ADR#002
Topic
What customer information do we need to collect, and what should the overall user flow be?
Problem
Presenting all available properties without first understanding the customer's needs results in an overwhelming and inefficient experience. Customers must manually browse numerous listings, many of which do not match their preferences or financial capability.
Without collecting essential information such as preferred location, budget, financing method, down payment capability, and other purchasing preferences. The system cannot provide personalized recommendations. This leads to irrelevant property suggestions, longer decision-making time, and increased effort for both customers and property consultants.
Decision
Instead of presenting every available property, the application guides users through a questionnaire to collect key information such as:
- Preferred location
- Budget and financing method
- Buying problems and promotion needs
Using this information, the system evaluates available properties and generates personalized recommendations.
Recommendation Formula
The property recommendation score is primarily calculated based on two key factors:
- Location Preference. The system compares the customer's preferred location with each property's location. Properties that match the selected location receive a the score.
- Budget Preference. The system evaluates whether the property's price fits within the customer's target budget. Properties that are match to the customer's budget receive a the score.
Other data about the customer's purchasing challenges, such as limited savings, income constraints, or credit history, are used to help our agents recommend properties with suitable promotions or financing programs.
The overall recommendation score is determined by combining these two factors. Properties with the highest scores are ranked first and presented as the most suitable recommendations for the customer.
Reasoning
Using these three factors provides a balance between recommendation accuracy and implementation simplicity. The model is easy to maintain, transparent for business users, and can be expanded in the future if additional recommendation criteria are required.
Discussion History
13 July 2026
Define the User Preference Flow
Discussed the overall user journey for collecting customer preferences before recommending properties. The objective is to gather sufficient information about each customer's needs, financial capability, and purchasing preferences so the system can generate personalized property recommendations.
| Participants | Roles |
|---|---|
| Martien | Product Coordinator |
| Kartika | Customer |
17 July 2026
Finalized the user flow by simplifying the customer journey to better align with the current business process. Unnecessary steps were removed, and the flow was streamlined to improve the user experience while remaining consistent with existing business operations.
| Participants | Roles |
|---|---|
| Martien | Product Coordinator |
| Kartika | Customer |
20 July 2026
Changed the flow so that users no longer receive property recommendations immediately. Instead, users send a WhatsApp message, and the KAHU agent will send them a link to their personalized property recommendations.
| Participants | Roles |
|---|---|
| Martien | Product Coordinator |
| Kartika | Customer |