Can the recruitment and training of employees be handled by digital tools? The use of software in human resources

Table of contents

Abstract

Table of contents

Introduction and objectives

1 Short description of the current status of using software and algorithm-based solutions in HR

2 Description of solutions
2.1 Recruiting and Sourcing
2.2 Internal training
2.3 Conclusion of the shown examples

3 Advantages, disadvantages and necessary requirements towards software and algorithm-based solutions

4 Conclusion and Outlook

5 Reference list

Abstract

Digitization has reached the human resources sector. Whereas in the past every application was done by hand, reviewed and the selection of applicants was done completely by people, today there are various providers of software and algorithm- based solutions to automate almost all areas of personnel selection as well as training and further education. On the one hand, all this should save costs and avoid the risk of the position being filled incorrectly, and on the other hand it should be beneficial for both the company and the future employee. The solutions are available. But now only a few processes are completely digital. Company websites and job portals are the most important recruitment channels with almost 60%, but only around 6% of the top 1000 companies in Germany use digital systems for personnel selection.

Solutions such as chatbots and intelligent algorithms can perform standard HR tasks such as answering questions, automatically creating Jon ads and sorting incoming applications. Furthermore, products are offered that can completely analyze the applicant and thus provide the Recruiter with the most comprehensive picture possible.

The solutions simplify many tasks and increase the speed of processes in human resources. However, it should also be noted that an algorithm only carries out statistical analyses and can never draw conclusions like a human being. Furthermore, an algorithm cannot evaluate emotions like a human being. If an algorithm is used, strict requirements must be met in order not to misjudge or discriminate against any applicant.

If HR departments use algorithms and software-based solutions in a targeted manner, they can improve the speed of internal processes and thus gain more time for dealing with applicants and employees.

This assignment deals with the topic of digitization in human resources. Available solutions are presented, and their advantages and disadvantages discussed. Furthermore, a holistic consideration of the advantages and disadvantages of digital solutions and requirements is mentioned.

Introduction and objectives

Digitization is progressing and covering all areas of industry. Personnel is here no exception. Almost all areas of human resources, from the search and selection of employees, administration and accounting to internal training and further education, are now software-supported or analyzed to improved or being replaced by automated programs and algorithms.

The general advantages and disadvantages of digitization in the personnel sector have already been and are being examined and highlighted in various papers and studies. This assignment therefore does not focus on the general view, but rather on the targeted use of the technology and software solutions for individual HR tasks. The assignment is structured in four chapters.

In the first chapter a short description of the current status of the use of software and algorithm-based solutions in human resources is given. In the second chapter, a description of individual solutions for selected topics in the personnel area is then provided. The structure will follow the same order as in real human resources. First, solutions for searching for suitable candidates and for the selection and testing of applicants are analyzed. Then the analysis of solutions and programs for already employed employees for internal training and further education as well as for instructions is carried out. The area of personnel administration and accounting is not discussed here, as these solutions and applications are already established.

When the solution and applications are analyzed, a general description is given, and selected examples are used to illustrate them. In these examples, the advantages and disadvantages of these software or algorithm-based solution are discussed.

The third chapter summarizes the advantages, disadvantages and the necessary requirements of using software and algorithm-based solutions as well as the problems that arise and their solutions.

In the concluding fourth chapter, approaches and further investigations are considered as well as an outlook on a possible further development of the use of software and algorithm-based solutions like Big Data or learning algorithms.

1 Short description of the current status of using software and algorithm-based solutions in HR

The digitization of processes meanwhile affects human resources as well. Whereas in the past every application was done by hand, reviewed and the selection of applicants was done completely by people, today there are various providers of software and algorithm-based solutions to automate almost all areas of personnel selection as well as training and further education. Almost 85% of all applications today are sent electronically by email or submitted via a web portal.¹

Software or algorithms sort and filter the applications and generate in some cases automatic confirmations and replies even before a person sees the application. Once a suitable applicant has been selected, he or she undergoes various tests. Language and behavior recognition, intelligence and stress tests as well as tests of interests are to be used to create a comprehensive picture of the candidate based on which the personnel manager will find the most suitable candidate for the vacant position. On the one hand, all this should save costs and avoid the risk of the position being filled incorrectly, and on the other hand it should be beneficial for both the company and the future employee.

The solutions are available. But now only a few processes are completely digital. After a survey by the monster.de corporation although company websites and job portals are the most important recruitment channels with almost 60%²,but only around 6% of the top 1000 companies in Germany use digital systems for personnel selection.³ Compared to the last survey of monster.de, the number is even declining.⁴ One reason for this are the stricter rules on data protection, as the systems actively use a variety of online sources and the function of the algorithms cannot be reliably monitored.⁵ However, increased growth is expected in the use of software and algorithms for special issues such as advance consultation via chatbots.

If the employee then works in the company, a constant further training is necessary due to constantly changing task areas and activities. Due to the ever-increasing cost pressure and the problem of finding a date for face-to-face training, this area is now also being digitized. Training and further education as well as instructions take place by App or at the computer. Thus, the employee can freely decide when and at what speed he or she completes the training or instruction. Digital solutions are also available for this and the market is growing strongly. Up to 2023, market growth of up to 10% per year is expected for the industrial applications segment alone.⁶

Since it is generally expected that digitization will continue to grow despite the current low level,⁷ selected examples of some systems already on the market will be examined in the following.

2 Description of solutions

2.1 Recruiting and Sourcing

In the area of recruiting, there is enormous potential for the use of algorithm-based solutions such as bots or chatbots for communication.

Job Vacancy

Before a job vacancy is published, various people must first be asked about the exact requirements. Once the job advertisement has been created, it must be published on the company website and on various portals such as monster.de, Xing or LinkedIn. A person must invest for this an enormous expenditure of time, which is not available for the genuine personnel work. This activity can be done by a bot.⁸ The employees involved can inform the bot of the requirements and the bot creates a draft for a job advertisement based on the specifications. After approval by humans, the bot publishes the job advertisements on the individual platforms.

Research for candidates

If the company itself becomes active, the recruiters usually go to job platforms such as Xing and Linkedin to find suitable candidates. The chances of finding a suitable candidate are high. XING and LinkedIn together have almost 27 million members.^{9 10} But this takes time. According to a study of Smartrecruiters, recruiters spend nearly 63% of their time in research, screening and messaging.¹¹ Here, too, a bot can provide support. The recruiter gives the bot the requirements and required qualifications and let it search the platforms. If suitable candidates are identified, the bot informs the recruiter and possibly also takes over the first request to the candidate.

Communication with candidates

As mentioned above, recruiters spend 63% of their working time on search, screening and communication. This communication can also be taken over or supported by a bot, in this case a chatbot. According to the Clustaar Chatbot Platform, almost 65% of all requests can be handled by a chatbot¹², since most questions are of a general nature and often expect similar answers. Thus, the recruiter can be relieved of work that he can freely use again without any disadvantage for the applicants.¹³

Selection and preselection

If suitable candidates are found, an algorithm can preselect or sort the applicants. Furthermore, the automatic creation and sending of an acceptance or rejection can be created, as well as the coordination of possible job interviews. Especially in the area of one-click applications, i.e. the type of application in which an applicant only transfers his data from an existing online network such as Xing to the company database, such a bot is helpful. The bot can sort the data records received and match them against the specifications of the personnel department. The HR manager only enters the process when a person becomes necessary and a bot or algorithm can no longer be used.¹⁴

Testing

Once a candidate has been selected, the company naturally wants to learn as much as possible about the person and his or her motivation. In addition to the already established personality tests and skills tests, software offers new possibilities. By using software, it is possible to assess the candidate's attitude and motivation based on the analysis of the CV texts or cover letters.¹⁵ The same procedure can be used for speech recognition. The algorithm analyses the person's choice of words and pitch and then gives an assessment. The company Precire¹⁶ has developed a software for that purpose that can be implemented in already existing personnel environments and can also be integrated into telephone communication during telephone conversations between the personnel department and the applicant and gives the personnel manager a live assessment directly.

Another special form is the skill testing for special tasks or jobs. E.g. for pilots it is necessary to know their behavior and skills even in stressful situations.¹⁷ Hereby software solutions can be used. They simulate special situations and measure for example the reaction time, the number of tasks completed in a time period and the number of right and wrong decisions within a time period. Combined with a human tester this test results can predict very precise if a person is suitable or not for a job or a task.¹⁸

2.2 Internal training

If the person is then employed in the company, the tasks of the personnel department do not end here. The employee must be instructed internally in accordance with legal requirements on various topics such as occupational health and safety and must continue to receive ongoing training in order to keep up with the constantly changing requirements. However, these measures should not disrupt ongoing business operations and should incur as few costs as possible as a result of the employee's absence, e.g. for classroom training.

Whereas this was not technically possible in earlier years, today there are various solutions for such an implementation. These range from simple training courses on the PC or via app to complex learning management systems, such as from Adobe¹⁹, where the entire training system within a company can be mapped. The employees themselves can freely decide when and how they want to complete a training course and can also freely divide up the learning pace. Managers can plan their employees' further training digitally and see which employees require more training or specific training by means of an analysis.²⁰ This eliminates the need for complex organizational processes to train employees together in one place, which also optimizes costs.

2.3 Conclusion of the shown examples

The examples presented are intended to optimize the work of HR departments and focus more on the tasks of HR departments where people are essential because of their skills, such as interpersonal understanding. All manufacturers or suppliers of these solutions merely state the advantages of their products and solutions. But one factor is rarely mentioned. Human resources is about human beings, not products.

It is therefore essential to consider both the advantages and disadvantages of these solutions and, above all, what quality and safety requirements must be met by such products and solutions.

3 Advantages, disadvantages and necessary requirements towards software and algorithm- based solutions

The most frequently put forward argument for the use of algorithms or software-based solutions is the increase of efficiency.²¹ Human resources staff can transfer simple standard tasks, which would otherwise be time-consuming, to algorithms. Thus, they can dedicate themselves to tasks in which human skills are necessary, such as conducting conversations, assessing employees and developing employees. This is also made possible by algorithms and software. According to a Monster.de study, currently only four out of ten applications are intensively examined by the personnel departments. Further it takes on the average up to thirteen days until it comes from the arrival of the application to a first contact.²² The implementation of such solutions can thus provide just this time to carefully examine all applications if possible.

At the same time, the use of such solutions should make the recruitment process free of personal prejudices and discrimination.²³ Applicants should only be selected according to the matching rate of their qualifications against the required requirements.

But an algorithm or a software is a bunch of code lines and is only comparing variables and does statistic analysis. A person, on the other hand, is capable of drawing conclusions and interpreting things. So what criteria are necessary to makes sure that an algorithm is helpful, choosing the right candidates and does not discriminate persons?

If an algorithm is to be used, it must meet the following criteria: validity, reliability, objectivity as well as the secondary criteria acceptance and fairness.²⁴ This requires a solid database against which an applicant is assessed. A pure decision in the sense of passed or failed is not permissible here, since no person ever completely fulfils all criteria. It is therefore necessary to transfer the criteria of the already existing suitability diagnostics according to DIN 33430 to such an algorithm.²⁵ This is the only way to ensure that his algorithm delivers the results that a human resources manager needs to support his decisions. Another task is to make sure that all data protection rights of the applicant are protected. The Recruiter must know where the information is from and if the information is legal to be used. Even in this field there is still a lot of work to be done, because one thing an algorithm as a tool cannot do: to evaluate the consequences of a wrong decision.

[...]

¹ cf. Weitzel et al. (2017a, p. 4).

² cf.Hays (2019, p. 8).

³ cf. Weitzel et al. (2018, p. 11).

⁴ cf. Weitzel et al. (2018, p. 11).

⁵ cf. Weitzel et al. (2018, p. 11).

⁶ Research and Markets Ltd. (2018).

⁷ cf. Weitzel et al. (2017b).

⁸ cf. Hundertmark (2018).

⁹ XING (2019).

¹⁰ LinkedIn (2019).

¹¹ cf. Cammack and Lavi (2018, p. 13).

¹² cf. Clustaar (2019).

¹³ cf. carerix.com (2019).

¹⁴ cf. Weitzel et al. (2017a, p. 13).

¹⁵ cf. ARTISANTALENT.COM (2017, p. 8).

¹⁶ cf. PRECIRE Technologies GmbH (2019).

¹⁷ cf. Arendasy, Sommer and Hergovich (2007).

¹⁸ cf. SCHUHFRIED GmbH (2019).

¹⁹ cf. Adobe Systems Software Ireland Limited (2019).

²⁰ cf. McGovern et al. (2018, p. 3).

²¹ cf. Weitzel et al. (2017b, p. 8).

²² cf. Weitzel et al. (2017a, p. 10).

²³ cf. Weitzel et al. (2017b, p. 8).

²⁴ cf. Rögner (2019).

²⁵ cf. Testkuratorium (2018).