How to use data analysis and algorithms
Data plays an increasingly important role in many business areas – so much so that many companies boast that they’re “data-driven” organisations.
Data-driven insights can enable companies to reduce bias, improve consistency, uncover hidden opportunities, and operate more efficiently. Businesses typically apply
these data-driven insights in areas such as customer relationship management, supply chain optimisation, product development and digital marketing.
Increasingly, businesses also use data-driven insights and algorithms to improve hiring decisions.
To help you understand how algorithms in HR work, here’s a snapshot of the state of this technology.
Data analytics and algorithms can inform recruitment, advancement and remuneration decisions. Most of these decisions are prediction problems. You’re predicting which candidate will perform best in a job; or what remuneration will retain an employee’s services.
Statistical algorithms can solve prediction problems powerfully. Whether you’re pinpointing which customers are most likely to switch to another company, or which released prisoners will most likely skip bail, statistical algorithms help decision-makers overcome psychological biases and make more informed choices.
However, it’s important to note that to create a statistical algorithm you need a substantial amount of raw data to analyse. If you’re using your algorithm to make predictions about an individual’s workplace performance, you need more data than is typically available from an external candidate’s LinkedIn profile and CV. Also, algorithms need time and substantial data sets to prove they work effectively.
Which is why it’s mainly larger companies, looking to recruit internally, who use statistical algorithms for recruitment decisions.
Also, remember that laws such as Australia’s Privacy Act and the EU’s General Data Protection Regulation (GDPR) govern how personal data can be stored, transmitted and used. Plus there are limits on the kinds of data that employers can collect, and how staff must be informed when it’s collected.
Network analyses are a common HR application of company data. For example, when similarly-qualified internal candidates vie for the same internal promotion, network analysis can support decision-making. These analyses generally require staff members’ permission. Without delving into the written content, the network analysis examines email and meeting history, which reveals who these candidates have been in touch with over the last, say, six months. This kind of analysis allows companies to compare their candidates’ connections within a company and which candidates keep company information flowing.
Larger companies with strong technical capacity may create “people analytics” tools such as algorithms. However, algorithms and data analysis for HR come with important provisos about their application. For example:
Research shows managers may exhibit a behaviour called “algorithm aversion” – even when an algorithm has proved to work successfully, some managers still prefer to rely on their intuition.
It’s not only managers who are sceptical of algorithms – applicants are too, and a majority say they don’t want to apply for jobs where an algorithm makes the decision.
People trust algorithms more completely when they understand precisely how they work. Managers need training to use a new HR algorithm confidently.
Algorithms can also incorporate bias – they rely on historical data (such as CVs and performance reviews) that already incorporates bias.
It’s imperative that individual managers own their hiring decisions and can explain why they’ve made a particular hiring decision. Therefore, it’s best to view statistical algorithms as a way to complement a manager’s decision-making, rather than replace the decision-maker. Using statistical algorithms in HR can help uncover hidden internal candidates, but it’s best for the manager to retain and own the ultimate decision on who gets the job.
Talk to a specialist
To elevate your HR recruitment practices and outcomes, and bring a 360o approach into the heart of your business, talk to 360HR today +61 2 9819 6324 | firstname.lastname@example.org.
To continue reading, click on the links to the right or