14 May 2016

Searching for Information Online Using Big Data to Identify the Concerns of Potential Army Recruits

Download Free Electronic Document
Full Document


0.5 MB 

Technical Details » 

Research Questions 
How have Army-related searches changed over time and across locations? 
What sorts of questions and concerns are prevalent in Army-related searches? 
How is the number of relevant searches related to the number of people who enlist? 

This report assesses empirical applications of web search data and discusses the prospective value such data can offer Army recruiting efforts. The authors examine three different tools — Google Trends, Google AdWords, and Google Correlate — that can be used to access and analyze readily available, anonymous data from Internet searches related to the Army and to Army service. They found that Google search queries can inform how interest in military careers has evolved over time and by geographic location and can identify the foremost Army-related concerns that potential recruits have. Moreover, by analyzing how search terms correlate across time, it is possible to predict with reasonable accuracy what non-Army related terms people are searching for in the months before or after an Army query. These queries serve as leading and lagging indicators of army-related searches and can offer a glimpse into the concerns of individuals near the time period when they are considering joining. The results suggest that search terms can serve as an indicator of propensity and can be incorporated into models to predict highly qualified Army accessions.

Key Findings

Google search queries can be used to better understand how interest in military careers has evolved over time and geographic location. 
It is possible to use these tools to identify the chief Army-related concerns that potential recruits have, including the qualifications for, procedures for, or benefits of enlisting. 
It is possible to predict with reasonable accuracy what individuals were searching for months before or after searching for Army-related terms. 
Including Google Trends terms in a model of factors influencing the number of Army accessions increases the predictive power of the model. 
Recommendations

Further research is required to better understand ways of using Internet search data to inform Army recruiting initiatives, with the eventual goal of drawing causal inferences and making accurate predictions. 
One promising path involves combining the Google search data with a secondary data source, such as the U.S. Census Bureau, which has information about the percentage of the military-aged population within a geographic region. 
In the same way that companies such as Amazon and Netflix recommend products or movies to individuals based on the behavior of similar shoppers, it is possible for the Army to share queried information with individuals based on what other similar individuals find most useful.

No comments: