A performance management system is a process driven approach to measure employee performance both quantitively and qualitatively. When done effectively the organisation can also detect skill and competency gaps of individuals or a population of employees and thereby organize training and mentoring for overall development.

Classical performance management system, an annual exercise, has been a 5- or 7-point rating on employee performance and attitude on various parameters marked by the direct supervisor and then validated by a senior functional manager and Human resource department. One of the limitations of the process is direct supervisor, when unsure, will mark a median grade that may not be reflective of the actual performance. Also, appraisal interviews are a source of discomfort for both the employee and the supervisor when a difficult discussion of employee weakness comes up. Also, the process being annual, the recency factor of employee performance over last couple of months would be the main decider.

With the advent of various digital tools at the turn of the century many of the paper-based practices have been shifted online resulting in scalability, faster turnaround and to certain extent avoidance of a face-to-face interactions. However, the annual process and the recency effect continued, leading to a certain level of distortion in the entire performance management system.

The development of data sciences and AI has addressed some of the concerns of a traditional performance management systems. Masson S. (2020) outlines some of the ways through which an AI powered system would help.

A Framework Driven Approach: This makes for easier recognition of employee efforts and performance by systematically organizing data.

Natural Language Processing and Sentiment Analysis: It helps to understand employee interactions with others and their reactions to the concerned employee efforts.

AI based algorithms continuously look at employee performance at real time and eliminates the annual ritual of appraisal.

Algorithms also track employee digital footprints and interactions over various digital tools like mail, calendar, slack etc. and integrate relevant information scattered across these applications to analyse performance.

Some advantages of an AI based performance management system would be as follows.  

Elimination of human error like personal bias, incomplete data compilation and acts of voluntary or involuntary omission and commission.   

Availability of comprehensive data to judge performance from various scattered sources.   

Ability to do a real time analysis of performance on ongoing basis to judge true performance and eliminate the recency bias.   

Better understanding of skill gaps based on data. Algorithms today can analyse data faster and give recommendations to managers that help in decision making.     

The real time analysis also helps employee to understand their limitations and achievements and help the managers to remove roadblocks to ensure better performance.

The deployment of AI system also has certain disadvantages.     

It requires substantial investment in terms of cost and time. A well-functioning machine learning algorithm requires substantial time to perfect and thereby a cost benefit analysis requires to be done before deployment.

The lack of human element of a performance management system using AI is a major limitation. It will take away the elements of empathy, emotional intelligence and cooperation and reduce performance management system as an output of data driven activity.

There are ethical concerns of tracking employee actions in real time over various digital applications. It raises the questions of privacy and intrusiveness and thereby may erode employee trust and morale.

AI system in performance management should bring in more transparency, aid accurate data-based assessments and remove biases including recency bias. However human touch will always remain the focal point of any performance management systems. AI should be an aid in decision making and not the decision maker. This would free up time for line managers for quality decision making and accelerate the overall performance of the individual employee and the team.

Aspan Maria (2020). This tech giant says A.I. has already helped it save $1 billion.