Sam’s Club: New Data Sets and AI-Powered Insights

Context

The problem: Sam's Club's Merch and Finance teams were spending an excessive amount of time manually moving data between disparate tools to create financial forecasts for shipping expenses. This manual, inefficient, and tedious process hindered their ability to generate accurate forecasts, make timely business decisions, and dedicate time to other critical job functions.

Causes: The root of the problem lay in data fragmentation. Shipping expense data was scattered across multiple systems (Power BI, Excel, One Stream, Green Mountain), making it difficult to access, analyze, and ensure accuracy. Manual data gathering and analysis were time-consuming, and there was a lack of granular and historical data.

Discovery: To understand the problem space comprehensively, a four-week investigative research period was conducted. This included:

1. User Interviews: Two in-depth interviews with Merch and Finance leaders explored the nuances of D2H, SFC, and DFC shipping forecasting processes.

2. Current Experience Audit: A thorough audit of the existing process was conducted with product and business partners to uncover bottlenecks, inefficiencies, and manual tasks.

3. Competitor Analysis: A meeting with Walmart's IBG design team provided insights into how they addressed a similar challenge with their driver tree solution.

Discovery findings

• Data was spread out. Shipping expense data resided across multiple systems (One Stream, Power BI, Green Mountain, etc.) hindering accessibility and analysis.

• Users expressed concerns about data accuracy and completeness, leading to inconsistencies and forecasting errors.

• Extensive manual data manipulation across various tools was required to create forecasts, consuming significant time and effort.

• Users lacked visibility into the factors driving shipping expense fluctuations, making it difficult to identify trends or optimize strategies.

• The current processes were cumbersome for conducting scenario analyses and evaluating the impact of different shipping decisions

Prioritizations

The problems I validated through research were critical, impacting both operational efficiency and strategic decision-making. Addressing these issues would yield significant ROI through:

• Increased Efficiency: Automating manual tasks and centralizing data would free up countless hours for Merch and Finance users, allowing them to focus on higher-value strategic analysis rather than data manipulation. This directly translates to cost savings and improved productivity.

• Improved Accuracy: A single source of truth, better data governance, and enhanced visibility into expense drivers would lead to more accurate forecasts. This enables better financial planning, reduced unexpected costs, and more informed business decisions.

• Enhanced Strategic Insight: Providing scenario planning capabilities and data visualization tools would empower users to understand the impact of various factors on shipping expenses, leading to optimized strategies and more insightful discussions with operations partners.

• Faster Decision-Making: Timely and accurate forecasts would allow senior leadership to make quicker, more informed decisions, reacting effectively to market changes and optimizing resource allocation.

Exploring solutions

Based on the research and identified needs, I presented two potential MVP approaches for a new expense forecasting experience to my team:

Driver Tree Interface:

• Concept: Similar to Walmart's successful approach, this solution would present financial packages as a hierarchical list (a "driver tree"). Users could easily navigate and pull specific data points like order volume or shipping expenses, with data pulled from a centralized back-end system.

• Key Features: Hierarchical data organization, drill-down capabilities, ability to adjust variables and see immediate impact on forecasts.

• Benefits: Simplifies data navigation, enhances transparency, provides users with greater control over their analysis, and offers a quick and accurate way to forecast expenses.

Analytics Dashboard:

• Concept: A centralized platform with robust data visualization tools and reporting capabilities.

• Key Features: Interactive charts and graphs, customizable dashboards, filtering options for granular data analysis, potentially integrated scenario planning tools.

• Benefits: Provides a comprehensive overview of shipping expenses, facilitates trend identification, supports data-driven decision-making, and centralizes access to key metrics.

Evaluating trade-offs

While both options offered unique advantages, the Driver Tree Interface was the more direct and efficient solution for the immediate critical pain points we were trying to solve. Walmart's success with a similar model validated its effectiveness. Its ability to directly address the lack of scenario planning, simplify data navigation, and provide granular control aligned strongly with user needs. The hierarchical structure also naturally lent itself to financial forecasting, where line items and their drivers are crucial. An analytics dashboard could be a valuable complementary feature in a later phase, but the driver tree directly tackled the core forecasting challenge.

Final solution

We moved forward with implementing a Driver Tree Interface as the core of the new expense forecasting tool. This approach, proven successful at Walmart, established a single source of truth by integrating data from various systems into a centralized platform. It enabled robust scenario planning by providing users with intuitive tools to adjust variables and analyze their impact on expenses, empowering them to make informed decisions. Furthermore, this solution automated manual processes, freeing up valuable time, and streamlining the overall forecasting workflow by optimizing data selection and reducing unnecessary steps.

By building this driver tree interface, Sam's Club could expect:

• Increased accuracy and timeliness of forecasts.

• Significant reduction in time spent on manual data gathering and manipulation.

• Enhanced ability to explain variances and trends to business partners.

• More informed and agile business decisions.

• Greater strategic capacity for Merch and Finance teams.

This approach directly tackled the most pressing issues identified, providing a robust, efficient, and user-centric solution that empowered financial forecasting at Sam's Club in a way that current disparate tools cannot.