Persuasion has been a cornerstone of education and business since ancient Egypt. Through the centuries, those who skillfully brought audiences to their own position were seen as wise scholars and merchants. Usually, they were well rewarded.
But global access to the Internet has created a virtual landscape where persuasion is open-sourced and citizens and consumers are overloaded with information. Too much information, rather than too little, is a problem decision makers increasingly face, and one that research suggests adversely affects innovation and productivity.
Internet users are inundated with incessant marketing as well as advertisements and promotional messages from businesses, individuals on the Internet, and social networks. Linda Stone, a former vice president at Microsoft who speaks and writes about attention, describes users as wanting to be a ‘live node’ on the information network. “In an effort not to miss any information received, the users are in a constant state of crisis alert termed ‘continuous partial attention’ resulting in compromised ability to reflect, to make decisions, and to think creatively,” writes Stone.
Crisis mode can be persuasive or dissonant in marketing and business; this depends on whether the immediate tension felt in the present can produce positive outcomes in the future. Managers in companies struggling to get projects completed on time and efficiently can be persuaded by offers of future rewards within the organization. A person looking for the best percentage rate on a home loan can be persuaded by a solicitation with the lowest rate. In both of these examples, the crisis is resolved by an offer that fits the needs of the recipient. This is what analytics can help accomplish: a best-fit business solution for competitive advantage. In other words, the use of persuasive analytics is to find the win-win.
Today, analytics inform most business decisions. But, like statistics, analytics are frequently applicable to only a small slice of the problem-pie and yield unpredictable results. Sellers are in the active business of prediction and persuasion, using analytics to cross consumer boundaries and blanket the market with messaging. To calibrate persuasion models effectively, companies buying and selling user information have to consider user attitudes toward persuasion. It is possible that less investment may yield higher returns if marketers understand the attitudes that underlie consumer behavior.
For instance, what is a recipient’s initial reaction to an advertisement’s intrusion on their Internet activity? Companies can use predictive analytics to make assumptions about consumers’ behavior, but may fail to understand their attitudes about method of persuasion. Measuring consumers’ attitudes as well as habits can reveal how consumers prefer to be persuaded, and assists advertisers in building long-range customer relationships.
Educational institutions also encounter a competitive challenge in determining how persuasion is received. Successful student recruitment begins with a school asking the right internal questions, broadcasting clear and consistent messages, and offering an accurate picture of what the institution is like and what it would be like to study there. Research-focused universities, such as Aarhus University in Denmark, focus on talent development and recruiting young researchers; Hong Kong University focuses on future collaborations among alumni and marketing the school as a strategic gateway to China. Interestingly, MIT, being oversubscribed, markets their scholarships for the less advantaged.
Leveraging predictive modeling and persuasive analytics successfully requires baseline knowledge of consumer attitudes, and then an adaptation to those attitudes. Understanding the patterns in consumer cues can enable effective sentiment analysis and help identify the most persuasive messaging. Metadata analysis can detect cues, such as words or phrases, that connote content similarity, which has been shown to be valuable tool in persuasion. For example, in a recent study tourist information was given to participants by presenters with accents similar to or different from their own. Participants overwhelmingly preferred presenters with accents similar to their own and viewed them as being more knowledgeable than speakers with different accents.
Sentiment analysis is the automated process of measuring people’s attitudes and emotions by evaluating large amounts of text or speech. The challenge in programming a machine to grasp the deeper meaning of sentiment is accuracy. Predictive decisions based on metrics or statistics alone risk inaccurate categorizations. Automation attempts to discern user attitudes, which are diversified across generation and gender, education-level and environment, culture and ethnicity. Challenges to accurate measurement include ever-changing language patterns, such as the shorthand of texting and emoticons, as well as attitudes associated with those patterns.
This is not traditional marketing, in which a business can calculate return on investment by comparing sales to marketing expenditures. The ROI from persuasive analytic efforts is hard to measure. The investments in time and money to predict behaviors and then persuade consumers to take actions requires more than Web optimization. Although specific optimization tools such as collection, analysis, and reporting can help, assessing underlying attitudes is an inexact science. “Just about anything that involves molding or shaping behaviors is persuasive, but persuasion should be the study of attitudes and how to change them,” writes R.M. Perloff, a nationally recognized scholar in the science of persuasion, in “The Dynamics of Persuasion: Communication and Attitudes in the 21st Century.” In a host of contexts, ranging from business and education to politics and health, the challenge to discover effective ways of influencing the ever-changing behavior of consumers begins with understanding attitudes.
