Case Study: Supporting a Cameroonian Startup with Decision-Making AI

Client: AgriTech Solutions Cameroon (Startup specializing in connecting small farmers to markets)

Problem: Uncertainties about the segments to target, the prices to apply and the optimization of limited resources.

1. Context

AgriTech Solutions had just raised €50,000 but was facing critical decisions:

• Which segment of farmers to prioritize?
• How to set attractive yet profitable prices?
• What marketing channels to use with a tight budget?

2. AI Decision Support Solutions Deployed

a) Predictive Analysis of Customer Segments

Use of a model combining: satellite data (production areas), local market buying habits, and mobile access. Result: Identification of high-potential regions (West and Littoral) and the most profitable crops (cocoa, plantain).

b) Dynamic Pricing

Algorithm analyzing: logistics costs, AgriMarket competitor pricing, and demand elasticity. Result: Optimal price range determined (15-20% below traditional intermediaries).

c) Optimization of Marketing Channels

AI simulation testing 3 scenarios: social media, SMS, rural radio. Result: Allocation of 70% of the budget to SMS (conversion rate 2x higher than expected).

• +40% of active farmers in targeted areas
• Gross margin increased by 25% thanks to optimized pricing
• 30% reduction in customer acquisition costs
• Fundraising of €200,000 facilitated by these KPIs

3. Results Achieved (6 Months Later)

1. Testimonial
   "AI saved us from costly mistakes. Without these insights, we would have targeted the wrong crops and wasted 60% of our marketing budget. Today, our decisions are based on solid data."

— Jean K., Founder of AgriTech Solutions

1. Tools and Methodology
   • Data Sources: Field surveys, FAO data, mobile money APIs.
   • Technologies: Python (Pandas, Scikit-learn), Tableau for visualization.
   • Approach: 3 collaborative workshops to adapt the models to local realities.

Case Study: Merchandising Optimization with AI for a Ready-to-Wear Retail Store in Douala

Client: Style & Co (Urban fashion store located in downtown Douala)

Problem: Inefficient inventory management, poorly targeted promotions, and suboptimal store layout.

1. Context

Style & Co was facing recurring difficulties:

• 30% of dormant stock (poorly placed or poorly promoted clothing)
• Low turnover rate on new items
• Underutilized store space (unidentified hot/cold zones)

2. Intelligent Merchandising Solutions Deployed

a) AI Heatmap Analysis of Customer Traffic

WiFi sensors and cameras analyzing customer paths + time spent per section. Result: Identification of "dead zones" and reorganization to increase visibility.

b) Dynamic Pricing and Targeted Promotions

Algorithms combining: sales history, local weather, and social events (festivals, concerts). Result: Automatic discounts on stagnant items (+12% turnover).

c) Planogram Optimization

AI generating virtual shelf layouts tested before deployment. Result: Better exposure of key products (+22% sales on those items).

d) Predictive Replenishment

Model anticipating demand based on local social media trends (via NLP analysis of posts). Result: Reduction of stockouts on bestsellers (-40%).

3. Results (4 Months Later)

• +35% overall sales thanks to optimized layouts and promotions
• 50% reduction in dormant stock
• Customer satisfaction rate increased from 68% to 89% (post-purchase surveys)

Testimonial
"AI has revolutionized the way we manage the store. We now know exactly what to highlight, when, and at what price. Our customers love the new products, and our margins are much better."

— Sarah M., Manager of Style & Co

1. Tools and Methods
   Technologies:
   • Computer Vision (analyzing customer traffic)
   • Python models (Prophet for forecasting)
   • Real-time dashboard (tracking KPIs)
   Approach:
   • Free initial audit
   • Team training on tools
   • Monthly performance tracking