“Cross selling in the banking industry, the value of recommendations among branches’ sales managers”


Cross selling entails selling products in the firm’s product line that the customer does not currently own. Previous studies have developed models, based on past customer behavior, to understand what products should the firm cross sell to which customers at which time. However, as recent researches showed two problems arise with the indiscriminate use of cross selling models based on past information. First certain, customers with persistent adverse behavioral traits (e.g., limited spending) might exhibit a downward spiral of unprofitable relationship the more the firm cross-sell them.  Second, customer behavior might change over time and the drivers of such changes might not be properly stored in a database.

Encouraging sales managers to cross sell their customers through recommendations (referral) to other business units might be a way to overcome the aforementioned problem. Real life cases showed that following this strategy firms can significantly improve their net profits. In this paper we study conditions under which customer cross selling can be improved through recommendations of sales managers across business units. We use a rich data set that combines the number of recommendations and successful outcomes among 252 sales managers who belong to different business units of a financial group. Our results indicate that, at the dyad level, cross-selling is not more effective among business units with a natural sequence of consumption (e.g banking first and investment then), that incentives are useful to increase cross selling rate across business units with natural sequence (e.g. banking first and travel service then), that demographic similarity does not increase the likelihood of cross-selling but strength of ties does, and that normative pressure matters. At the sender level we do not find support to the influence of opinion leaders and finally at the receiver level we get that there is a limit in his/her capacity to manage recommendations.

Guillermo Armelini, ESE Business School, Universidad de los Andes

Sebastián Maldonado, Universidad de los Andes

Erica Salvaj, Universidad del Desarrollo