新產品開發流程的模糊前端指概念形成階段，是新產品開發成敗與否的關鍵所在。現有論文雖有簡明的階段說明，但並未有明確的執行步驟可以遵循。新產品開發品牌廠商掌握目標顧客需求以開立產品規格；台灣產業承接品牌廠商的委外製造或設計為主，亟需有企圖心的廠商轉型從事差異化產品開發與建立品牌，以提高產品的附加價值。本文經由分析新產品開發的關鍵成功因素與現有模糊前端作業的優缺點，並借鏡市場導向價值創造的模式，據以建構國內適用且具原創性的需求導向模糊前端創新流程，包括三大程序與六個操作步驟。三大程序為：(1)目標顧客需求萃取；(2)差異化產品／服務概念建立；(3)研發藍圖展開。六個操作步驟為：(1)確定主題範疇；(2)趨勢觀察與分析；(3)趨勢轉需求情境；(4)需求情境找需求屬性；(5)需求屬性轉差異化產品／服務；(6)研發藍圖規劃。此流程以產品／服務差異化為主軸，在工業技術研究院歷經兩次大規模實作顯示成效顯著；操作流程前後經過三次修改，具邏輯性而易操作。

In new product development (NPD), fuzzy front end (FFE) refers to the process in formative stages of new product concept. FFE is key to the success of NPD. Not only can it improve the effectiveness of new product development, but it also saves time in new product development. Literatures touch upon this stage, but they barely take deep look into the clear processes in execution. The original brand manufacturers (OBMs) target customer needs to define product specifications. Taiwan’s industry consists mainly of OEMs (original equipmentmanufacturers) and ODMs (original design manufacturers). Ambitious manufacturers are encouraged to evolve themselves into OBM by differentiating products and establishing brandnames, so as to add value to their products. This paper analyzes the critical success factors of NPD, as well as the strengths and weaknesses of existing NDP models. The authors, by learning from the marketoriented value creation model, construct the originative need-based fuzzy front end of innovation process (nFFI) that fits the domestic ecosystem. The nFFI includes three procedures and six steps. Three procedures are (a) extracting target customer needs; (b) establishing differentiated product/service concept; and (c) planning research and development roadmap. Six steps include (1) setting the theme areas; (2) observing and analyzing trends; (3) transforming trends into demand scenario; (4)translating scenario into need attributes; (5) converting need attribute into products/ services; and (6) plotting R&D roadmap. The nFFImodel, which revolves around differentiated products/services, was tested and implemented twice in large scale at Industrial Technology Research Institute (ITRI). The operation process has been refined three times in logical and easy to use way, thanks to the feedback gained from the experiments at ITRI.