The Interplay Between Online Consumer Reviews and Recommender Systems: An Experimental Analysis

We all know the situation: We are looking at a website to find the perfect product. A recommender system helps us sift through all the different options but we also take a look at customer reviews to make the best pick.

Recommender system have been found to be more useful (Senecal and Nantel, 2004) and to increase choice quality and confidence for a customer (Aljukhadar et al., 2012). At the same time, reviews from other users have been perceived as more trustworthy (Senecal and Nantel, 2004). So, what do you think happens when both, a recommendation generated by a system and previous customer’s review are available?

This is what the researchers Daniela Baum and Martin Spann (2014) tested during an experiment with 1332 participants of different age groups and gender from a large European country. Divided into 15 groups, the (informed) participants were asked to pick a gift for a friend on a hypothetical webpage. They picked one low-involvement product

Figure 1

(clock radio / package of coffee) and one high-involvement product (digital camera/ hotel stay). Depending on the group, participants received different information when making their pick, here some examples (also see cover picture): some participants did not receive any recommendation or review; for some, the provided recommendation and the review contradicted each other, while in other cases they supported each other. Now, let us take a look at the most interesting results of the study:

  1. green-arrow  Unsurprisingly, if the recommender system’s suggestion and the customer review support each other, more customers follow the recommendation. This scenario is particularly effective for low-priced, low involvement products, because it simplifies decision-making and increases confidence in our choice (Aljukhadar et al. 2012; Baum & Spann, 2014)
  2. arrows   A review which contradicts the recommendation or praises an alternate product (figure 1), significantly decreases the probability of following the recommender system’s suggestion on average by 30% for both, low and high-involvement products.
  3. arrows  Negative reviews have the largest impact for experience goods. For example, when choosing a hotel, only 17% of participants still followed the system’s recommendation in the presence of a negative review compared to 68% in case of a recommendation and a positive review. The authors attribute this finding to the category “experience product” where previous customer’s experience is perceived to be more informative also considering the financial implications (Baum & Spann, 2014). While similar results were found for all the other products, only in this case results were significant.

What do these findings imply and why are they important? The results of this study show that by disclosing information wisely, firms can greatly influence decision-making influence customers. They decide about displaying reviews, how many will be shown and also which (type of) review will be made available. Still, firms are well-advised to keep showing negative reviews as well and must carefully consider them to maintain and improve customer satisfaction.

Would the findings of this study be the same in real life? It is likely, but this study is limited by the fact it lacks a lot of information which customer encounter in real-life: brands, different prices, different product looks, more information. The authors should have opted for an online experiment on a real platform; despite its drawbacks the results would simply be more realistic. Only further experiments and data analyses will be able to answer this questions fully.


Aljukhadar, M., Senecal, S. and Daoust, C.E. (2012). Using recommendation agents to cope with information overload. International Journal of Electronic Commerce, 17(2), pp.41-70.

Senecal, S., Nantel, J. (2004). The influence of online product recommendations on consumers’ online choices. Journal of Retailing 80, pp.159–169

Baum, D., Spann, M. (2014).  The Interplay Between Online Consumer Reviews and Recommender Systems: An Experimental Analysis. International Journal of Electronic Commerce  19(1). pp. 129–161

 Cover picture:

Baum, D., Spann, M. (2014). Appendix 1 The Interplay Between Online Consumer Reviews and Recommender Systems: An Experimental Analysis. International Journal of Electronic Commerce  19(1). pp. 129–161

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s