Founded in 1902 and headquartered in Birmingham, Alabama, Coca-Cola Bottling Company United, Inc. (CCBCU) is the third-largest Coca-Cola bottler in the U.S. With 10,000 associates across the Southeast, CCBCU is dedicated to the production, marketing, sales, and distribution of a diverse portfolio of beverages while continually improving to serve consumers and customers better. Recognizing the need for faster, more precise data-driven insights accessible to C-Suite leaders and field representatives and being the leading supplier and service provider in each local market, CCBCU aimed to streamline its data operations and make real-time information readily available.
01. Objective
- CCBCU aimed to develop a one-stop solution that provided easy and faster access to accurate data and insights, eliminating the need for multiple resources, and simplifying data searches for operational leaders.
- The solution needed to be accessible on mobile devices, enabling users to generate insights during field visits, customer meetings, and other on-the-go scenarios.
- Additionally, the interface had to be user-friendly to increase platform usage and adoption across the organization, thus improving overall decision-making processes.
- CCBCU partnered with Decision Point for an enhanced solution to revolutionize their decision- making processes.
02. Solution Implemented
- BeagleGPT by Decision Point was implemented at CCBCU, streamlining their decision-making process and leveraging AI for a user-centric approach to increase the adoption amongst the end users.
- Additionally, the solution leveraged Microsoft Teams integration to enhance user familiarity and engagement.
- This comprehensive approach ensured that users could seamlessly access data and insights, aiding in decision-making and performance by ensuring timely and relevant insights.
Development Approach:
- Integration of Multiple Datasets: Our approach involved integrating multiple datasets into BeagleGPT to monitor various KPIs critical to United’s operations on Microsoft Teams.
- Seamless Query in Everyday Language: BeagleGPT’s advanced semantic layer facilitated seamless data querying in everyday language and offered context retention for deeper analysis. Extensive metadata, nomenclature, and lingo training were conducted to make data accessible to users in their day-to-day business language, similar to how they would communicate with colleagues.
- Ensured Data Privacy: Row-level security and logical filters were activated to ensure data privacy and relevance.
- Personalized Proactive Nudges: Personalized nudges were sent to users based on various criteria. These nudges were tailored based on persona, usage history, data thresholds, FOMO, social proofing, educational needs, and past 4 weeks’ usage patterns. The User Knowledge Graph automatically curated user data needs, bringing intelligence from the user cohort providing individual user-specific views and further enhancing the user experience.
User Personas and Use Cases mapped:
We identified and created specific roles within United’s team in BeagleGPT, aligning them with access levels and responsibilities. The target roles envisioned using BeagleGPT were:
Various use cases were mapped for these user personas for a personalized approach, including:
- Sales Performance: To Track sales KPIs such as volume, revenue, volume vs. PY, revenue vs. PY and use trend analysis to track revenue performance.
- Cooler Servicing & Placement Analysis: To track cooler placements, cooler penetration, cooler service history details and other related details.
- Delivery Execution: To monitor and analyze the order volume, returns, daily deliveries, product on the truck, case fill rates, and outlet level information.
- OOS (Out-of-Stock): To track product, brand, or category availability.
- Budget vs. Sales: To monitor the difference between actual sales and budgeted sales values.
- Run Rate: To administer target achievement tracking through KPIs like achieved vs. monthly target along with the current run rate (volume achieved per day) or the required run rate (volume required per day).
- Share Performance: To check for the value share, volume share or the growth of the organization.
03. Challenges
During development, challenges included real-time insight generation, maintaining data accuracy, thorough testing of NLP queries, reducing response time for queries, improving data visualization on mobile devices, and introducing context retention. These challenges were addressed through continuous enrichment of NLP capabilities and enhancements in data visualization and user experience.
Impact:
- The current success of BeagleGPT includes increased adoption of data analytics use cases among 37+ pilot users.
- Increased platform usage and adoption through centralizing data access and enabling real-time insights generation answering 1700+ queries.
- Engaged users with 150+ personalized storytelling nudges.
- Streamlined processes, reducing time spent navigating dashboards from minutes to less than 15 seconds per query, thus enhancing business planning and identifying opportunities.
- The integration of multiple user personas and 6+ use cases transformed how CCBCU accessed and utilized data, fostering a data-driven culture within the organization.
- Improved efficiency through market research, enabling United to monitor fluctuations in critical KPIs, and operational insights into new product introductions and sales metrics.
04. Testimonials
