Prototyping Model

Definition

The Prototyping Model builds a working, scaled-down version (prototype) of the software early, based on incomplete/unclear requirements, so the customer can interact with it, give feedback, and help refine the actual requirements before full-scale development.

Steps

1. Quick requirement gathering (basic understanding)
2. Quick design — focus on user-visible aspects (UI, key features)
3. Build prototype
4. Customer evaluates prototype and gives feedback
5. Refine prototype — repeat steps 3–4 until requirements are clear
6. Engineer the actual product based on the finalized requirements

Requirements Gathering → Quick Design → Build Prototype → Customer Evaluation
        ▲                                                          │
        └──────────────────── Refine ◀────────────────────────────┘
                                  │
                                  ▼
                    Develop Final Product

Types of Prototyping

• Throwaway (Rapid) Prototyping — prototype is discarded after requirements are clarified; final system built from scratch
• Evolutionary Prototyping — prototype is incrementally refined into the final product

Advantages

• Helps clarify unclear or incomplete requirements
• Reduces risk of building the wrong system
• Improves customer involvement and satisfaction
• Early visual/functional feedback

Disadvantages

• Customer may mistake the prototype for the final product (expects it "finished" too soon)
• If not managed well, throwaway prototype code sometimes gets reused into production (poor quality risk)
• Can be costly and time-consuming if requirements keep changing
• Developer may make implementation shortcuts (skip proper design/testing) to build the prototype fast

When to Use

• When requirements are vague, unclear, or evolving
• Projects involving significant user interface design
• Where user involvement is critical to correctly capturing needs

---

Exam tip: A common question asks to compare Throwaway vs Evolutionary Prototyping — Throwaway discards code, Evolutionary builds on it toward the final product.