With internet still coming up as a strong way of commerce in this day and age, researchers looked into ways consumers’ purchase decisions can be influenced by each other. Chen et al. (2011) started looking at online word of mouth and online observational learning as ways this might happen.
Word of mouth (WOM) is the way people talk about a product, for instance in the form of a review or a rating of the product. So by word of mouth, you are affected by other consumers’ opinions. The two most important aspects of word of mouth are valence (positive or negative message) and volume of the messages (amount of reviews, ratings etc.).
Observational learning (OL) is the way people might be influenced to buy a certain products based on the actions of other consumers. This information contains only actions, and no reasons for these actions. So observational learning might contain less information than word of mouth, because actions speak louder than words this information might be perceived as more valuable and trustworthy.
For this experiment the researchers looked at data from Amazon.com. Amazon went through some changes in policy, which led to a unique situation for Chen et al. (2011) to look at. Amazon temporarily removed the observational learning part of a section of their website. This was in the section of the website where they sold camera’s. After taking away the observational learning information it was reintroduced later. This lead to two studies that were interesting to look at and made the researchers able to double-check their results: going from both WOM and OL to just WOM (1) and going from just WOM to both WOM and OL (2).
The researchers looked at the removal of positive and negative OL information separately. Removing negative OL information leads to an insignificant drop in sales, while removing positive OL information from the site leads to a significant drop in sales. On the side of WOM information, the researchers found that negative WOM information (a low product rating) has a more significant impact on sales than the positive WOM information (a high product rating). On a more general note the study found that the effect when both WOM and OL are present on the site, both types of information synergise and lead to a greater effect then both effects separately. Furthermore the effects of WOM and OL decline over time.
So what can we learn from this study and apply in our businesses? The different effects of positive and negative OL information could help very popular products while not hurting niche products. Also, platforms like Amazon can take advantage of this by offering this information. The results also indicate a complimentary effect from providing both WOM and OL information. You could take advantage from this by designing your information offering with both in mind. Finally WOM and OL effects drop over time in the products lifetime. So it’s important to provide good WOM and OL information in the early stages of a product.
Chen, Y., Wang, Q. and Xie, J. (n.d.). Online Social Interactions: A Natural Experiment on Word of Mouth Versus Observational Learning. SSRN Electronic Journal.