The Value of Applying a Data Driven Behavioral Sentiment Analysis for Business Sustenance and Improving ROI in Disrupted Marketplaces 

Authors: Dr. Aaron Wester, Linar Molotov, Dr. Huiyu Qian, Dr. Rachna Kumar 

Insights from analyzing data that compares behavioral influencers such as the “frustration” of the consumer with the purchased product, and purchase experience before and during the global pandemic have provided critical insights for organizations to navigate challenges during disruptive environments. 

Ice hockey legend, Wayne Gretzky, was once asked what his secret was to being so successful in the sport. His response was simple and poignant: “a good hockey player plays where the puck is. A great hockey player plays where the puck is going to be.” An article in Real Money magazine by Lang in 2021, drew a parallel between Wayne Gretzky’s winning strategy, business, and money matters by advocating that businesses should stay ahead of the puck proactively, and predict and adapt. Easily said, hard to execute. But businesses can accomplish a similar feat from a return-on-investment perspective by utilizing data-driven modeling for evaluating behavioral influencers and using predictive probability modeling to pivot their business or strategy as needed. The COVID-19 global pandemic negatively impacted all businesses by disrupting supply chains, offline and online sales completions, worldwide distributions, and inventory availability among other essential factors. Some companies perished and went the route of Chapter 11 Bankruptcy because they could not weather the disruptive storm and stay ahead of the puck. Companies such as Hertz, J. Crew, GNC, Pier 1 Imports, Brooks Brothers, Century 21, California Pizza Kitchen, Gold’s Gym, ALDO, Bed Bath and Beyond, and Sizzler just to name a few.

Amazon showed they stayed ahead of the game, reporting an $8.1 billion in profit, representing an increase of 220% year-over-year. We analyzed Amazon purchase data during Covid’s chaotic disruption through sentiment analysis. Since tensions ran high during a global pandemic, increased consumer frustrations disrupted the shopping experience leading to higher churn volume, higher product returns, diminished brand loyalty, and ultimately fewer sales and business decline or close. We analyzed behavioral data from Amazon reviews in the U.S. before and during the pandemic from three key shopping categories on Amazon.com:  Children’s Category (e.g., Toys & Games Department), Home Category (e.g., Home & Kitchen Department), and Clothes Category (e.g., Clothing, Shoes, & Jewelry Department). Using 500 consumer ratings for 116 best seller products, data was scraped from Amazon.com’s department-specific product pages for purchases between Jan 2018 through March 2021 (1 year prior to COVID-19 and 1 year during Covid-19). The data was scraped using ParseHub, through the Amazon Web API via Amazon Affiliates access, and JungleScout. The data included:  star rating, time stamp of purchase, location of purchase, type of product bought, type of customer, and how many products customer found useful.  

For the categories of Children’s, Clothes and Home purchases, we found significant increase in the level of frustration in the shopping experience during the COVID-19 global pandemic when compared to previous year across all behavioral parameters studied. Organizations may benefit from the knowledge that consumers now, more than ever before, have a hyper awareness and sensitivity to product related frustrations tied to both product purchase quality and overall experience. B2C brands such as Amazon should then spend more for allaying consumer frustrations overall, including the purchase and product experience. Amazon focused substantially on increasing trust, warmth of purchase experience, emotional dependency, and decreasing purchase obstructions. We believe behavioral sentiment analysis is critical for Amazon and other B2C providers to stay ahead of the puck through targeted and actualized self-sustaining measures and tactics.