Key results
brand-specific dashboards deployed in under 8 weeks, enabling SKU-level visibility across product performance, fulfillment, media ROI, and customer retention for Amazon, Walmart, and Shopify clients
analyst hours saved per month by automating QBRs—cutting report turnaround time from 3 days to under 1 hour and eliminating manual Excel work
at-risk SKUs identified in 3 months using predictive inventory alerts—helping prevent stockouts and protect revenue during critical sales periods
improvement in ROAS forecasting accuracy achieved by training ML models on historical campaign data and integrating insights into Tableau scenario planning dashboards
About the Client
A U.S.-based Amazon & Walmart Brand Growth Agency managing retail media, fulfillment, and analytics for 12+ consumer brands across Amazon, Walmart, and Shopify. The agency oversees 1,600+ SKUs and supports $100M+ in annual marketplace GMV.
Key Challenges
No proactive view of stockout risks across 1,600+ SKUs, forcing brand ops teams to rely on static inventory dumps and gut-based PO decisions
Time-intensive QBR creation across 12 clients, requiring analysts to manually consolidate sales, campaign, and fulfillment data from multiple sources every week
Inability to compare media efficiency across Amazon Ads, Meta, and Google, leading to suboptimal budget allocation and missed ROAS optimization opportunities
Lack of SKU-level visibility into performance drivers, making it difficult to identify margin-draining products or conversion issues at a granular level
Our Approach
Omnichannel Data Architecture on GCP
- Centralized data from Amazon, Walmart, Shopify, Meta, and Google Ads into BigQuery using Cloud Functions
- Standardized schemas for sales, marketing, SKU metrics, inventory, and returns using dbt
- Created blended KPIs such as TACoS, net margin, ROAS, and session-to-order rate to support cross-platform visibility
SKU-Level Inventory Forecasting
- Built ML-based forecasting models using historical sales velocity, PO lead times, seasonal weights, and promotion calendars
- Deployed predictive dashboards in Tableau highlighting SKUs at risk of stockout or overstock within 14–30 days
- Prioritized replenishment recommendations based on projected sales loss, holding cost, and fulfillment delay risk
Automated Client Reporting at Scale
- Designed modular Tableau dashboards tailored for brand managers, category leads, and account teams
- Replaced Excel-based QBRs with auto-refresh scorecards for product, channel, and campaign insights
- Reduced turnaround from 3 days to <1 hour per client—standardized across all 12 brands
ML-Based Marketing Mix Modeling (MMM)
- Trained ML models using campaign-level data to forecast marginal ROAS and detect saturation points
- Built channel-specific ROI curves for Amazon Ads, Meta, and Google
- Integrated into Tableau as a live scenario planner for budget reallocation and pre-campaign simulations
Implemented Solution
Omnichannel Sales Performance Dashboards
- Real-time SKU-level reporting across Amazon, Walmart, and Shopify
- Channel, brand, and category filters with week-over-week, YoY, and campaign overlays
- Profitability and return rate segmentation at the product level
Inventory & Forecasting Layer
- Tableau dashboards showing SKU health, coverage days, and replenishment risk
- Inventory exception reports filtered by warehouse, fulfillment type, and vendor
- Integrated restock logic using live PO arrival feeds and sell-through velocity
Retail Media ROI & MMM Simulator
- ML-based MMM estimating true media impact across platforms
- Saturation curve visualization to guide budget shift decisions
- Scenario planner allowing teams to forecast ROI lift for proposed budget changes
Customer Cohort & Retention Dashboards
- Cohort waterfall charts segmented by acquisition source and time-to-repeat
- LTV trends by SKU, campaign, and discount band
- CRM and ad-platform integration for re-targetable high-value and at-risk segments
Technologies Used
