Blogger interest and online buzz around new product launches is tightly linked to paid media spending, according to a new study by The Nielsen Company, which suggests that marketing strategies built around the separation of advertising and paid media from pure word-of-mouth tactics can be severely misguided.

After analyzing blog buzz volume, ad spending, purchase intentions and actual product sales, Nielsen units Nielsen BuzzMetrics, a leader in buzz analytics, and BASES, a provider of new product forecasting and consulting, found the best predictor of buzz for newly launched consumer-packaged goods is a large ad budget.

The study evaluated nearly 80 new products across several subcategories, launched in the U.S. between 2005 and 2006. On average, the top 10 percent of products in terms of buzz spent nearly $20 million on paid media, while the companies that generated the next 40 percent of buzz spent an average of $15 million and the companies that generated the bottom 50 percent spent an average of only $5 million.

“Splintering media sources along with emerging consumer-generated media are challenging one-way, mass-media advertising models,” says Robert Mooth, vice president of product development for BASES and an author of the study.  “However, our analysis shows that traditional mass media continue to play a critical role for most CPG brands. What has changed is that online buzz and consumer expression have entered the fray, resulting in a complex yet inseparable relationship within the overall marketing mix for many types of products”.

Not all CPG subcategories and products generated buzz at the same level.  In fact, 10 percent of brands accounted for 85 percent of total CPG buzz in the study. Over-the-counter drug brands have higher buzz, partly driven by consumers’ higher level of involvement. Edgy brands (such as Red Bill, Altoids and Viagra) were also among the top 10 percent of products with the most buzz.

The survey also found that most buzz tends to occur very early in relation to a new product launch, with peaks in buzz preceding peaks in sales two-thirds of the time in the launches studied. 

And for the select products that generate substantial buzz, there is evidence that buzz volume can positively influence sales. In a regression-based sales forecasting experiment, incorporating actual buzz levels results in more accurate forecasting models, by as much as 20 percent.