Will big data allow HR to predict the future?

Will big data allow HR to predict the future?
By Mike Toten on 21 July 2014 The advent of big data will provide an opportunity for HR to exert more strategic influence on management by using the data for predictive purposes, according to a recent UK article.

The article, published on the website Personnel Today, says HR will need to change the way it handles metrics and data, and this may mean more resources will be required for data analysis, eg the use of specialists in the field.

Move from reporting to predicting

The fundamental change would be a move from merely reporting the information collected towards using it to predict future developments. The biggest opportunities will be in high-volume repeatable processes such as recruitment.

Initial steps

The authors suggested many HR functions should already be in a position to take the following steps in relation to recruitment data:
  • Analyse the data collected on previous, current and potential future employees. This will provide a starting point to commence data modelling;
  • Assess the relevance of all data to the target audience of senior management. Which metrics do they really need to see?; and
  • Dig deeper. For example, analyse what the data says about recruitment in specific business units, occupational categories and locations, and what the trends are predicting for each one. Using an ongoing “scoreboard” approach will help to attract management attention.
As an example of the “scoreboard” approach, turnover metrics have traditionally been historical data, ie they recorded what has already happened. They can also be used to predict what will happen in the next quarter, year, etc. A “relative turnover risk” profile score can be compiled for each section.

How to collect big data

Most HR functions have access to HR-related data, such as payroll, absenteeism and turnover records. The article says that they need to become more purposeful in collecting data, and this will often require access to data specialists.

It adds that two types of specialists are required, and are equally important:
  • “Cops” – those with expertise in compliance and enforcement;
  • “Assassins” – those who are agents of change and (positive) disruption within HR. They are the ones who will have to lead a change to predictive analysis, but currently the vast majority of data specialists within HR are in “cop” roles.
Four issues require attention:
  • Volume – ability to handle an ever-increasing amount of data;
  • Velocity – the need to analyse it and act on insights quickly;
  • Variety – ability to handle it in a range of formats; and
  • Veracity – the need to ensure it is accurate and reliable.
HR will need to be able to handle data in both structured and unstructured formats.

Structured format refers to traditional records stored in a defined and orderly way, such as payroll and absence records.

Unstructured data is spread across multiple media and not held together by anything obvious. Examples include social media discussions, video/DVD CVs, etc.

Responses to article

Readers who commented on the article suggested that HR training and degree courses should include content on how to collect, understand and analyse organisational data.

There is the potential risk that otherwise HR practitioners will be encouraged to “spit out” data and predictions prematurely, and if the predictions are often wrong that will undermine the credibility and influence of HR.

Another prerequisite is the need for HR practitioners to clearly understand how a particular business unit functions and is performing financially. Being able to add people-related data to the financial data can then allow a constructive conversation about how HR initiatives can make a positive contribution.

Further information

Big Data: HR Needs to Stop Reporting and Start Predicting”, published on Personnel Today (UK). 


No comments yet. Be the first.