Summary
A prominent consumer products company specializing in household items faced challenges stemming from inaccurate and delayed trade promotion analysis, hindering its growth potential. To tackle this, Sonata implemented AI-Powered Trade Promotion Analysis Solution across the company's marketing and sales teams, benefiting over 50+ users responsible for managing trade promotions and analyzing their effectiveness. The solution is expected to bring an additional $1 million in revenue and $500k in cost savings within the first year of implementation.
Client Overview
The client is a prominent US-based consumer products company specializing in household items, specifically in the food and beverage packaging sector. With an annual revenue of $4B in 2023, their customer base includes households and consumers, restaurants and foodservice providers, retailers and grocery stores, commercial kitchens and catering companies, as well as online retailers and e-commerce platforms.
Pressure Points
Growth challenges due to inaccurate and delayed trade promotion analysis
Challenges in understanding the relevance of ongoing trade promotions
Data analysis was time consuming due to reliance on other teams for reports, leading to decision-making delays
Needed assistance in formulating future deals and strategies
Solution
The solution included a customizable dashboard for tracking active deals, an automated analysis tool for monitoring KPIs such as ROI and % trade efficiency, and cost per incremental unit.
The solution also combined Generative AI capabilities to provide insights and personalized recommendations. Visual trends and insights were also provided to enhance the decision-making process.
Design Highlights:
Key design considerations are.
- Building responsible AI foundation platform consisting of below mentioned modules using Sonata’s Harmoni.AI Platform
- Leverage Amazon Bedrock's advanced capabilities like chain of thought (CoT) to fine-tune the Generative AI model on the client's historical trade promotion data, enabling it to provide highly relevant and personalized recommendations based on the specific context and nuances of the client's business.
- Usage of LLM Ops for the efficient deployment, monitoring, and maintenance
Harmoni.AI Foundation Modules
- Zero-Trust security framework
- Service usage optimized for cost and performance
- Logging & Monitoring
- Account represents all resources for one project/application
- Hub-Spoke model: Transit Gateway preventing direct access to resources
- Templatized for quick provisioning of new project environments
- Sonata Ready Assets for rapid engineering
Service Catalogue of Approved Use cases
- VPC
- Private subnet
- Lambda functions - Most API Workloads
- EC2 - Any Custom Workload
- S3 - Blob storage, Single Page Application Hosting
- SNS - Event Requirements
- DynamoDB - Storage
- Sage Maker - ML Requirements
- Bedrock – Generative AI
- AI Services - Various Cognitive Needs Depending on Identified Use Case
- IAM for Granting Access to the Account
- AWS Config for Managed and Custom Rules to Manage Resource Configurations
- CloudWatch and CloudTrail - Monitoring and Auditing
AWS Landscape Details:
An Organizational Unit for Sandbox
An Account per Innovation project
Results that Speak Volumes
By optimizing trade promotions and resource allocation, the solution is expected to help an additional $1 million in revenue and $500k in cost savings within the first year of implementation.
20% time saved
for analysis, allowing marketing teams to focus on strategy and promotion execution
10% ROI increase with better understanding and optimization of promotions, leading to improved resource allocation
Impressive reduction in data processing time & enhanced UX with a streamlined One-Lake architecture and Power BI Direct-lake
