Study Suggests New Ways of Using Online Buzz to Forecast New Product Sales
Thursday, May 15th, 2014
Companies can significantly improve the forecasting accuracy of forthcoming products’ performance by analyzing the history of online consumer buzz growth before product launch, according to a paper from researchers at the University of Georgia Terry College of Business.
Pre-release buzz refers to consumers’ online conversations (e.g., in blogs and online forums) about a forthcoming product before its launch, according to the paper, which is published in the most recent edition of the journal Marketing Science.
“Consumers frequently talk about forthcoming products. Unlike conventional consumer conversations that occur in person, online buzz is curated and can be tracked in terms of date and time, and with the decline in computing and storage costs, it has become increasingly affordable for companies to keep a record of the history and evolution in consumer buzz over time,” said Guyiang Xiong, assistant professor of marketing at UGA and co-author of the study.
Because buzz evolves over time and varies greatly from product to product before release, it contains valuable information that can forecast new product sales and predict new product performance.
Notably, the dynamic evolution of pre-release buzz data allows accurate prediction of a forthcoming product’s sales long before it hits the market.
“In fact, the sales forecast obtained from analyzing the shape of only partial history of buzz evolution, say, ending two months before product launch, can be more accurate than that using the total volume of all buzz up until the day of product release,” said Sundar Bharadwaj, the Coca-Cola Company Chair of Marketing at UGA and co-author of the paper.
This approach to accurately forecast new product sales well before its launch allows sufficient time for managers to make necessary changes to the product or its marketing plans. By way of contrast, the traditional forecasting, which depends on product sales during the first few weeks to predict sales later on not only delays forecasting but also severely restricts the ability of the firm to modify product.
In addition, the research finds product pre-release consumer buzz can be used to predict stock price of the firm at the time of product launch.
“Pre-release buzz also provides stock market investors with important information to predict the chance of product success,” Xiong said.
“It turns out that, before product launch, stock prices quickly reflect changes in pre-release buzz, said Bharadwaj. “The availability of pre-release buzz reduces the level of surprise in the stock market at the time of product launch, since the investors have already adjusted their expectations based on the buzz before product launch.”
The authors examined the blog and forum postings of more than 650 new video game released between 2009 and 2010. They build on recent advancement in statistics research to analyze the shape of each product’s pre-release buzz evolution path.
The authors also examined the factors influencing pre-release buzz, providing guidance for firms as how to manage pre-release buzz evolution to enhance new product performance.
“Sales forecasting before product release has been an important but challenging task, especially for new products without prior sales history,” Xiong said.
According to this research, to accurately predict new product success, managers should not only monitor how much pre-release buzz there has been in total, but also keep tracking the trend in pre-release buzz over time. In fact, in order to understand the changes in consumer interest or purchase intent before product launch, it is much more cost efficient to track online buzz than to conduct repeated consumer surveys.
“Moreover, given the financial value relevance of buzz, managers should consider more effective means of communicating about it to the stock market investors,” said Bharadwaj.