Consumer Goods (CPG)

A replenishment analyst at a major CPG company opens Retail Link on Monday morning. She needs to review fill rates across 400+ SKU-store combinations before the weekly order cycle closes. The data she's looking at is six days old. By Thursday, she'll have her analysis ready. By then, the stockouts she's flagging have already cost a week of lost sales.

This is the rhythm of CPG analytics at most companies. Not broken in a dramatic way — broken in the slow, compounding way that erodes millions in revenue annually while everyone agrees "our data could be better."

The syndicated data bottleneck. Nielsen and IRI are the lifeblood of CPG planning. They're also notoriously slow. At P&G Canada, standard Nielsen extractions took up to 20 days through the UI — with frequent timeouts and failures that forced restarts. We bypassed the entire front end, connecting directly through ODBC, automating extraction through KNIME, and cutting delivery from 20 days to 3. The impact went beyond convenience. P&G's sales teams sold with data their competitors hadn't seen yet. That information advantage contributed to winning Walmart Category Captaincy in three major categories and an estimated $5M annual POS uplift.

The negative feedback loop nobody sees. Automated replenishment sounds like the answer. Often it's the source of the problem. At Walmart Canada, we identified a pattern across P&G's Femme Care category where the automation was systematically destroying its own forecasts. Products that required manual shelf stocking from backroom inventory — bulky items that didn't fit on the shelf in full case quantities — showed declining scan data when store staff fell behind on restocking. Retail Link's algorithm interpreted the low scans as low demand. It ordered less. Sales dropped further. Forecasts spiraled to single digits for products that should have been among the store's top performers.

We built a custom On-Shelf Availability algorithm using RetailLink POS data, inventory levels, and R/Knime workflows. It distinguished genuine low demand from availability-driven low sales across hundreds of SKU-store combinations. The fix was counterintuitive: strategic overstocking. Pilot results in 20 stores were visible within two weeks, leading to all-Canada expansion and then cross-category application to Shaving — P&G's highest-margin category. Total impact: $3M incremental revenue in four months, 10% stockout reduction, 5% improvement in on-time deliveries.

The reporting tax. Every hour an analyst spends assembling data is an hour not spent interpreting it. We automated P&G's weekly replenishment reporting pipeline — from Retail Link extraction to phantom inventory detection to analyst-ready dashboards delivered every Monday — and recovered 120+ hours per month of analyst capacity. In-stock rates improved 1% across the Walmart Canada portfolio. At that volume, 1% moves significant revenue.

Route-to-market visibility. For regional CPG brands, the challenge isn't just shelf performance — it's knowing which routes, distributors, and retail partners actually drive growth. We've worked with regional producers to digitize sales route data, analyze basket composition by territory, and surface the distribution patterns that separate their expanding markets from their stalling ones.

What the engagement delivers. A governed data layer where Nielsen, POS, and internal supply chain numbers reconcile automatically. Analyst workflows that start with insights, not extraction. Replenishment logic that distinguishes between demand problems and availability problems. And a clear connection between what happens at the shelf and what your planning team sees on screen.

Diagnostic questions. How many days old is the syndicated data your team makes decisions on? Could your analysts tell you today which low-performing stores have a demand problem versus a stocking problem? When your replenishment system reduces an order, does anyone verify whether the sales decline that triggered it was real?