How CCPL ai Helped an Indian FMCG
Giant Improve Data-Driven Decision Making
The Client
A national FMCG company with food and home care brands sold through roughly 4,800 distributors and reaching outlets across 22 states.
The Problem
Every region ran its own version of the truth. The western zone team tracked secondary sales in one Excel format, the southern zone used a different one inherited from a distributor management tool nobody else had adopted, and the east hadn’t moved off paper-based collection reports until the previous year. When the CXO office asked for a consolidated view ahead of a quarterly board review, finance and sales ops spent close to nine days manually reconciling numbers — and even then, two zones’ figures didn’t tie back cleanly, which meant a chunk of the board deck went in with a footnote instead of a number.
What We Built
We started with the unglamorous part: sitting with each zonal team to understand what their spreadsheets actually meant, because the same column header sometimes meant different things in different regions. From there, our engineering team built a data pipeline pulling from the zones’ distributor management systems and the handful that were still on manual uploads, standardising everything into one schema before it hit the reporting layer.
The dashboards we built weren’t one-size-fits-all – the CXO view showed national trends and zone-level comparisons, while regional sales heads got outlet and distributor-level detail relevant to their patch. We also flagged data gaps automatically instead of silently filling them, since the client’s leadership specifically didn’t want a system that “smoothed over” missing numbers.
The Results
By the second quarterly board cycle after go-live, the consolidated report was ready in under a day instead of nine, and for the first time in three cycles, every zone’s numbers tied back without a footnote. One unexpected outcome: the anomaly flagging caught a distributor in Telangana whose order pattern had quietly dropped 40% over two months, something that had gone unnoticed at the regional level until the dashboard surfaced it.
Why It Worked
This wasn’t really a technology problem at the start – it was a trust problem. Each zone trusted its own numbers and was wary of someone else’s system overriding them. The work that mattered most was the unglamorous mapping exercise before a single dashboard got built. Once people could see their own numbers reflected accurately in the unified view, adoption stopped being a fight.