Retail & FMCG Analytics for Logistics & Supply Chain
Most FMCG planning teams are running weekly forecasts from data that's already four days old. In a market where promotions shift demand overnight, four days is the difference between a stockout and a writeoff.
Logistics operations in India, the GCC, and Southeast Asia share a common data challenge: high transactional volume, multi-party execution (3PL, 4PL, last-mile carriers), and a fragmented visibility picture. The shipper sees the TMS. The warehouse sees the WMS. The carrier sees their own system. Nobody sees the end-to-end picture in real time. Most logistics analytics implementations start by answering a question that should already have a live answer: did this shipment arrive on time? Demand forecasting that runs on last week's Excel export is not a forecasting system. It's a liability.
Delivering data and AI for industrial operations
Manufacturing, FMCG, supply chain, and logistics
India · GCC · Singapore & Malaysia · UK · North America
Amit leads every engagement personally
What we hear from operators
The problems we solve
These aren't hypothetical pain points assembled from industry reports. They're observations from actual plant floors, warehouse ops, and finance desks — written down because they come up in almost every first conversation.
Demand planning runs on weekly data exports
The planning team exports sales data from SAP weekly. Loads it into the forecasting tool — or more often, into Excel. Applies the forecast model. Sends to supply chain. By the time supply chain acts on it, the underlying demand has already shifted. Promotions calendar isn't connected. Seasonality isn't modelled at SKU level. The forecast is always chasing reality.
Trade promotion ROI is measured after the fact, if at all
Most FMCG companies spend 15–25% of revenue on trade promotions. Most can't tell you which promotions drove incremental volume and which just accelerated existing demand. The promotion data is in one system, the sales data in another, and nobody has connected them with enough granularity to calculate a reliable promotional lift factor.
Distributor sell-out data arrives too late
Sell-in numbers are visible in real time from the ERP. Sell-out from distributors arrives weekly or monthly in a spreadsheet. The gap between what you shipped and what actually moved off shelf — the one that tells you whether you have a distribution problem or a demand problem — is often invisible until the next reorder cycle.
How we work
Our approach
01
Connect SAP, POS, distributor, and promotions data
We unify sell-in from SAP, sell-out from distributor feeds and retail POS where accessible, promotions calendar, and any Nielsen or market data in use. One integrated demand signal per SKU per channel per geography — updated daily, not weekly.
02
Build demand sensing, not just demand planning
A statistical baseline at SKU level. Promotional uplift modelled from historical promotion data. Seasonality and event calendars integrated. Anomaly detection that flags unusual sell-out patterns — the kind that indicate either a stockout at shelf or an unanticipated demand spike — before the reorder cycle catches up.
03
Automate the planning process and the exception workflow
Forecast review meetings become exception-focused rather than data-compilation exercises. Planners spend time on the 20% of SKUs and promotions that drive 80% of forecast error, not on reformatting Excel exports. Replenishment recommendations generated automatically from the integrated demand model.
What changes
Outcomes
These are specific, measurable shifts — not benefit statements. Every outcome listed here has been achieved with a client.
Forecast accuracy: typical 60–65% → 75–80% at SKU-week level
SKU-level statistical forecasting connected to real sell-out data consistently outperforms Excel-based planning by 15–20 percentage points on MAPE. The gap widens during promotional periods.
Promotional ROI visibility: post-event analysis → pre-event modelling
Trade spend decisions informed by modelled ROI before the promotion runs, not just reviewed after. Incremental volume separated from demand acceleration across historical promotions.
Distributor sell-out lag: weekly spreadsheet → daily automated feed
Sell-out visibility updated daily for distributors with API or SFTP capability. Weekly for the remainder. The planning team stops waiting for the monthly distributor report.
Technology stack
Common questions
What buyers ask us
These are questions that come up in almost every first or second conversation. If yours isn't here, it will be in the first call.
We've tried statistical forecasting before and the planners didn't trust it.
Forecast adoption is a change management challenge, not a modelling challenge. When a statistical forecast contradicts what a planner knows from their market experience, the planner usually wins — and usually should. The approach we take is to make the forecast explainable, not just accurate. Planners need to see why the model says what it says, and have a structured way to override it with their judgement captured as data. Over time, the model learns from those overrides.
Our distributor data is inconsistent and often arrives in different formats.
This is universal in FMCG. We build a distributor data normalisation layer that handles different file formats, different product codes (matched to your master data), different time frequencies, and different levels of granularity. The clean output feeds the demand model regardless of what format the distributor sends.
We operate across multiple countries with different currencies and seasonality patterns.
Multi-market FMCG analytics is what we do across GCC, India, and Southeast Asia. Currency handling, market-specific seasonality calendars, different promotional calendars by market — these are standard requirements in the data model, not edge cases to be handled later.
Further Reading
Practitioner insights on this topic
Data Lakehouse for Logistics: Bringing Freight, Warehouse and Fleet Data Together
Freight in one system, warehouse in another, fleet telematics in a third. How a Data Lakehouse brings freight, warehouse and fleet data into one governed platform for real-time analytics and AI.
Read article →
Microsoft FabricMicrosoft Fabric for Logistics and Supply Chain: A Practical Overview
Most logistics operations still run on disconnected ERP, WMS, TMS, GPS and Excel — and the team spends more time assembling data than acting on it. How Microsoft Fabric closes that gap: one governed platform for real-time visibility, leaner inventory and lower transport cost.
Read article →
Supply Chain & LogisticsMulti-Tier Supply Chain Visibility: Why Tier-2 Supplier Data Never Reaches Your Dashboard
The supplier scorecard reads 98% on-time-in-full — then a line stops for ten days because a coating house you have never heard of loses a furnace. Why tier-2 data is missing for commercial reasons, and four routes to partial visibility.
Read article →
Other industries
Retail & FMCG Analytics in other industries
The same foundation, applied to a different operation. What changes is the metric that matters and the system it comes from.
Ready to move
Start with a conversation, not a proposal
First call is 30 minutes with Amit. We ask about your systems, your team, and your most pressing operational problem. You get a clear view of where the gap is and what closing it looks like. No slides. No pitch deck. No obligation to proceed.