The Impact of External Word-of-Month Sources on Retailer Sales of High-Involvement Products 2014-09-18
Overview Problem Process Dataset Schedule
Problem Hypothesis 1. External WOM sources have a significant influence on the sales of high-involvement products at online retailers. Hypothesis 2. External WOM sources have greater influence on the sales of high-involvement products at online retailers than retailer- hosted internal WOM sources.
Process Empirical Model −𝐿𝑛(𝐴𝑚𝑎𝑧𝑜𝑛𝑆𝑎𝑙𝑒𝑠𝑅𝑎𝑛 𝑘 𝑖𝑡 )= 𝛽 1 𝐿𝑛(𝐴𝑚𝑎𝑧𝑜𝑛𝑅𝑎𝑡𝑖𝑛 𝑔 𝑖𝑡 + 𝛽 2 𝐿𝑛(𝐶𝑛𝑒𝑡𝑅𝑎𝑡𝑖𝑛𝑔 + 𝛽 3 𝐿𝑛(𝐷𝑝𝑅𝑒𝑣𝑖𝑒𝑤𝑅𝑎𝑡𝑖𝑛 𝑔 𝑖𝑡 + 𝛽 4 𝐿𝑛(𝐸𝑝𝑖𝑛𝑖𝑜𝑛𝑠𝑅𝑎𝑡𝑖𝑛 𝑔 𝑖𝑡 + 𝛾 1 𝐿𝑛(𝐴𝑚𝑎𝑧𝑜𝑛𝑁𝑢𝑚𝑅𝑒𝑣𝑖𝑒 𝑤 𝑖𝑡 + 𝛾 2 𝐿𝑛(𝐶𝑛𝑒𝑡𝑁𝑢𝑚𝑅𝑒𝑣𝑖𝑒𝑤 + 𝛾 3 𝐿𝑛(𝐷𝑝𝑅𝑒𝑣𝑖𝑒𝑤𝑁𝑢𝑚𝑅𝑒𝑣𝑖𝑒𝑤 + 𝛾 4 𝐿𝑛(𝐸𝑝𝑖𝑛𝑖𝑜𝑛𝑠𝑁𝑢𝑚𝑅𝑒𝑣𝑖𝑒𝑤 + 𝜃 1 𝐿𝑛(𝐴𝑚𝑎𝑧𝑜𝑛𝑃𝑟𝑖𝑐 𝑒 𝑖𝑡 + 𝜃 2 𝐿𝑛(𝑆𝑒𝑎𝑟𝑐ℎ𝐼𝑛𝑡𝑒𝑟𝑒𝑠 𝑡 𝑖𝑡 + 𝜃 3𝑖 𝐿𝑛(𝑁𝑢𝑚𝑂𝑓𝐷𝑎𝑦𝑆𝑖𝑛𝑐𝑒𝑅𝑒𝑙𝑒𝑎𝑠 𝑒 𝑖𝑡 )+ 𝜌 𝑖 + 𝜀 𝑖𝑡
Process Prepare Dataset Examine the correlation among key independent variables Empirical model analysis Robustness Check
Dataset Paper Trendata Amazon daily data(sales rank, review ratings, review number, release dates, price, search volume) on Canon and Nikon(Camera&Photo) from 2007.6 to 2007.10 Cnet, DpReview, Epinions daily date(review ratings, review number) Amazon daily data on Cell Phones(560), Televisions(33), Smart Watches & Accessories(316), Home Office Furniture(126), Household Cleaning(146) from 2014.11 to 2015.1 Data volume: paper’s product number is about 112, amazon’ product number of cell phones is about 560 Product type: paper’s just about high-involvement, trendata can also do analysis on low-involvement products Analysis Cycle: paper’s cycle is about 5 months, trendata is about 1 or 3 months Search Interest: Googles Trends search volume about a keyword in a day. Example: https://www.google.com/trends/fetchComponent?hl=en-US&q=html5,jquery&cid=TIMESERIES_GRAPH_0&export=3&date=3/2014+3m
Schedule 2014/09/23 determine types of product to analysis 2014/09/30 learn to crawling product info from Epinions 2014/10/17 learn to crawling product info from Cnet, DpReview 2014/10/23 crawling data; learn to get search volume from Google Trends 2014/10/31 learn to analysis using crawled data 2015/01/01 crawling data