These days everyone seems to have a view on recruiting technology. But with the proliferation of new solutions that we are currently seeing, comes an increase in marketing spin and a battle for thought leadership amongst industry commentators which makes it difficult for employers to find the signal in so much noise.
So what is actually going on and what are the real trends that people need to be aware of? Who better to guide us through than someone who is responsible for actually building some of the software itself. Hessam Lavi is both the founder of Jobspotting and the Director of Product for SmartRecruiters.
In the interview we discuss:
• What is happening in recruiting technology in 2018
• The critical talent acquisition issues that technology is attempting to address
• Recruiting automation and the emergence of automated sourcing
• The evolution of recruitment advertising
• The vital importance of data
• The potential for bias in machine learning
Hessam also gives us his take on the future of recruiting technology and explains why recruiters should be optimistic about the evolution of their role.
Subscribe to this podcast in Apple Podcasts
Transcript:
Matt Alder [00:00:00]:
Support for this podcast comes from Smart Recruiters, the hiring success company Smart Recruiters is an enterprise grade talent acquisition suite designed for hiring success. Move beyond applicant tracking with a modern platform that provides everything you need to attract, select and hire the best talent. From candidate relationship management, sourcing and recruitment marketing to screening, selection and offer management experience. A talent acquisition suite with a user experience that candidates, hiring managers and recruiters all love. Companies from Kelly Services to Visa to Bosch leverage smart recruiters to achieve hiring success and expand their business. Visit smart recruiters@www.smartrecruiters.com to find out why companies across the globe consider them the number one ATS replacement.
Matt Alder [00:01:14]:
Hi everyone, this is Matt Alder. Welcome to episode 142 of the Recruiting Future podcast. These days everyone seems to have a view on recruiting technology, but with the proliferation of new solutions that we’re currently seeing comes an increase in marketing spin and a battle for thought leadership amongst industry commentators, which means it’s very difficult for employers to find the signal in so much noise. So what is actually going on and what are the real trends to be aware of? Who better to guide us through than someone who’s responsible for actually building the software itself? Hessam Lavi is both the founder of Jobspotting and the Director of Product for Smart Recruiters. He shares some brilliant insights in this interview, so make sure you’re listening very carefully. Hi Hessam and welcome to the podcast.
Hessam Lavi [00:02:11]:
Hi, good morning Matt.
Matt Alder [00:02:13]:
Great to have you on the show. Could you just introduce yourself and tell everyone a little bit about you and what you do?
Hessam Lavi [00:02:20]:
I’m Hessam Lavi, I’m working as a Product Director at Smart Recruiters. I joined a company around two years ago when Smart Recruiters acquired a startup that I founded in 2013. I’ve been within the HR tech business since 2011 and I kind of got in sort of by accident. I previously worked at Google in Dublin with Search and I decided to move to Berlin together with my wife and landing in Berlin I realized I couldn’t speak German and had struggled to find a job and understand what’s going on in the job market. So I just to help myself and my friends that were also looking to move to Berlin, I started a job board and very simple job board that was in English because there was no other similar site and I just ran that as a side project for a couple of years and it kind of grew organically and now it’s one of the biggest job boards in Berlin. And that kind of got my eyes on the whole recruiting and online industry around jobs and job search. And based on that, I got in touch with a couple of my ex Google colleagues that moved to Berlin and we founded a company called jobspotting, which is still a recommendation tool for job seekers. And then throughout the years that we were running that, we got to know small recruiters who are interested in the technology and of matching and AI and machine learning and how that can help recruiters and job seekers. And then we joined forces back in 2016 and I’m based in Berlin. As I mentioned, we have an office here focused on engineering and yeah, that’s pretty much what I do.
Matt Alder [00:04:06]:
Fantastic. And obviously in that time there’s been a huge amount of change in recruiting that’s kind of being driven by technology. What I’m noticing at the moment is more and more people are talking about recruitment, you know, recruiting automation, and how technology is changing the way that recruitment works. What’s interesting though is not many people are actually talking about what’s happening, you know, what exactly is happening and how is technology changing it. From your kind of perspective as someone who’s helping to develop and create these products, you know, what’s going on, what do people need to be aware of?
Hessam Lavi [00:04:49]:
I mean, generally there’s a couple of problems in the market that are increasingly becoming a challenge. And this is why technology is starting to play a bigger role. One of those things that it’s just becoming harder to find, to find the right talent. There has been some studies which show that more than 50% of all the talent acquisition professionals find that screening candidates are the hardest part of the job. In addition to this, I also see that because of the way online job platforms and recruiting software has changed, they made it increasingly easier for candidates to apply to jobs. That might sound like a good thing, but it has a side effect of recruiters just being bombarded with an increasing amount of application resumes that they need to manually check. So just the problem of signal versus noise is just increasing massively and pushes it down to recruiters to manually manage it. And because of all of those two problems, we also see that bias has become a major concern. And the subjective decisions and actions together with the negative bias has a pretty big negative impact on teams and organizations. And many companies are, are spending a huge amount of time to try to fight bias in recruiting because it’s pretty obvious that diverse teams are generally performing better. And based on those problems, we Obviously look to technology to solve those. And some of the things that we’ve seen happening is of course around automation first and foremost. And automation is really about offloading the grunt work, like things that recruiters don’t want to do. Because to be honest, I don’t believe any recruiter got to this business to read resumes all day long. I think it’s a people business. Recruiters really like to reach out to people, negotiate with them, get to know them and bring them on board. But because of the increasing amount of workload in terms of screening, they just have a decreasing amount of time to spend on it. And having automation with technology, it really helps us. And this is, we’re not talking about a vision for the future. This is already happening in the market. There are several companies and software that take care of automated screening, automated scoring of candidates, so that recruiters can really start working with focusing on the most promising candidate first. The second part is around automated sourcing. So not only managing the incoming applicants, the inbound into your pipeline, but also the outbound part of it doesn’t only help recruiters to find matching candidates, but it also helps you to rediscover candidates in your own databases as well as outbound, to connect to external databases and analyze the content, figure out who are the best potential candidates are, and even reach out to them through automated way through chatbots and automated messaging to engage with them and bring them on board. So that’s about the actual recruiting part of it. Another part that I find really interesting is about advertising because job advertising has remained highly manual and fairly unscalable. If we’re looking at how advertising happens in other industries, e commerce and basically everywhere else, we see that job advertising is still not performance based and most organizations just still spray and pray and hope for the best. So there is a massive opportunity that we see to apply this modern ad technology to make job advertising performance based, targeted and also automated to really bring it to the next level.
Matt Alder [00:08:54]:
That all makes perfect sense and I think it illustrates just how much is going on at the moment. And as you say, all these things are currently possible. They’re not some kind of futurist vision about where we’re going to be. Just to sort of pick up on one aspect of that. You were sort of talking about automated matching and automated sourcing. How does that actually work in practice? How does a machine get good at knowing what a potential good candidate looks like? And also in terms of bias, can a machine be biased as well? What’s your kind of experience and Thoughts on those?
Hessam Lavi [00:09:41]:
So let’s start with the first, first questions about how this can be done. Generally this is a difficult technical challenge similar to Google and search and other popular search and recommendation technology. This problem is something that would never be solved and we will work on this for years. But generally there are certain things that are new technology available that can make that automation faster. So one of the things we do at Smart Recruiters, we recently launched a product called Smart Assistant. And the way we do it, and I think different companies, different software have different approaches. What we do is that we spend the vast amounts of time on our engineers to actually normalize data and clean up data, because that’s what we see is the biggest wins in terms of every time we receive a resume. We will obviously parse and analyze that profile data. And the biggest challenge there is to make sure that we have a fairly good understanding of the candidates skills, work experience, background, education background, and generally what skills they have demonstrated throughout their career. And this is not very fairly straightforward, especially because we’re dealing with fairly unstructured data. You know, you talk about resumes in PDF formats or word formats or LinkedIn profiles or any other social accounts that we can access to and there is no standard way of describing your experiences. And generally it also comes down to the single individuals because some people are very good at describing what they do in their resumes and frankly, other people are not good at selling them themselves. So you’re dealing with a huge amount of ambiguous data that you need to clean up and normalize first before you can actually do anything smart about it. So we started looking at this problem back in 2013 with the first startup that I founded because we applied similar, similar thought process on job descriptions because we want to recommend the best jobs for the candidates and to be able to do so we had to analyze job descriptions which also suffer from similar problems that, you know, some, some job. Many job descriptions are fairly weak. They’re either stuffed with keywords to make sure that they attract as many candidates as possible, or they can be very, very ambiguous and not necessarily talking about the skills that they require and things like that. So that’s what we’re doing most of the time. A huge amount of effort is going into normalizing the data, finding valuable data sources in the market, bringing it in and trying to enrich our own data set so that every time we see a new candidate, every time we see a new job description, we can very quickly classify it and also additionally enrich it with additional data that would help it to attract the right candidates or attract the right jobs. And based on that, of course, there is no one type of technology that can be applied for the rest of the matching process. We use a variety of different technologies as machine learning, matching, we use natural language processing, deep learning, and there is a fairly complex set of different tools that can be applied to different problems. And depending on the problem that we have, we’re looking at all kind of available technology out there. To move on to your second question about if machines can be biased, and that’s absolutely correct. And this is one of the reasons we’re spending such a huge amount of effort on the top of the funnel in terms of the incoming data to make sure that the bias that are in there within that data doesn’t affect the rest of the process. Because obviously if we would train our machine learning algorithms on detecting patterns based on biased data, it’s absolutely correct that machine learning will also be biased and we will reinforce that negative effect into the process. There are different ways to remove bias. I mean, first of all, it’s fairly challenging and I would rather say impossible to completely remove bias from the process because there’s going to be some degree of market bias in terms of the way jobs are described, the way that we expect candidates to describe themselves in the resumes. That is fairly hard to move away from what we do at Smart Recruiters and within my team is that we trying to expand our data set as much as possible because with a larger data set you reduce your risk of having bias and you’re balancing out the different biased over time. So one of the way we do is, as I mentioned, we started jobsplatting in 2013 and we were active in 12 countries. So we have access to on access to job descriptions for five years in 12 different countries, which helps us to have a fairly good understanding of a broad understanding of similar job descriptions, what those particular jobs are requiring in terms of skills and educational background in different countries that kind of broaden our spectrum in terms of variety of mindsets and views. And on top of that, we are, in terms of matching, we’re really moving away from looking at individual candidates and instead every time we receive an application and analyze it, we try to enrich it with additional data so that we make sure every candidate has a fair chance of being discovered and being considered for a job in addition to it. The last thing I want to say is that I don’t expect software and AI to fully replace the recruitment process. I think there has been some misunderstanding that AI will completely take over their jobs and be able to automatically hire, automatically reject. So if the AI system is biased, it will automatically infuse that. Instead, we see it as kind of an AI assistant, like somebody who takes care of the groundwork. All the manual tasks that a recruiter has to do today so that you leave more time for recruiters to actually do their jobs. And we believe that burdening them to not burdening them with manual work that they do today, it will leave them more time to really learn about a candidate, to really dig deep and reach out to them so that they would not have to fall back to their negative biases, which is also another way to encourage diversity hiring.
Matt Alder [00:17:02]:
That’s pretty interesting. And I think you start to touch on the sort of core of my next question, really, which is, you know, which is about the role of the recruiter. What are recruiters going to be focused on? Are there any sort of precedents from other industries where, you know, technology has sort of massively changed roles? Should recruiters be optimistic or should they be pessimistic about their future?
Hessam Lavi [00:17:29]:
I don’t see any reason for being pessimistic. I definitely, I’m excited about the future. I’m excited about what technology, how technology is changing. Recruiters work today because as I mentioned, to be honest, I don’t think anyone got into this business to do manual repetitive work. You know, these are tasks that nobody likes to do, but somehow we all have to have to do it. But you know, obviously with what, what do we do when sourcing, screening and sorting candidates are done by robots? I actually had this conversation recently with our own head of people at Smart Recruiters, Sarah Wilson, and she said something that really stuck with me, and I’m going to quote her here. She said, inserting AI into recruiting process is a huge win for talent acquisition professionals because instead of spending their time on repetitive administrative tasks, recruiters can finally focus on the part of the job that matters the most. Connecting to the right candidates with the right opportunities. I personally think she’s hitting the nail on the head because there is a human element to recruiting that can never be done by a machine. And if you’re looking at other industries that has gone through the same transformation, let’s take medicine as an example. With the computer vision and data analysis on medical imagery, now doctors can really focus on the important work of patient care, because introducing that technology into the process doesn’t replace the doctors. It’s actually opening up for them to spend more time on care and the next steps instead of actually trying to Figure out what the potential causes are. We have seen a similar thing happening within the marketing industry because if you’re looking at five to 10 years ago and the tasks that a marketing professional would do, obviously most of those tasks have been automated through automated, automated advertising, automated targeting, real time bidding, and a huge amount of those tasks that were done manually, it just are currently done by machines. And instead those professionals are not out of work, they really have changed the skills that are required by them and really focusing on a much more creative work and much more analytical work instead of actually the process oriented workflow that they had before. So in terms of recruiting, I see that disturb automation will replace many of the transactional aspects of our jobs today. Things like screening, matching, interviewing, also scheduling and even onboarding. So human skills are becoming increasingly important. Things like conversational skills, networking skills, but also community building, coaching, being able to influence candidates, negotiating, closing candidates, and also the whole relationship development with your community is going to be interesting. So there’s going to be a big challenge there because this is again not about a future vision. Automation is happening today. There’s several companies out there, several software vendors that easily have automated many of the manual tasks that are happening in recruiting today. And so we just need to embrace that change and not necessarily trying to fear it and instead looking at how we need to change our way of thinking. Because there’s no point in trying to compete with robots and technology. Instead we just need to focus on what we humans are good at.
Matt Alder [00:21:23]:
So as we’ve both said, everything that we’ve sort of talked about in the conversation so far is stuff that’s happening right now and stuff that forward thinking companies are already adopting. As a kind of a final question, sort of some final thoughts. Where are we going next? If this is the present, what does the future look like? What sort of technologies and things do you have on your radar?
Hessam Lavi [00:21:48]:
I mean generally I think with the hype around AI and machine learning that is in the market today, I see similar thing happens past years around mobile technology. You heard startups and companies talking about their mobile roadmap and being mobile first. Similar to that, I think within a couple of years AI and machine learning will just be a standard part of any software, including recruiting software. So all software will have to become smart because if they’re not, they then just absolutely going to be killed by the competition out there. So that’s one thing that all software will become smart. Second is just the way that we recruit and the way that both sides of the market is changing is happening much more rapidly. So not only as recruiters, we have to get used to automation and change the way we work. We also need to take into account that the candidates are the way they look for jobs, the way they manage their careers also going to change massively. We talked about intelligent advertising and how most companies are still posting their jobs to job board and hope for the best, and I think that’s definitely not something that’s going to work in the future. It needs to be much more focused on a candidate relationship management because I think candidates today behave much more like consumers than the typical job candidate that we might be thinking of. They value the convenience and personalization much more. They see companies much more like brands that they identify with and want to work for. There’s also another area where I think candidate relationship management is going to be a major area for future changes.
Matt Alder [00:23:48]:
Hessam, thank you very much for talking to me.
Hessam Lavi [00:23:50]:
It was a pleasure. Thanks Matt.
Matt Alder [00:23:52]:
My thanks to Hessam Lavi. You can subscribe to this podcast in itunes or via your podcasting app of choice. The show also has its own dedicated app, which you can find by searching for Recruiting Future in your app store. If you’re a Spotify user, you can also find the show there. You can find all the past episodes@www.on that site. You can also subscribe to the mailing list and find out more about working with me. Thanks very much for listening. I’ll be back next week and I hope you’ll join me.
