The Future of Talent Management
While other departments (marketing, finance and supply chain management) have embraced data-driven methodologies for generating business-critical strategic decisions, HR management (and, by proxy, talent management) has lagged behind (Harris et al., 2011).
In the era of big data, talent management has finally become the subject of mathematical modelling, which requires a large body of credible data in order to function (Ashton & Morton, 2005; Davenport et al., 2010; Lawler, Levenson & Boudreau, 2004). Indeed, the HR department usually holds a record of all personnel from the date potential candidates apply for positions in the organization, through to their promotion and development until they leave employment. This longitudinal dataset is ideal for data mining and modelling, and there are now software applications on the market which attempt to manage talent within the organization (e.g., Oracle Embedded Analytics Solution and SAP Success Factors Workforce Analytics). This analytics software trawls through HR data and compiles a list of the causes and effects of employee turnover. Thus, they can forecast which personnel are at risk of leaving by correlating the different factors discussed in this chapter and offer solutions for talent retention. For example, factors such as a salary, career development opportunities, commute distance and relationship with the manager are assigned weightings in risk level such that if the total risk reaches a certain level, the software will trigger an at-risk alarm so that HR and senior management can consider remedies for retention (Harris & Craig, 2011). More sophisticated modelling can analyse effective and ineffective career paths (i.e., those that will retain talent and those that will cause them to leave; Davenport et al., 2010). To date uptake for talent analytics applications has been taken by relatively few organizations (e.g., Google, Convergys and AC Milan; Harris et al., 2011) but it is expected that more organizations will be using their personnel data more productively and effectively to obtain predictive insight for talent management in the future (Bersin, 2013; Corsello, 2012). It is important to note that since talent analytics is a model of the potential factors affecting the lifetime of an employee and relies on large datasets to run correlations and look for causality, the onus is on providing reliable data - as the jargon goes, ‘garbage in, garbage out’. Updating the dataset with newer external data can make the analysis richer and predictions more accurate, leading perhaps to the creation of a new type of talent management consultant: the talent modeller.