Business Case
A fashion store with too much choice
The story: A clothing store has thousands of items. Customers walk in, feel lost, fill a basket, then put it down and leave. The few who buy rarely come back. The owner wants each customer to feel like they have a personal stylist.
In AI/AWS terms: AnyCompany, an online fashion retailer, has high cart abandonment and few repeat purchases because customers are overwhelmed by choice. It wants a personalized shopping experience.
For the exam: The problem is too much choice. The goal is personalization.
Business metrics to track
The story: The owner will count: how many visitors actually buy, how much each purchase is worth, and how many customers come back.
In AI/AWS terms:
- Conversion rate: purchases per site visit
- Average order value: dollars spent per transaction
- Customer retention rate: percentage of returning customers
For the exam: Conversion rate, average order value, and customer retention measure a personalization project.
Solution
The story: The store trains a personal stylist on its own sales history and customer reviews, and on what each customer liked, clicked, and bought. The stylist writes item descriptions that speak to each shopper, gives personal advice, works with the store's "you might also like" shelf to change suggestions and offers on the spot, and keeps learning new trends by itself.
In AI/AWS terms: An LLM that writes dynamic product descriptions, gives personalized shopping advice, and improves automated interactions:
- Fine-tuned on transactions, customer feedback, and interaction data (likes, clicks, past purchases)
- Integrated with a recommendation engine to adapt products and promotions to each customer in real time
- Continuous learning from new customer data and fashion trends, without manual intervention
For the exam: This case uses fine-tuning on the company's own customer data, plus a recommendation engine.