The one perfect job title? There’s no such thing, is there, Stephan Schmid?

Stephan Schmid from Jobvector knows a thing or two about job titles – data scientist, statistician, computer scientist or perhaps AI developer after all? – Four different job titles can ultimately describe quite similar skills. For recruiters, this is precisely the problem: a job advert might be a perfect technical fit for someone – yet still be overlooked because the person doesn’t feel at all drawn to the chosen job title. Stephan Schmid from Jobvector is tackling this very challenge. Using AI, jobvector analyses job adverts and candidate profiles based on skills, aiming to match jobs with suitable candidates regardless of the exact wording.

In this interview, we discuss why there can’t actually be ‘THE’ perfect job title, how a traditional job advert becomes a ‘smart job advert’, and why even small differences such as ‘electrical engineer’ or ‘electrical engineering graduate’ can make a difference. Let’s get started:

SAATKORN: Stephan, please introduce yourself briefly to SAATKORN readers. And what exactly does jobvector do these days?

jobvector is a job board for qualified specialists and executives. Originally, we had very strong roots in the STEM sector. In recent years, however, we’ve continued to develop our own AI technology for skills-based applications, thereby broadening our scope significantly. A major focus for us today is skills-based recruitment: in other words, the question of how we can match a role, based on the skills actually required, with the people who possess those skills.

SAATKORN: You talk about the ‘smart job advert’. What makes it smarter than a normal job advert?

A smart job advert is empowered by AI technology to do more than a traditional job advert. The first step is identifying the target audience. Our technology reads and analyses the content of a job advert and attempts to understand the information it contains within its professional context. This includes, for example, the job description, the role profile, the department and, above all, the relevant skills. On this basis, a skills-based match can then be made between the role and suitable candidates.

SAATKORN: So does that mean the job title alone no longer determines who is suitable for a role?

Exactly. That’s actually the essence of skills-based matching. People can be suitable for the same role even though their previous job titles are completely different. For example, if I have a vacancy in the field of data science, depending on the knowledge required, a data scientist, a statistician, a mathematician, a computer scientist, an AI developer or a big data engineer could all be suitable candidates. What matters isn’t whether their CV lists exactly the same job title as in my job advert. Anyone with the relevant skills may be a good fit for the role.

Skills rather than job titles: Who is really suited to the job?

SAATKORN: How does your AI actually identify which candidates possess these skills?

Of course, both sides need to be considered. On the one hand, we analyse the job advert and try to understand as precisely as possible the technical requirements it entails. On the other hand, candidate profiles are created based on various signals. These can include, for example, CVs, job seekers’ search queries and their behaviour when looking for work. Demographic data plays no role in this. It’s all about qualifications. This information can then be matched against the analysed job adverts.

SAATKORN: At the same time, you say that the job title, of all things, is one of the most important performance drivers for a job advert. Isn’t that a bit of a contradiction?

No, not at all. The job title is extremely important for attracting attention and ensuring the advert is found. It often provides the initial impetus to open a job advert in the first place.

The only problem is: there isn’t one single ‘perfect’ job title that works for all suitable candidates. If I know which different people are professionally suited to a role, I can also address them in different ways. And that’s exactly where it gets interesting.

Why there is no such thing as THE perfect job title

SAATKORN: So the answer to the age-old recruitment question ‘What is the perfect job title?’ is actually: it doesn’t exist?

Exactly. – It always depends on who’s reading the title, where they see it, and even what device they’re using to read it.

Is the person currently actively browsing a job board? Are they searching on Google? Are they on social media or another website, and perhaps only vaguely considering a change of job? And are they viewing the title on a desktop or on a smartphone? – All these factors can influence which job title works best at that precise moment.

SAATKORN: You gave a lovely example of this in your talk, using the analogy of a parcel. Could you explain that again?

You can think of a job advert as a parcel containing something really interesting. The contents of this parcel might be a very good fit for different people. But if the wrong recipient’s name is on the parcel, it might not even be opened. A statistician might be the perfect fit for a role in terms of her expertise. But if the ‘parcel’ simply says ‘computer scientist’, she might think: ‘Are they even talking to me?’ If I address the same person using a title that suits them, the likelihood that they’ll even look at the job increases.

SAATKORN: And that’s why you change the job title depending on the candidate?

When displaying a smart job advert, the title can be personalised. Naturally, we don’t simply alter the company’s original job advert without asking. But when we reach out to suitable candidates via different media channels, the technology can select an appropriate title variation. For example, a rather general ‘Project Manager’ can become an ‘Electrical Engineer as Project Manager’ for a suitable target group, and the title can also be supplemented with a relevant specialist field.

Different target group, different title, different channel

SAATKORN: How precise is this personalisation? Does using ‘Electrical Engineer’ rather than ‘Electrical Engineering Graduate’ really make a difference?

Such nuances can indeed be relevant. From a technical perspective, ‘electrical engineer’ may be a very accurate term. However, if people search for ‘electrical engineer’ far more frequently, this search behaviour should be taken into account. Of course, it’s even better if we know exactly what a specific person has actually typed in. Then the way we address them can be tailored precisely to that. That might sound like a minor detail. But it’s precisely these performance factors that can determine whether a job advert appears higher up in the results, or indeed whether it’s noticed and clicked on at all.

SAATKORN: And does the platform also play a role?

Absolutely. Every channel has its own rules, its own search algorithms and, in some cases, different character limits. A headline that works on one channel therefore won’t necessarily work just as well on another. The technology can take into account which information is particularly important in which medium and which headline variant makes sense there. This applies, for example, to job boards, Google, social media, partner networks or other advertising formats.

SAATKORN: So far, we’ve mainly been talking about active job seekers. What about people who aren’t even using a job board?

That’s also an important point. A person may be a very good fit for a role in terms of their skills, even though they aren’t currently actively searching on a job board. Perhaps they’re currently browsing social media or another website on the internet. If this person has been identified as a suitable candidate, a job advert can also be displayed there – again, with a message tailored as closely as possible to that person. This means we reach more than just those who are currently actively searching for a specific job title.

SAATKORN: And once the application has been received, is the AI’s work done?

Not necessarily. The matching process can still be helpful afterwards. Once applications have been received, the relevant information can also be utilised within ATS functionalities. For example, the AI can provide insights into how well the applications received match the role. In this way, the technology can not only help recruiters draw attention to suitable candidates for a role, but also provide additional information during the subsequent process.

Over 30 per cent improvement in performance

SAATKORN: What are the actual benefits of personalisation in the end? Do you have any figures on this?

Our data shows that, on average, customising the job title using the smart job advert results in an increase of just over 30 per cent. To us, this clearly demonstrates the impact this personalisation can have on the performance of a job advert.

SAATKORN: If recruiters were to take just one idea from our conversation to apply to their next job advert – what should that be?

Don’t assume there’s one perfect job title that will appeal to all suitable candidates. A role can be relevant to people with very different professional backgrounds and job titles. If we focus more on skills and understand how the right people actually search for jobs and want to be approached, we can broaden the target audience profile – whilst communicating in a more relevant way.

SAATKORN: Stephan, thank you very much for the interview!

Further links

• Stefan’s full presentation is available HERE in the 4INSIDER media library.
• Fancy joining the next SAATKORN Online Conference live? You can register HERE.
• Stephan Schmid on LinkedIn
• Jobvector website