Product Recommendation - Handbook

Type: LP (Linear Programming)

This handbook explains the Product Recommendation sample problem in the LP Black Box platform.

---

The Problem

Scenario

A small online store wants to recommend a personalized product bundle to a customer with a $200 budget. The customer has expressed interest in three categories: Electronics, Fitness, and Books. Maximize customer satisfaction while staying within budget.

Product	Category	Satisfaction Score	Price ($)
Wireless Earbuds	Electronics	8	50
Smart Watch	Electronics	9	120
Yoga Mat	Fitness	6	25
Resistance Bands	Fitness	5	20
Novel (Bestseller)	Books	7	15
Cookbook	Books	4	12

Your Goal

Maximize total satisfaction score while staying within budget and customer preferences.

The Constraints

1. Budget: Cannot spend more than $200

   Formula: 50×earbuds + 120×watch + 25×mat + 20×bands + 15×novel + 12×cookbook ≤ 200

2. At least one Electronics item (customer explicitly asked)

   Formula: earbuds + watch ≥ 1

3. At least one Fitness item (customer explicitly asked)

   Formula: mat + bands ≥ 1

4. At least one Book (customer explicitly asked)

   Formula: novel + cookbook ≥ 1

The Variables

Variable	Meaning
earbuds	1 if Wireless Earbuds included, 0 otherwise
watch	1 if Smart Watch included, 0 otherwise
mat	1 if Yoga Mat included, 0 otherwise
bands	1 if Resistance Bands included, 0 otherwise
novel	1 if Novel included, 0 otherwise
cookbook	1 if Cookbook included, 0 otherwise

---

How to Use

Step 1: Load the Sample

Select Product Recommendation from the dropdown.

Step 2: Solve

The solver finds the optimal bundle that maximizes satisfaction while respecting budget and category preferences.

---

Try It Yourself

• Change the budget to $150 and see which items get dropped
• Add a constraint to include at most 3 items
• Add a new category (e.g., Home) with products

---

This demonstrates how linear programming can power personalized recommendation systems.