Gunnar Brune from the AI.Hamburg network spoke with Andreas Wartenberg, Chairman of Horton International and Member of the executive board of Hager Unternehmensberatung.
Gunnar Brune/AI.Hamburg: Andreas Wartenberg, what is your business and how do you use artificial intelligence:
Andreas Wartenberg: Hager Unternehmensberatung, Partner of Horton International is one of the largest executive search companies in Germany with 100 permanent employees. We are focused on top and middle management positions. To fill a position, we do direct search. For us, that is all about finding, evaluating and assessing candidates and presenting the best possible and most suitable candidates for the vacancies to the client. For this it is important not only to be able to achieve a professional and technical fit, but also personality fit: who fits to whom on an interpersonal level and who fits into which company situation – the keyword here is culture. Or: who fits into a particular management task: return on investment, new growth strategies/establishment of new services, business reduction/bankruptcy management. These are very different scenarios that corporate units or business units can get into. No single person has all the skills to solve all tasks. Therefore, the demands on the people we find are often very individual and connected to the company, the teams and the acting persons.
In order to achieve this fit, the use of artificial intelligence is expected to increase in the coming years. There are already some existing applications and more and more new software tools and solutions in which AI is integrated and where different tasks are solved with AI. Therefore it is crucial for us to deal with these topics and technologies. On the one hand, to know what is available, which technologies and solutions, software products and services are adaptable for us or even created exactly for our industry. On the other hand, to find out what added value can actually be generated with them.
Gunnar Brune/AI.Hamburg: Can you give us an example of artificial intelligence you really work with every day?
Andreas Wartenberg: Among the applications that have already integrated artificial intelligence today is the personnel search engine portal LinkedIn. Especially the product “LinkedIn-Recruiter”. This is a very humanized service. It is a tool that personnel consultants and headhunters all over the world use. Through this special search tool you have more search possibilities than the normal LinkedIn licensee. For example, the search results are already pre-sorted by AI. If I enter a keyword like probability of change, AI predicts it and I can sort the results according to it. Such a thing is already working very well with artificial intelligence today.
Gunnar Brune/AI.Hamburg: What is the most exciting AI project of Hager Unternehmensberatung/Horton International at the moment?
Andreas Wartenberg: We have a wide variety of fields of activity. There is not just one project that we are dealing with, because artificial intelligence will develop great innovative power with us. There are and will be new tools and services that make things possible that were not possible before. Let me give you some examples: Let’s take the subject of face recognition. Executive search is not just about identifying a person. We want to know when they are lying, when they are telling the truth, when they are insecure, fibbing, going off the rails or exaggerating. These are all things that can be read in the face, in the facial language. There are very few people in the world who can do this, they do observe features in the face, the movements of the facial muscles: when the eyebrows are raised, when the corners of the mouth are lowered, they always pay attention to the facial expression at certain prominent points on a person’s face in response to certain questions. Today, this is done with a video-supported interview, which is analyzed by AI.
For example we can sit opposite of each other, with a camera recording the face of the person opposite and there is software running there that does the analysis. This can only be done via AI in connection with Big Data – like almost all AI solutions are connected to Big Data. If you don’t have Big Data, then AI is not worth the effort, because you need a big, big data pool for it.
Artificial intelligence will make possible what was previously impossible. It will make it possible to compare many thousands of application documents in detail. Take the cover letter, the choice of words, the sentence structure, the choice of words. They give information about the personality I am dealing with. Is it someone who is confident and can write freely, is it someone who pays attention to formalities, is it someone who has copied a text of a manual, are there many standard sentences in it or is it individual? Is the applicant courageous, does the applicant use daring, i.e. challenging vocabulary or use a lot of question marks or exclamation points in his cover letter? Or acts le applicant like the lamb that offers itself, saying: “I am a loyal follower, that’s why I write the way I write.” Big Data, in conjunction with AI, makes it possible to filter out just that, even with a large number of applications, and ask what it means.
These were two examples of AI in recruitment, another topic is chatbots. We all know the Google example of a hairdresser appointment that is arranged by telephone with a talking chatbot. To a certain extent, this also works for pre-screening, i.e. in the upcoming interview phase. Or during the initial contact with a candidate: Here too, it would be possible to place an intelligent, talking chatbot in the online service and have an initial selection of candidates carried out according to previously defined questions.
Where does that make sense? This is already being done today. A typical scenario would be if you have a large number of applicants for a smaller number of positions. For example, when a large telecommunications company wants to open a new call centre in a geography of this world. Imagine 100 positions for which 800 candidates have to be interviewed. Having this done by a chatbot saves a lot of human resources and of course money.
Gunnar Brune/AI.Hamburg: But surely this is not used in the search for a CEO? Where does artificial intelligence bring added value here?
Andreas Wartenberg: It will probably never be used for CEOs, because we don’t have a situation where we have a large number of candidates, because we do not act in field of masses and volume and therefore have to filter with these technologies. However, there are many technologies that could be relevant and that we may have to consider for what they can do for us. If I can’t use the AI-driven chatbot in the pre-selection of a selection, I might still be able to use it as a global company for standard requests. Companies also receive unsolicited applications en masse. And if we want to talk about the current situation, in a pandemic leading to mass unemployment, many people will be looking for new jobs in all sorts of places to apply and reposition themselves.
With AI-supported tools, we can handle also in executive search an unexpected rush without blocking internal resources. This also applies to companies, of course. If a large car manufacturer lays off a lot of people, people will apply to another car manufacturer, especially if there is a local presence. For example in southern Germany. Then one could consider whether such a chatbot tool would make sense to pre-select.
Gunnar Brune/AI.Hamburg: These are exciting opportunities for the whole world of human resources and executive search. You do executive search. What can we expect from artificial intelligence in the near future?
Andreas Wartenberg: As already mentioned, there is not “the one” project we are currently working on. Rather, we are in a scouting and screening process in which we are looking at various tools. By the way, there are also tools that are reshaping the topic of matchmaking. This is a very exciting area, where AI will be used in various forms in the future. I put a job description in my AI tool and the tool searches independently in different portals using its own AI-driven keywords. Or we can go one step further: I enable the AI-Tool to access my current team for my top management level, from the candidates I know to be managers I am satisfied with. The AI-Tool searches in the portals that exist in this world for other candidates that fit well with my performers and suggests them to me. This is a matchmaking, which is of course necessary and in this new form shortens our processes.
In this area, we work closely with a start-up company to follow this path together. The idea is to put the job description directly into the tool. In a fraction of a second, the tool captures and indexes the job description. Then our database and external databases such as LinkedIn or XING are searched with it. This saves a lot of manual work, because there are often different relevant keyword combinations: Head of Creative, Creative Director, Chief Creative Officer etc. These have to be researched in the manual work in the different sources, which leads to several lists, which then have to be matched again manually. Here, the machine search is much quicker to find a relevant selection of candidates in a single list. Currently, the machine still has to be managed relatively tightly for this, but the process is already much faster and puts the consultant in a position to better advise the client.
But this is only one of several projects. The topics with additional analysis such as face recognition or evaluating cover letters are very important and charming for us because they provide additional information to assess the candidate. In addition to question techniques, interview techniques, CV analysis, references that we can obtain, this is additional information that helps us to assess people. And this is the central point for us: the assessment of people and their ability to fit into companies, functions, teams and cultures, which are already there.
Gunnar Brune/AI.Hamburg: That means you are better at identifying candidates and assessing their specific suitability.
Andreas Wartenberg: At least it’s quicker and it gives additional information about the person.
Gunnar Brune/AI.Hamburg: How mature are the solutions you mentioned?
Andreas Wartenberg: All the companies and examples I mentioned last are companies in start-up mode. We deal with these companies, we scout that, we try to use these things. Specifically, it has to be said that it will be some time before the video-based, AI-evaluated face reading interview becomes standard. The AI-assisted evaluation of cover letters will probably be faster, and even faster will be the topic of matchmaking.
Gunnar Brune/AI.Hamburg: This means that the current AI application is primarily used within the framework of the major search portals. And they actively scout when they can embed completely new solutions into their working methods?
Andreas Wartenberg: Yes and no. There are the big portal operators, e.g. LinkedIn. They will soon offer many of these topics themselves and get them up and running. And there are also many, many HR-Tech start-ups that have many other ideas, where AI can contribute a substantial part. Of course, the question is which of them will remain independent and which of them will be integrated by a large portal operator or another large technology provider.
Gunnar Brune/AI.Hamburg: How far are you involved in the technologies on which these innovations are based. Artificial intelligence uses machine learning to a great extent. Are you involved in training the algorithms?
Andreas Wartenberg: That varies. Mostly not. If you enter development as a partner early on, you are often also a data supplier and thus a “co-trainer” of algorithms. In matchmaking, for example, there is a German start-up with whom we work exclusively to integrate the solution comprehensively with our own database. Then, using the job descriptions in our own database and external portals, we search for suitable candidates in seconds. This is an area where we get involved at a very early stage of the trial. And this cooperation naturally also helps the start-up with this technology – here again the keyword Big Data – to evaluate Big Data in a meaningful way. We are rather reserved when it comes to other topics like video analysis. The problem with our discussions is that the candidates often talk to us very openly about their career, their life, individual wishes, successes and also failures and understandably have little interest in participating in such experiments for new technologies. Such an evaluation of big data and video data is often done remotely at startups in America. We are therefore dependent on these startups developing their tools with candidates from their own cultures and other people or tasks.
Gunnar Brune/AI.Hamburg: How much insight do you have into the mathematical part of artificial intelligence, the selection of algorithms and mathematical models?
Andreas Wartenberg: This is happening outside of us. We don’t talk about the individual algorithms, we talk more about the validity of the results and whether they have been scientifically validated. Which algorithm is behind a face recognition is not relevant for us. What is relevant for us is whether there are other universities or scientists who confirm that it is validated and reliable.
An example from the past: Decades ago, people started using graphological expertise for the evaluation of personalities. But this has never really caught on. Nevertheless, there are competitors who do it regularly. You have to be careful with all methods: Not everything that can be counted counts, but many things that can’t be counted count. Not everything that I can capture with AI and Big Data is relevant, and much of what I can’t capture with them is relevant.
In this context the question is often asked “Will technology ever replace us?”. Will it come to the point where one day there will be no need for executive search consultants because the AI-based systems will solve everything by themselves? No. That will not happen. Because there are also a lot of things that you can’t evaluate at all through AI and that are still important. For example: sympathy. I can work extremely well with great people and achieve success, even though they may not have the best qualifications for this position from their CV and statistical analysis. Maybe just because I want to, because I have the feeling that we all can do that together. These are things that an AI can’t take away from me. Computers can do what we have taught and trained them to do, but that is always limited. We, on the other hand, can think further. Therefore, all Big Data and AI-based procedures in our field will always be complimentary, but never replacing.
Gunnar Brune/AI.Hamburg: Will there be areas of personnel consulting that are characterized to a greater extent by artificial intelligence and areas that have less of it?
Andreas Wartenberg: Artificial intelligence gives us a deeper picture of a person, which allows us to make a better assessment. So it is clear that everything that comes out of the computer should be revised and evaluated again. It is probably the case that the higher the area of responsibility of the position, the more complex the area of responsibility, the more the person will be involved. The lower the position or the more limited the range of responsibility, the more likely I am to be able to map the assessment via technology. The requirements for simple positions, in terms of the individuality of people and the personality traits of people, are somewhat lower. Here, technical and craft characteristics count a little more. The higher up the positions are located, the more important are other qualifications than technical and manual skills. The CEO of an international company does not necessarily have to come from the same industry. That is what I am talking about right now.
Gunnar Brune/AI.Hamburg: Your industry is characterized by people who sell insights in human nature – something German has the word “Menschenkenntnis” for. In the future, will you have to bring together knowledge of human nature and an understanding of – another German term – “Kollege Computer” (coworker computer) in order to arrive at the best recommendation?
Andreas Wartenberg: Yes, we will bring that together. In the future, we will enrich the topic of people’s ability to judge with further information and facts from our colleague Computer. I think we should always have a critical view, even if information comes from an artificial intelligence.
Gunnar Brune/AI.Hamburg: Your company deals intensively with artificial intelligence. What do you recommend to those who are not yet doing so: e.g. a HR manager who does not yet dare to do so, or a personnel consultant who continues to focus solely on personal work?
Andreas Wartenberg: Start and get started. Technological progress is unstoppable. You have to start and see opportunities. Ask yourself: What can this mean and what can it do for me? What you don’t do, others do. That doesn’t mean adapting everything and using everything. It means looking at what opportunities are available to me.
There are many in our industry who say that artificial intelligence is a great danger, which declassifies many in our industry or makes life difficult for many because companies can suddenly find personnel all by themselves. I do not believe that. I believe AI provides us with additional and complementary tools that we are obligated to use well and skillfully. Because then we have an advantage and we have used our opportunities. But we have to start doing that.
Gunnar Brune/AI.Hamburg: What are they planning next?
Andreas Wartenberg: We have two or three topics we are working on. I am also looking forward to the completely new topics that are coming up. As already mentioned, we are already working with various AI solutions. We have a very exciting collaboration with a start-up in the area of matching with AI.
Furthermore, we have formed a scouting team that has been screening the world for more innovative technologies since the beginning of the year: What is out there, what can make sense in our industry, what is already directly intended for our industry? We can already see that very exciting topics are emerging.
Last but not least, we are intensively dealing with the future of personnel consulting/executive search and are working on a study on this topic, to which we are also consulting external experts. We are asking: How do we search? What is being searched for? Who is being searched for? What will this look like in the future? The topic of artificial intelligence will also play a very central role, because the future will not take place without this topic. I am therefore particularly curious about the results of this study.