Powering Performance Management with Gen AI

Traditional approaches to Performance Management typically involve annual, or at best quarterly, reviews to evaluate set objectives and employee performance against them. Challenges include lack of or dated performance feedback, significant time spent collating and chasing feedback, reliance on highly subjective input and that objectives set at the outset rarely remain current throughout the entire performance period. Managers and employees often feel burdened by the process, and that’s before the opportunity to have the right conversation!

Here's some ways in which AI can help address the challenges

• Timely, admin-light feedback – using AI’s data mining and analysis capabilities, information from a host of sources can be used to generate summarised feedback, for example from emails and other internal applications. No longer something that only happens annually, information can be generated on demand quickly, saving significant administrative time and helping to build a progressive view of performance over time. This ongoing feedback approach ensures that issues are identified early, and associated changes made.
• Less subjective, richer picture – as data can be extracted from multiple sources, a broader set of inputs informs performance evaluation. This minimises ‘point in time’ feedback, or particularly personal views, adversely influencing the overall evaluation.
• Targeted career and development planning – AI can be applied to review overall performance history to inform a view of employee skills, experience and gaps for their current or future role/career path. Algorithms can also be used to identify development interventions to address any gaps – imagine how such data can inform a targeted and effective manager/employee conversation!

Points for Consideration

• Data and data usage – as with human processes, AI is reliant on good, accurate and objective data. It’s not a magic solution that takes away from the ‘rubbish in, rubbish out’ principle. The benefits are realised for those organisations who regularly cleanse and review their data sources and wrap appropriate controls, risk protocols and security standards around use of that data. Specifically, given the sensitivity of performance related data, regular monitoring of AI generated information must be undertaken, to ensure accuracy, fairness and adherence to policy/standards.
• Employee buy in – for what is typically a very ‘human’ process, many employees will be concerned about use of AI in performance management, as well as access to some data sources. Involving some employees in the definition and design process is critical to reassure the wider organisation of the role of AI algorithms, the data sources accessed and how the data will be used, and the governance process around use of AI.
• The right conversation – Given the role that AI can play in reducing administrative activity, the outcomes of performance related conversations will only improve if managers have effective engagement with their employees. This means using the information to agree actions required to meet goals, understand the employee perspective and agree interventions that will support achieving those goals. Managers, in turn, must be given the coaching and development to equip them to have successful performance conversations.

In summary, organisations can realise significant benefits through AI adoption; however, trust in the target application will have to be built through effective engagement and robust management of the associated risks.