AI and Analytics for Business
Voices, Minds, and Decisions — Dataset from Bazaarvoice
Yuping Liu-Thompkins and Anh Dang’s proposal specifically looks at how emotions embedded in reviews affect consumers and how that may change depending on where the consumer is along the purchase path.
I was in the car stopped at a red light when I found out that the proposal from Anh and I had been chosen for the Bazaarvoice Research Opportunity through AIAB. I pumped my fist in the air and shouted “Yes!!!” My daughter sitting at the back of the car looked at me with anticipation and asked, “Mommy, what happened?” How do you explain the excitement of getting a rich research dataset to a four-year-old? I replied, “Mommy is excited because she is going to start on a really big project. It’s gonna be a lot of fun!” Showing her understanding with a nod, she threw me a typical barrage of questions: “Are you going to build a big statue? Are you going to need a lot of pencils? Can I see it when you are done?” I laughed and promised her that she would see the finished project when I am done, but it was probably not going to be done in one day.
Later that day when I saw Anh, her face radiated the same excitement that I was feeling, and she gave me a big hug to share the camaraderie of success. By now you may be wondering what exactly we were so excited about. Just to be clear, being selected for this AIAB Research Opportunity does not carry a big cash prize, but it gives us something much more valuable: a dataset from Bazaarvoice about consumers’ review and web browsing behavior. OK, I admit it…we are nerds who would get so excited over obtaining a mountain of data about seemingly mundane behavior. But these seemingly mundane records of behavior could say a lot about how consumers shop, how they read reviews, and how the reviews they see factor into their purchase decisions.
Today few people would probably argue against the importance of online reviews. Most of us have at one point or another looked for reviews of a product or a business, and have more or less based our purchase decisions on the reviews we saw. But how do reviews really affect us? Why do we seriously consider some reviews and not others? How do the contents of a review affect us, beyond simple star ratings? In some ways, each review is like a mini-story with its own conflict and resolution: some happy and some unhappy. How do we choose which story to listen to? How do these stories translate into what we will do ourselves? These are some of questions that this Bazaarvoice Research Opportunity is trying to address.
Our proposal specifically looks at how emotions embedded in reviews affect consumers and how they may change depending on where the consumer is along the purchase path. The idea originated from Anh’s course project in my doctoral seminar. Together, with other Amazon.com “junkies” in the class, we talked about the kinds of reviews we see and how they affect our own purchases. During that conversation, Anh pointed out the relevance of emotions as reflected in some reviews and how they may affect us beyond simple positive or negative ratings.
This was more or less dream talk as we realized the difficulty in getting that kind of microscopic data. Then a few weeks later AIAB came along with the Bazaarvoice Research Opportunity, and it was exactly what we needed. The incredibly rich dataset tracks exactly what consumers see on each page, from product information to the content of reviews. More interestingly, it captures the information by screen. So if there are many reviews for a product and a consumer did not bother to scroll very far down a page, we know those later reviews did not get read. What made us truly excited, however, is the availability of order information for the same individuals. Therefore, if a consumer reads some reviews and ends up buying the product, we can link the browsing/reading behavior to the purchase decision. We can also see other product alternatives the consumer browsed and considered. All in all, the data captured a complete picture of consumers’ interaction with the retailer website, from browsing product information, reading reviews, to the eventual purchase decision. It was the perfect solution to our data problem. Without hesitation, we submitted our proposal. Echoing the AIAB webinar title for this Research Opportunity “He said, she bought”, we named our proposal “He Yelled, She Smiled”.
The rest is history. Now with data in hand, we are eager to start our journey of exploration. Having worked with AIAB on quite a few such Research Opportunities, I know we can expect a productive and helpful experience from both AIAB staff and the data sponsor. At the conclusion of the project, we will get to present our findings to the big shots at Bazaarvoice. This is the kind of opportunity that I personally love – an opportunity to turn ideas into top-quality academic publications and at the same time make a real impact on marketing practice (what more can I ask for?). For Anh, a first-year student at Old Dominion University’s marketing doctoral program, it is great encouragement that her ideas are valued, and it provides fertile ground for her to refine and play with her ideas. And of course the possibility of a very good publication at the end.
Today as I sit typing these passages, I still feel the initial excitement I felt in the car a couple of weeks ago. Instead of all the pencils that my daughter imagined I would need for the project, I know I will be spending a lot of time crunching data. But to a geek, that is fun!