應用小世界網路之新科技車輛購車選擇模型:以油電混合車為例

Abstract

本研究結合小世界理論模型與動態離散選擇模型預測新科技能源車輛進入市場的市占率與滲透情形，並選定屬於漸進化科技產品-油電混合車作為案例分析。模型涵蓋個體消費者偏好與口碑影響力進行總體市場預測與探討受到油價、車價、政策等外在變數對於市占率的非線性影響關係。本研究以實驗設計與敘述性偏好設計問卷並利用羅吉特模式校估樣本資料，車型方案選擇包含傳統汽油車和油電混合車。模式校估結果指出初次購車與非初次購車消費者共同的顯著影響變數為車輛售價、油價和口碑影響力，非初次購車消費者相較於初次購車消費者更重視油價與口碑影響。不同於非初次購車消費者，初次購車消費者會受到車輛加速度表現和消費者本身環保意識影響。在政策誘因方面，提供報廢老舊車輛並換購油電混合車者之現金補助政策可有效鼓勵非初次購車消費者選購油電混合車款。本研究也探討不同來源訊息的口碑影響力在購車決策中所扮演的角色，結果顯示車商運用口碑行銷策略會比透過廣告的行銷策略效果更佳，在不同訊息來源中，以周遭固定討論對象最具影響力，次者為網路口碑影響力，另外也發現負面訊息的影響力略高於正面訊息的影響力。

我們以美國油電混合車歷史銷售資料驗證模式的可行性，市占率預測結果平均約高估0.88%，具有不錯的精確度。根據台灣人口統計資料進行的模擬結果顯示油電混合車車價高於汽油車價1.56倍以下是消費者較可以接受的價格，可以作為車商訂定促銷優惠價格的參考，而油價在32.91元/公升時是實施促銷方案較好的時機。網路的口碑訊息會在不同油價下對油電混合車市占率產生約5~10%的影響，然而不論在任何油價趨勢下，產品滲透率達20~25%時對於提高消費者購買油電混合車意願的效果最佳。一旦負面訊息經由網路散播，會造成產品上市初期市佔率成長極為緩慢，並產生 9~25%市佔率損失。政策模擬結果顯示隨油課徵燃油稅(交通稅)是有效且符合經濟效率的政策方案，若課徵3%交通稅，油電混合車市占率可達5~11%的提升效果，並可達5%產品滲透率提升效果；若課徵5%交通稅，油電混合車市占率提升9~17%，並可達10%產品滲透率提升效果

We develop and empirically test an agent-based vehicle consumer choice model on hybrid electric vehicle (HEV) market share and market penetration. The model combines small world theory and choice-based conjoint adoption model, using individual-level preferences as a basis for the prediction and exploration of sensitivities and nonlinear interaction between various potential influences. We take into account word of mouth (WOM) effect and specific agent willingness to adopt into the HEV market estimation. This study examines factors and policy incentives that are most likely to influence first-time and non-first-time buyer choices for Hybrid Electric Vehicles (HEVs) in Taiwan. Questionnaire surveys are based on experimental design and stated preference methods. Choice alternatives of car type include conventional gasoline vehicles and HEVs. Each option is described by a varying set of vehicle attributes and economic incentives. Calibrated results of logit models indicate common significant variables for both buyers are purchase cost, gasoline price and word of mouth (WOM). Nevertheless, differing from non-first-time buyers, first-time buyers would be also influenced by acceleration performance and their own environmental consciousness. On the other hand, incentives such as offering cash subsidy for disposing old cars would encourage non-first-time buyers to adopt HEVs. We also discuss the role of word of mouth from different sources in affecting car purchase behavior. Results indicated manufacturers apply WOM for marketing strategies is better than advertisement. The WOM from close people has the most powerful effect among all sources, while the secondary is from internet. Moreover, negative WOM has more impacts than positive WOM. The estimated market share is just 0.08% higher than actual US HEV market share, validating our model yields high precision. Using Taiwan demographic data, our simulation results indicate the acceptable HEV price is about 1.45 times that of internal combustion engine counterpart and it is good timing to do promotion when the gasoline price is NT$32.91/L. With 20-25% of HEV market penetration, it would effectively encourage consumers to adopt HEV. The positive influence of internet WOM effects on HEV market share is 5-10%, however if the negative WOM spread through internet, it will make 9-25% loss of market share. For policy leverage, adding 3% transportation tax will increase 5-11% HEV market share and 5% market penetration; adding 5% transportation tax will increase 9-17% HEV market share and 10% market penetration。