What is an agent?

Anything that perceives its environment through sensors and in return acts upon it(Russell and Norvig 1995).

What is an intelligent agent? An agent that displays machine learning abilities.

Does perhaps Amazon Alexa, Apple’s Siri, Google Assistant, Microsoft’s Cortana ring a bell?

In “Research framework, strategies, and applications of intelligent agent technologies (IATs) in marketing applied” the authors attempt to define how are these intelligent agent technologies used in the context of marketing and how can marketers understand and exploiting them. First step towards that direction was to try and establish a marketing centric definition. Hence, Intelligent Agent Technologies are according to authors:

Systems that operate in a complex dynamic environment and continuously perform marketing functions such as:

• dynamically and continuously gathering any data that could influence marketing decisions
• analyzing and learning from data to provide solutions/suggestions
• implementing customer-focused strategies that create value (for customers and firms)

The second step was to classify all marketing applications of IATs in a way that would demonstrate relationships and differences among them. A useful and understandable tool for researchers and managers, the proposed marketing taxonomy is depicted below:

 

To answer all these research questions the authors reviewed the existing literature and then conducted 100 in depth interviews with managers form 50 randomly selected companies. Two independent researchers analyzed the interview data, that were then used to shape the taxonomy and the below framework.They also made some propositions that would help researchers and mainly managers to utilize IATs and ultimately drive company performance.

Overall, implementing the right IAT can assist the progress of numerous marketing functions permitting companies to achieve a sustainable competitive advantage. Both firms and customers can benefit from them. Companies are in a position to understand and put customers’ interest first (through collaborative filtering, personalization, recommendation systems) and in return gain customer loyalty and trust.On the other hand IATs offer consumers value, by providing them with convenience, better information, customized selection and less information overload (e.g price-comparison engines or agents that configure and customize their computer systems on the basis of their preferences).

Strengths and Weaknesses:

Since there was no concrete research or a fully developed theory surrounding IATs in marketing and subsequently no certain phenomena or existing theoretical frameworks to test, the authors rightfully opted for the grounded theory approach.So in contrast to the traditional research method they tried to construct a theory by discerning which ideas and concepts are repeatedly used in the interview data. These patterns were then grouped into categories that formulated their theory and shaped both the taxonomy and the framework.

Although the authors reasonably based their analysis on grounded theory, whether they applied it correctly is another question. The fact that they reviewed the existing literature in order to formulate the interview questions somehow conflict with the grounded theory methodology. The goal of this approach is to discern natural patterns. However, the used questionnaires possibly inhibited this since they kind of predisposed the managers’ answers since the queries were literature related.

“Given further progress in recommender systems (or other means of reducing costs for the customer), a situation might arise in which a “ready-made” solution provided by the system delivers higher preference fit than a customer-designed product—which, on the other hand, delivers the advantage of enabling “I designed it myself” feelings.” (Franke, N., Schreier, M. and Kaiser, U, 2010). This poses a very serious question for companies. When it is preferable to let an agent customize, decide or recommend a product/website? How quickly and how frequently should the agents respond and adjust to user needs? Ultimately what is more beneficial for both parties, implementing agents or give consumers the freedom to tailor products and and websites to their needs. Perhaps, technological advancements and the machine learning capabilities of IATs could soon enable companies them to successfully distinct these two categories of consumers and accordingly present them the proper interface.

 

References:

Franke, N., Schreier, M. and Kaiser, U. (2010). The “I Designed It Myself” Effect in Mass Customization. Management Science, 56(1), pp.125-140.

Kumar, V., Dixit, A., Javalgi, R. and Dass, M. (2015). Research framework, strategies, and applications of intelligent agent technologies (IATs) in marketing. Journal of the Academy of Marketing Science, 44(1), pp.24-45.

Russell, S. and Norvig, P. (1995). Artificial Intelligence: A Modern Approach. 1st ed. Prentice Hall, p.31.