation by integrating drip savings into our design. I understood the power of small, consistent actions in driving significant behavior change. By building saving habits through manageable, regular contributions, I transformed these small acts into a routine. Celebrating each success, we could reinforce this positive behavior, smoothly guiding users towards more substantial financial achievements. This demonstrated the strategic design thinking at the core of our solution.
We Get You = We Got You
I aimed to understand participants' emotional drivers, enhancing their enthusiasm for 'fun' goals and offering encouragement through social proof and detailed spend analyses for more enduring objectives.
Takeaway 3: Personalization for the Win
At first, adding friction to uncover customer savings goals seemed backward since I've always aimed to simplify user interactions. Yet, this method turned out to be incredibly insightful for tailoring our approach to diverse financial objectives, from emergency funds to retirement savings. It helped me understand the emotional drivers and barriers our users faced, enabling me to personalize our services more effectively. This experience has reshaped my approach to future projects, highlighting the value of intentional friction in gaining deeper user insights.
Takeaway 4: Working within a Limited Lens
I recognize the challenge of having only a partial view of our customers' financial landscapes due to their diverse accounts outside our purview. Despite this, I believed in maximizing the value of the data we had, like spend history, to offer insights and motivate customers to consolidate their financial accounts with us. This strategy not only leverages our current capabilities but also showcases our potential to deliver more comprehensive support with greater account visibility.
One of the challenges of this project was the fragmentation and subsequent lack of visibility to the complete financial picture of our customers. We mitigated this by utilizing the existing capabilities, such as Spend History to inform their Savings Goal.
Overall Learnings
Reflecting on our Google Design Sprint, the journey was immensely positive, and oh-so-much fun! Fueled by the invaluable access to Subject Matter Experts (SMEs) which allowed us to swiftly navigate through our activities, we were able to leverage our existing wealth of past knowledge and insights to move quickly to address the specific problems Savings presents. The Sprint setup was crucial for breaking free from the inertia our team was experiencing, enabling us to create nimble, simple prototypes and get feedback to move quickly. The sprint's pace, coupled with the depth of ideation and the clarity of structure—resulting in actionable deliverables—was cathartic. Despite our team being transitioned off the project towards the end, with the Anticipate team picking it up to completion, I viewed it as a success, allowing us to reframe our workflow and reinvigorate our process.
In the year the Smart Assistant Launched, we achieved:
1.7M
Unique User Submitted Queries
2.1M
Transactions Conducted through Smart Assistant
87%
Questions Successfully Fulfilled