The Connected Consumer

Product Seeding: Word-of-Mouth Effects For and Beyond the Focal Product

Yakov Bart


WOM, Word-of-Mouth, Seeded Marketing Campaigns, Social Media, Spillover Effects, Viral Marketing

download pdf article

Seeded word-of-mouth (WOM) marketing campaigns are very popular
Marketers know that consumers are much more likely to trust their peers over traditional advertising when it comes to purchasing decisions.  Therefore, the idea of using product samples to promote new offerings is increasingly common. In the most recent industry survey by the American Marketing Association and the Word of Mouth Marketing Association, over 70 % of marketers said they have used or plan to use consumers to disseminate product-related content.
In a classic seeded WOM marketing campaign, a company sends product samples to a selected group of consumers, typically referred to as seeds, and encourages them to try the product and share the product information and their own opinions with other consumers. Seeds typically spread product-related information in the form of posts in online forums, on social media websites or as reviews on retail websites – and lots of these transmissions can be tracked by brands through online and social media monitoring.

Obvious and less obvious effects of seeded WOM campaigns
Usually companies concentrate on measuring diffusion of information about the product that has been seeded, the focal product. For example, if a campaign consisted of the Chanel brand sending new lipstick samples to seeds, then the marketing managers would typically focus on monitoring online communities related to cosmetics and collect data on how this campaign affected consumer conversations about this new Chanel lipstick only.

In our research, we demonstrate that such a narrow focus would be a mistake. Consumer conversations on products related to but different from a focal product are also likely to get affected. In other words, seeding a product can result in various spillovers. For example, seeding Chanel lipstick may lead to more or fewer conversations about other Chanel products or about lipsticks marketed by other brands. Moreover, WOM about a focal product among consumers in one segment could affect product-related or brand-related conversations among consumers in other segments. For example, while Chanel is likely to choose seeds that are expert cosmetics users, WOM generated by consumers in this segment may also affect how much folks talk about it in other consumer segments, such as more novice cosmetics users (see Figure 1). In our study (Box 1) we identified several effects:


Box 1: Studying WOM spillover effects of seeding cosmetic products
In order to empirically estimate the WOM spillover effects, we have collected a unique data set covering product-related discussion forums for cosmetics products from Naver, one of the largest internet portal sites in South Korea. We chose this field setting because of the large size of the South Korean cosmetics industry (about $1 billion in sales) and because it is common for cosmetics companies to use product seeding campaigns for promoting new products. Also, in this industry consumers often engage in online conversations about their experiences and new products that come in frequently, and there are many brands with products across multiple categories. Finally, Naver’s cosmetics forums are divided into a members-only forum for expert users and an open forum primarily populated by relatively novice users.

In our data collection, we’ve obtained a list of all products that were promoted in seeding campaigns run by cosmetics companies between February 2008 and November 2011. The sample included 390 products by 192 different cosmetics brands across 11 cosmetics product categories, (such as lipstick, toner and nail polish). We collected all posts mentioning these promoted (focal) products in both expert and novice forums and then collected all posts in expert and novice forums that mention these focal products’ brands or categories within the same time period. To identify spillover effects, we estimated a 16-equation multivariate model.

Effect of seeding on focal product WOM
As expected, seeds started spreading WOM through discussing their experiences with a new product, and, in doing so, this raised interest from other consumers. These consumers started sharing and generating new WOM about the focal product as well. The impact of seeding on focal products’ WOM both among experts and novices was positive and statistically significant: In both groups, conversations and posts about the seeded products increased. This mechanism is fundamental for the success of any seeding campaign and has long being established by past studies.

Segment spillovers
We found that the effect of seeding on WOM about the focal product is amplified through positive segment spillovers from organic WOM in both segments: Nonseed focal product WOM among experts positively affects nonseed focal product WOM among novices and vice versa. At the brand and category levels, segment spillovers were also present, albeit in a one-sided manner: Brand-related organic WOM among novices positively affects organic WOM among experts, and category-related organic WOM among experts positively affects organic WOM among novices.

Brand and category spillovers
More surprisingly, we found that WOM from product seeding campaigns triggered negative spillover effects across brands and categories. In other words, more conversations about the focal product reduced the “off-topic” conversations about other brands in the same category as well as other products of the same brand. For example, the more seeds talked about the focal Chanel lipstick, the less non-seeds talked about Chanel products in other categories and the less they talked about other lipsticks offered by competitors.

The negative spillovers are interesting because spillover effects in the advertising literature are primarily positive. Despite it being highly likely that exposure to a focal product triggers brand- and category-level thoughts, such cues do not seem to materialize as additional nonseed WOM about brand- or category-level topics. We believe that this happens because, in response to getting exposed to information about a focal product, consumers may focus on very specific aspects of the focal product, such as particular product-specific attributes or functionalities. In line with construal level theory and past studies on cognitive processing, such adoption of a concrete and narrow perspective in consumers’ minds may be suppressing more abstract brand- and category-level thoughts that might have otherwise occurred.

This pattern can likely be ascribed to the following reasons. First, as companies tend to encourage seeds to describe their experiences with a new focal product in as much detail as possible, the seeds may talk more about specific product functionalities and features rather than about overall desirability. Second, seeds are typically expected to describe the usual sequence of obtaining a focal product sample, unwrapping it, inspecting it, trying it and examining whether the product worked as intended. In doing so, it is also more likely that they will adopt a narrower perspective.


Marketers tend to consider only positive spillovers to be beneficial for a company, but negative spillovers should not be immediately classified as “bad news.”

Product type and spillover effects
Our understanding of the psychological mechanism behind the negative brand and category effects has allowed us to predict for which products these effects would be stronger or weaker. Specifically, the effects should be stronger when it is easier for seeds to describe their product experiences in a functional sense. Products of a more functional nature, such as face cream or mascara, are relatively easier for seeds to describe in more concrete and narrow terms that focus on attributes, functionality and performance. It would be much harder to describe primarily hedonic products (such as perfume) in such way. Consequently, marketers should expect that negative brand and category spillover effects will be stronger when the focal product is of a more functional nature. Our empirical results support these arguments.

The positive aspects of negative spillovers
Marketers tend to consider only positive spillovers to be beneficial for a company, but negative spillovers should not be immediately classified as “bad news.” There are upsides to this effect that managers can use in their favor.

  • Negative spillovers help concentrate attention on one specific product
    Companies may value the ability of product seeding campaigns to help nonseed consumers focus on the focal product and reduce buzz about their other products. For example, when introducing a new product, this could help focus as much attention as possible on that new product. Alternatively, product managers may be concerned about “competing” against other product managers in the same brand for WOM when running a product seeding campaign – the negative brand spillover effects suggest that this is unlikely to be the case, particularly for novice consumers.
  • Negative spillovers reduce attention on lower-quality products 
    It may also be helpful when there is substantial variation in product quality or desirability across products under the same brand. With reduced buzz about less-valued products through negative spillovers, a company can mitigate the danger that negative brand associations will contaminate WOM and consumers’ perceptions about a focal product.
  • Negative spillovers can help distract attention from a competitor’s product
    Perhaps most importantly, the presence of negative category spillovers opens possibilities for strategic product seeding against the competition. It suggests that companies could benefit from product seeding – not only through the positive effect on WOM for the own focal product but also through the negative effect on WOM for competitors’ products in the same category. In other words, this implies that managers could strategically deploy product seeding to reduce an impact of a competitor’s WOM campaign.  Moreover, negative category spillovers may benefit companies eager to increase their market share.

All in all, we hope that these findings will help companies better understand the full range of consequences of product seeding strategies and use them to profitably leverage both positive and negative spillovers across segments, brands and categories.



Yakov Bart, Assistant Professor of Marketing, Northeastern University, Boston, MA, USA[H1]  y.bart@northeastern.edu

Further Reading

Chae, Inyoung;  Stephen, Andrew T.; Bart, Yakov and Yao, Dai (2017); "Spillover effects in seeded word-of-mouth marketing campaigns." Marketing Science, Vol. 36, No. 1, pp. 89–104.