Talent Management solutions are designed to improve HR areas such as performance management or career and mobility management. Most talent management solutions include artificial intelligence to help match demand and supply at skills and to individualize career support.
The 2021 barometer of the ANDRH nevertheless shows a relatively modest equipment in software: 30% of companies with 300 to 5,000 employees, and 62% of companies with more than 10,000 employees.
What improvements does artificial intelligence have to offer to serve talent management?
What improvements can be made to Talent Management tools?
The improvements that HR teams need today in their daily tools concern superior pedagogical support on the possible uses, and the reliability of the data on which they make decisions.
Support in the use of talent management tools
The innovation represented by artificial intelligence incorporated into talent management solutions first requires change management and sustained training for HR users, employees and managers.
Software publishers are responsible for conducting regular workshops during the project and after deployment. Linnet Kotek, Senior Consultant at Tempo&Co, insists on the various training methods (motion design videos, guided tour), communication and measurement of the tool's adoption to provide the necessary support to users. In order to support users, Neobrain offers a 21-day course: 88% of our users are active after 2 weeks.
Pedagogy around the use of artificial intelligence for the HR domain
The perception of AI in the HR sphere involves a number of ambiguities. While it is a given that the number of tasks performed with the help of algorithms will rise from 30% to 40% in the near future, resistance to change persists among HR managers, who are afraid of losing control of tasks they value. The purpose of AI is not to automate jobs, but tasks. In fact, 74% of employees are in favor of having an assistant provide this service. The results of several studies show expectations centered on suggesting the right training schemes (78%), evaluating skills (71%) or even career management (68%).
We are giving you free access to the Replay of the Webinar"Talent Management: how to do better with less ?"
We will also outline a framework you can adopt to optimize your investments.
HR Data Reliability and Scalability
The volume and reliability of the HR data required for the AIself-learning process is essential. Machine learning, which is at the origin of analytical, predictive and individualization improvements, produces relevant results after an initial "training" phase. Therefore, in order to successfully implement a talent management solution we recommend 3 steps:
- 1ststep: Carry out a first phase of recovery of existing HR data (repositories of skills, business repositories, performance data, results of annual interviews, ....)
- 2ndstep: Guarantee the homogeneity of HR data formats from different sources (excel files, HRIS core, best of breeds, ...)
- 3rdstep: Build a 2-month training model from an initial sample of HR data (addressing the population with whom the talent management solution will have the most impact first is often recommended).
To deliver real benefits in talent management, the use of AI is a step-by-step process.
Artificial intelligence answers to Talent Management
Artificial intelligence in the HR function, or in any other field, consists in having a machine imitate the cognitive abilities of a human being. It must be considered as a form of intelligence that becomes relevant with practice. In order to demystify its functioning, it is crucial to understand that the human being at the origin of the instructions that are initially provided to the AI has a key role in its future relevance. Thus, an organization with multiple and diverse perspectives is better equipped to generate an artificial intelligence that is neutral of any judgment, and efficient in the results it produces.
Artificial intelligence carries as many ethical risks as equal opportunity opportunities depending on its configuration. Each human acts according to cognitive biases. AI can reduce them as long as the diversity of thought of the team at the origin of this intelligence is real.
In order for Talent Management solutions to be more widely used, artificial intelligence will have to convince and evolve in three areas:
- The ability to learn faster (active learning)
- The ability to understand unstructured language (BERT)
- The ability to model the transferability of a skill
Artificial intelligence must learn faster (active learning)
Artificial intelligence needs training to become efficient (notion of"Machine Learning"). This technology will evolve on its ability to learn with little user feedback thanks to "data labeling". The very high consumption of data to learn cases not encountered in the past is now facilitated by a tool like Amazon Mechanical Turk. The labeling technique consists of outsourcing the annotation process to third parties to gain speed.
- Talent Management and training: to multiply a human "task" such as tagging skills on training or building a succession plan.
- AI adapts training programs more quickly, thus accelerating the adaptation of your employees' skills .
Artificial intelligence must understand unstructured language (BERT)
Unstructured language consists in understanding unstructured data to make it intelligible. This unstructured data is very common for HR, it is qualitative information that requires a lot of collection and processing. For example, feedback from an employee survey, a job description, .... Google's BERT technology continues to progress in understanding the meaning of words in order to classify and exploit them.
- Recruitment and training: be able to understand a job offer to extract the skills, constraints, imperative qualification levels and treat the applications objectively.
- Take from a training presentation the skills that will then be integrated into the employees' profiles.
Artificial intelligence must model the transferability of a skill
Artificial intelligence still needs to progress in its ability to predict future developments. Today the technology is able to anticipate future realities from past elements, the next step is to be able to reinforce these predictions from different scenarios. For example, determining how long it will take to acquire skill B, if I already master skill A.
- Make the employee an active participant by optimizing suggestions for training, job offers and professions.
- Optimize recruitment: identify candidates with similar skills
The development of artificial intelligence goes hand in hand with the transformations experienced by HR. Among these fundamental changes, AI can help facilitate decision-making, make employees actors in their careers, and promote a learning culture. The Gartner panel reveals that 13% of companies are using an HR solution with AI, and two-thirds of them intend to double the scope of their projects. The expected objectives are as follows:
- 62% of them are looking for decision making based on reliable data
- 57% see it as an opportunity to improve the employee experience
- 56% are looking to automate repetitive tasks
- 51% believe they will achieve substantial cost reductions.
The European Parliament report of February 10, 2020 speaks of"automated decision-making", which challenges the concept of AI as a simple tool and goes beyond our recommendations. AI calls into question the organization of work and the management of employees. It will be a subject of negotiation that will tend to become more widespread in the years to come.
Why is talent management becoming essential?
There are several reasons why talent management is essential and must be integrated into talent management policies:
- CEOs now see this area as a source of significant contribution to the company's economic performance.
- The talent management policy is no longer limited to the individuals with the most potential, but to all employees.
- These subjects are the subject of a significant desire for training within the HR function: 7% say they are "experts" in anticipating the management of skills, a figure to be compared with recruitment, where 25% have this level of expertise.