FMCG & Retail | Sales Performance & Territory Optimisation
Body ML price elasticity models + competitor monitoring + conversational analytics
Revenue growth from territory realignment against equivalent prior period
About
Horizon Consumer Brands is a fast-growing FMCG company with a direct sales force of over 300 field representatives operating across 8 states. The company sells through general trade and modern trade channels across personal care and home care categories. Sales territory performance had long been a source of tension — with leadership lacking the analytical infrastructure to distinguish between territory potential and sales rep performance.
Industry
FMCG & Retail | Sales Performance & Territory Optimisation
Company size
1,000 – 5,000 employees
Founded
2005
The Company
A direct sales business growing faster than its planning infrastructure
Horizon Consumer Brands is a fast-growing FMCG company with a direct sales force of over 300 field representatives operating across 8 states. The company sells through general trade and modern trade channels, with a portfolio spanning personal care and home care categories.
Sales territory performance had long been a source of tension across the business. Some territories consistently outperformed targets while others chronically missed — and leadership lacked the analytical infrastructure to determine whether these differences were driven by territory potential, representative performance, or both. The business was leaving revenue on the table without a clear way to quantify how much, or where.
The challenge
Forecasting on an unreliable foundation
The sales planning team was producing monthly forecasts using a combination of historical sales data and anecdotal input from regional managers. Forecast accuracy was poor — averaging below 40% at the territory level — which meant that inventory allocation, sales incentive structures, and resource deployment decisions were all being made on an unreliable foundation.
Territory design itself was a further compounding issue. Sales territories had been drawn years earlier based on geography and historical boundaries, with no systematic analysis of market potential. Some high-potential markets were being under-resourced while representatives in low-potential territories were hitting activity targets but unable to generate proportional revenue.
Sales incentive design was suffering as a direct result. Representatives were being evaluated against historical averages that reflected territory design rather than individual performance — creating frustration and making it impossible to identify who the real high performers were.
The Solution
Forecasting and territory optimisation in a single integrated intervention
Seven Billion deployed a two-phase analytical intervention: a forecasting model to improve sales prediction accuracy, and a territory optimisation model to realign resource allocation to market potential.
For forecasting, an ensemble model was developed using Facebook Prophet for trend and seasonality decomposition, combined with a Random Forest model to capture SKU-level and territory-level performance patterns. The combination outperformed either approach in isolation, particularly on territories with irregular seasonality or recent distribution changes.
For territory optimisation, a Linear Programming model was built that took potential scores — derived from market size data, current penetration rates, and competitive activity indicators — and optimised territory assignments to maximise total addressable revenue within existing headcount constraints. The model produced a concrete reallocation recommendation with projected revenue impact quantified at each territory level. Output was delivered through a sales intelligence dashboard giving regional managers real-time access to territory-level forecasts, actual-vs-forecast tracking, and the territory potential heatmap used to drive the reallocation recommendation.
The Results
A 9% revenue improvement and a fundamentally better-run sales operation
Sales forecast accuracy improved by 30% across the territory network, with the greatest gains in territories where the historical model had been systematically biased by distribution changes. Territory realignment, implemented in phases over two quarters, drove a 9% improvement in total revenue versus the equivalent prior period — the highest organic growth rate the company had achieved without new product launches or market expansion.
Regional managers reported a significant reduction in time spent preparing and reconciling monthly sales reports, with the dashboard replacing a three-week manual consolidation process. Sales incentive design improved measurably as a result of the more granular understanding of territory potential — with representatives now evaluated against meaningful benchmarks for the first time.
The commercial impact extended beyond the numbers. The business now had a systematic way to diagnose territory performance and act on it — a capability that had been entirely absent before the engagement.
The commercial impact has been significant. But more importantly, we now have the tools to keep improving — and a sales team that finally trusts the targets it is working against.
VP Sales, Horizon Consumer Brands
KEEP READING
Explore more studies on AI, analytics, and enterprise intelligence.
