All right, we're going to go ahead and get started. I'm Doug Anmuth, JP Morgan's Internet Analyst. I'd like to welcome President and CEO of Recruit Holdings, Mr. Hisayuki Idekoba, or Deko, as he's widely known. Deko is a lot easier for me to.
Yeah, it's easier for me.
Okay. All right. Recruit is a global leader in the HR industry, best known for its leading jobs and hiring platforms, Indeed and Glassdoor. Listed on the Tokyo Stock Exchange, Recruit serves users and clients in more than 60 countries and generated more than $23 billion in revenue in fiscal 2024. Deko has played a central role in shaping Recruit's global strategy. After building leading Internet businesses in Japan, he led the acquisitions of both Indeed and Glassdoor, helping establish Recruit as a leader in HR technology. Today, the company supports more than 3.3 million employers annually and reaches more than 600 million job seekers worldwide. The platform helps more than one person get hired every three seconds, which certainly shows the global impact.
Since 2021, under Deko's leadership, Recruit has evolved into a global technology powerhouse, leveraging AI and data innovation to redefine how people get jobs and how employers find talent. Welcome, Deko.
Thank you for having me.
All right. Maybe just to start, for those in the audience who might be a little bit less familiar, could you briefly tell us what Recruit Holdings is today in your own words and how you describe the company's core mission?
Yeah. Recruit Holdings, we are basically now focusing on simplifying hiring. That is our mission. When we think about the HR industry, which is a $300 billion-plus industry, but it's including like 60% or 65% of human labor manual cost, it's very difficult to find that big industry with such a high percentage of human labor manual cost. What we are believing, we believe, is basically, you know, how can we simplify hiring with using AI and technology and data to reduce manual work? That's what we are focusing on.
Okay. The acquisition of Indeed was clearly transformative for Recruit. What was it about the company back then that gave you the conviction that it could become the global leader that it is today, especially when a lot of others in the industry expected kind of other large Internet players and search giants to kind of dominate the space?
Yeah. First of all, I was very lucky to meet with great founders. You know, at that time, we started to explore companies, how we can expand outside Japan. I started to meet with a bunch of HR companies, you know, meet with CEOs and founders. You know, probably I met hundreds of companies. Most of CEOs told me how they are monetizing employers or what kind of products are they selling to employers, employers, employers. When I met with founders of Indeed, they're talking about obsession of, you know, job seeker experience. Their mission was, is, still is, we help people get jobs. When we built, when I built Internet businesses when I was in Japan, I really believed that Internet technology should be able to, you know, like democratize or empower consumer experience.
I really liked the founders, and I tried to convince them to join Recruit. At that time, Indeed was roughly $80 million revenue, and we had 40 engineers and product guys. You know, since then, you know, 15 years, it took 15 years, but we grew 100x bigger.
Okay. All right, let's talk a little bit about the labor market outlook. Recruit and Indeed's data is often seen as a leading indicator for the global labor market. What trends are you seeing from employers right now? How are those insights shaping your view on the 2025 outlook, especially just given the forecast for continued softness in U.S. job openings?
Yeah. First of all, I need to touch on the longer-term big trend for all that's happening in all the developed countries, not only in the U.S., which is less supply of the labor force because of the aging population. In the U.S., 2018, 2019, we saw almost zero growth of the working-age population. From 2020, we added like roughly 4.5 million-5 million-ish workers on top of 160 million workers. 90% of them were immigrants. In 2008, 2009, the U.S. was adding 3.5 million-ish working-age population every year. That situation of the labor market itself is so different compared with 2008, 2009. We think we will see continuous weaker hiring demand for the next 12 months. Roughly, we are thinking that 10% less hiring demand. As I said, the supply of labor force is extremely weaker than, you know, 2008, 2009.
As a result, we're not thinking we're going to have like 8%-9%-ish high unemployment rate. It's not going to be happening. It might be like 5%-5.5%. People might say it's a mild recession. What we are thinking is like if the unemployment rate is 5%-5.5%, which is the last 20-30 years' average, if people can keep their jobs, probably it's not going to be so bad as a recession. Again, like we're still thinking that weaker demand for the hiring.
Okay. So what's the impact of all of that on Recruit? Your HR tech revenue grew last year despite declining U.S. job postings. You're now forecasting U.S. revenue to be slightly up this year, even with another 10% decline in job openings. How are you able to, you know, kind of recoup this revenue growth from job posting volumes, and how sustainable is it?
Yeah, I can say what we have been trying is the shift from quantity to quality, meaning in the past, probably, you know, 15 years, we have been selling quantity, meaning, hey, employer, if you want to get more applications, you need to pay money. That is a typical advertisement, job advertisement business. Now with AI improvement, we can say quality, if you want to buy quality, we have AI screening tool or AI matching tool or AI even interview tools, whatever, to pick a good candidate. For example, we're literally sending, you know, the big companies like, you know, the people are having, you know, conversations here, we're sending 2 million-3 million applications every year. I'm sure they have thousands of recruiters, but still they need to pick good candidates out of literally 2 million-3 million candidates.
Right now, what we are trying to do is, hey, if you want to pick good candidates out of a million candidates, this is a tool. What we're trying to do is how can we sell more qualified applications. Of course, you know, there are so many small businesses which really need additional applications. At the same time, you know, even these small businesses have some open positions, which is requiring very selective skills or experiences. There should be a good fit with these AI tools.
Okay. You've also talked about that you expect to achieve record profits again in fiscal 2025, and that's despite potential market challenges. Can you just elaborate on your approach to efficiency and productivity this year, and particularly how AI is being leveraged internally to drive more of that improvement?
Yeah, AI is clearly a game changer, and especially for, you know, like us, like, you know, two-sided marketplace model, meaning, you know, when we think about our sales guys or client support, they're talking with employers to advise, hey, you know, if you're posting this job, this budget, you know, this should be like this size of the budget, you know, or like, you know, maybe you should change this job title to be like this or blah, blah. That should be able to, you know, be changed with AI agentic experience. We're going to have more agentic experience for employers and job seekers and to improve our internal productivity. You know, I have been saying, you know, almost every team like, hey, you guys need to rethink and redesign all operations from scratch with AI. How can you redesign it?
You know, for example, like 1/3 of our programming codes were written by AI now, and it's going to be 50% pretty soon. Meaning we need to have a good vision, okay, two years later, even the next year, how we are going to work with AI. That should be the center of our new design. I think AI change is big.
Okay. Let's stick with AI. You showcased, I think last week, some of your new product experiences that you're building, and they seem to go beyond kind of the traditional job search. Maybe you could talk about those a little bit and just how you see AI evolving the services that you offer to both job seekers and employers.
Yeah. For example, in the U.S., a lot of job seekers are applying 20x , 30x , 40x , literally 20 in a month, 20x , 30x . It should not be the experience for job seekers with AI agent. What we are trying to have is like, you know, when we think about actual experienced recruiters, if you're getting this support from recruiters, you will never like apply 30x , 40x . What we're trying to do is how can we have like carrier advisor-ish experience with the AI agent. For example, like, hey, I know you're applying this job, but, you know, maybe you should think about your current skill, and maybe, you know, you should try to apply this job instead of that job to get this additional skill.
You know, maybe you can apply this job later or whatever. These types of conversational experience will be very important for job seeker side and also employer side. Right now, you know, small businesses don't know how to write down job description, what kind of skills or experiences should be required, or hourly wage, how can they set and benefit. These kind of fundamental things should be, you know, answered by AI agent. We're having a lot of AI agentic experiences for job seekers and employers that will upgrade our experience for them.
Okay. You're in a pretty unique position just as a very large two-sided talent marketplace, vast amounts of proprietary data about both job seekers and employers. How does that data and scale enable you to build AI solutions that others cannot easily replicate?
Yeah. First of all, of course, the size of the two-sided marketplace itself is a good differentiator. You know, 600 million profiles and 3.3 million employers are logging in to post their job and check billions of billions of applications. That itself is having a huge power to train AI model. On top of that, when we think about next two years, three years, AI agentic experience, the important thing is how can we personalize experiences for job seekers and employers. For example, like if somebody said, oh, we built super smart AI, you should ask all questions to this AI. You know, but still, even if it's so smart, this AI has to ask a bunch of questions. Hey, Dad, what did you do before? What kind of experience do you have? Like, what do you want to do for the next three years, whatever?
Like, they need to know you. To personalize the experience, the piled-up historical data for each job seeker is very important. Also for employers, you know, even if the employers are posting the exact same job title, we saw totally different preferences from each employer. Each employer is very unique in what kind of people they're hiring. It is also very important to have piled-up historical data to personalize their experience. I think when we're having more AI agentic experiences for them, it's a great differentiator.
Okay. So your business in Japan, you have some unique insights from your placement businesses. How does combining your online job platforms with the traditional placement services help you strengthen your ability just to make job search and hiring simpler and to deliver a better user experience?
Yeah. You know, we have the largest recruiting agency, offline recruiting agency business in Japan. We merged this offline business with online job board business a while back. What we found is the, you know, the mix or collaboration between offline and online is very important to evolve AI, you know, power. What I'm saying is, like, you know, when we checked very experienced recruiters, how they're navigating job seekers, how they're supporting employers, this is great. Meaning like today's AI cannot fight against these very experienced top-level recruiters. What I'm saying is we need to try to catch up these great recruiters. To do that, we need data, how they're doing their jobs. So these, you know, we need to build each parts, but still AI agent is not that great.
You know, if we can have a bunch of data from offline data, I think that would be the best way to upgrade AI agentic experience. I think it's going to be a key how we can actually change offline experiences with the data.
How do you tie that together, you know, from an AI agentic perspective? It feels like, you know, the big question on everyone's mind about AI products is how do you plan to monetize these agentic experiences?
Yeah. Everybody is asking monetization about AI. I know we need to do that. But, you know, basically, as I said, like $300 billion+ HR industry, the big part of this industry is a recruiting agency model, meaning they are charging 20%-25% of first year salary. I'm sure people are, you know, have experience to get the support from these recruiters. But when we asked job seekers, just 4% job seekers were telling us they got support from recruiters, meaning 96% of people, first of all, just 30%, roughly 30%-ish people in the U.S. have college degree. So 70% people, you know, don't have college degree. And there's like 4,000 universities in the U.S. And I don't know how many universities' names do you guys know. It's a tiny, tiny percentage who can get the actual support from recruiters.
You know, what I'm thinking is like we should be able to replicate recruiters' experience with AI. Then, you know, it should be cheaper. It should be easier for all job seekers to get this benefit with the AI. There should be, we're just still charging less than 1% of first year salary through our platform. If we're delivering really qualified candidates, we should be able to monetize 1.3%, 1.5%, 2% of first year salary. What I'm thinking is right now we should be focusing on to build actual value, which is delivering qualified candidates. If employers are saying, oh, you guys are great, then there should be a way to monetize our product. The focus right now is, of course, we're going to do a bunch of tests how to monetize it, but we need to build value first.
Okay. Let's talk about just the AI impact on the labor force. Just as we think about that, you know, that impact of AI, just how it reshapes things on a global basis. What is your long-term perspective on AI's impact on job creation, jobs that might get displaced over time, and the skills that are going to be most in demand in the future?
Yeah. Actually, what we're seeing with our data, which is like software engineer hiring demand decreased 35% compared with 2019 pre-COVID level. The software engineer is the lowest hiring demand compared to 2019. It's probably clear that software engineering job is even today, like it's impacted by AI improvement. But, you know, on the other hand, there are so many, you know, researches showed that, you know, for example, like not only software engineering job, like accounting jobs or legal jobs, these jobs were highly ranked. It's going to be replaced by AI. But when we checked these data, legal jobs, accounting jobs, 20% higher hiring demand compared with pre-COVID level. It's not happening yet. Also, you know, even 10 years ago, like driver jobs will be replaced by automating driving car compared with, you know, pre-COVID level.
We're seeing 40% higher hiring demand for driving jobs. Right now, what we're thinking is if the job is 100% online, it will be relatively easier for AI to replace. If the job has 5%-10% requirement to have human-to-human interaction, it's not easy for AI to replace. At least right now, we're not seeing it. Even, as I said, like literally like higher demand for these positions we'll see. What we're thinking is it will take time to see this change and actual impact from AI. Also, yeah, probably I should stop here.
Okay. I think we have a mic in the back. If anybody has a question, feel free to raise your hand and we'll get a mic around. Let me keep going with a couple more, though. You have a large staffing business. We hit on that a little bit. Why not bring staffing more into HR tech? And what's the strategy for that segment?
Yeah. Maybe the same thing as I said right now, but the staffing business has a lot of manual work. Of course, we're trying to have more AI to automate each process of recruiting agencies and staffing agencies. Still, staffing agencies have a lot of manual parts, which is very difficult to replace with AI right now. Of course, we've been doing a bunch of tests. We're going to invest more. Right now, it's probably healthier for us to, you know, keep these business units as a separated unit. You know, we are having more and more module type of approach, meaning like, hey, you know, staffing agencies, can you use this AI interview as a part or AI, you know, screening as a, you know, solution for this part? I think in the long term, probably we will see more results.
You know, we will think about the actual collaboration or reorganization. That's how I'm thinking.
Okay. We have a question over here if we can get a mic over. Thank you.
Hello. Thank you for this. Question on what you were saying regarding the candidate and assisting them with AI, with the agentic AI. Who's going to pay for that?
Who's going to pay for that? Better candidate.
What does that make a better candidate fit? Is that the idea that you can match better to what the recruiter actually wants or whoever's recruiting?
Yeah. Good question. As I said, we're sending like 2 million, 3 million applications. Every time I talked with the CEO, like, hey, actually we're sending this much application. They would say like, what? 2 million? What is that? Almost all the time, hey, how many recruiters do you have? Most of the CEOs don't know. Most of the time, like 2%, 3% of their whole cost were associated with talent acquisition or HR. When we think about the future, what we're trying to do is how can we simplify recruiters' jobs inside recruiters, outside recruiters? It doesn't matter. Especially in the recessionary time, all CEOs will think about how can we downsize the talent acquisition team. It happened in 2008, 2009.
Actually, 2008, 2009, after that, 2010, 2011, 2012, what we saw is huge uptick of RPO business, which is recruiting process outsourcing business, because they cut a lot of talent acquisition team. We'll see. Fundamentally, the biggest cost for employers are internal operational cost, and not only the advertisement cost. Advertisement cost is actually a tiny part. Fundamentally, when we think about the future, like the next three, five years of AI change, I think we need to talk more about the actual cost, not only the advertisement cost. It might take time, but that's how we need to approach with employers. Is that answering your question?
Yes, it does. Thank you.
Thanks.
Yeah. I had a question, which is really following up on his, where just a question on the alignment of interests, right? Today, if you think about on the one hand, having important relationships with recruiters themselves, right? They're kind of representatives of the supply side. On the other hand, let's just say we have Recruit's AI, which is basically disintermediating the relationship between the supply and the demand. If I take that to its kind of logical conclusion, the agent should have the interests of the applicant in mind first, right? Not first the recruiter or the company. The applicant may even be willing to pay for this. I would pay for it, obviously, if I got a perfect job that I was interested in. These interests may not align between the recruiter and the applicant and Recruit itself, the company.
It could even become adverse, right? If, for example, I'm applying to a job and the recruiter finds out that Recruit said, hey, this job may not be for you, that creates some tension. How could you, how are you thinking about that over time?
Yeah. Great question. In a perfect world, we want to try to build a perfect world, meaning like everything should be dictated by supply and demand, right? For example, we're talking with the employers, which is having a sugarcane factory in the middle of nowhere. It's a big factory. They are having like a bunch of, you know, machines, but these are old machines. They have to hire, you know, who have experience of this old machine. It was very difficult to find this position. They have to pay a lot more to find out this position. At the same time, the supply itself is so limited. Meaning like maybe, you know, what we need to think about is like, hey, you need to pay more to even like asking job seekers to relocate to this middle of nowhere.
That type of solution should be required because there's supply and demand. At the same time, job seekers want to get this job to step up ladders. But, you know, employers prefer to have somebody who already has the experience. There's a supply and demand issue all the time. We have been testing some small tests. How can we charge job seeker side? At the same time, again, like as I said, like how can we charge employers? The important thing is how can we actually make satisfied hires? Again, like if we can have better supply and demand estimation model, we should be able to get the price tag for each hire better. Right now, we're still probably scratching the surface. Right now, it's just a one-sided demand-side auction model.
Meaning if we do not have two, three buyers for this auction model, middle of nowhere sugarcane factory, it is very cheap. You know, from the supply demand point of view, we need to charge way more. Same thing for job seekers. Sorry, this job is having a lot of applications. Probably you should be, you know, stunned out a little bit. Maybe better to pay, whatever. Again, it is going to be, yeah, probably, you know, we need to calculate and we need to improve our marketplace model. That is basically how I am thinking for the next 10 years. Is that answering your question? Sorry.
Okay. Great. Just want to ask you more about your strategic vision as we're wrapping up. What's the most important takeaway you want investors to understand about Recruit's long-term vision?
Yeah. You know, as I said, we're seriously trying to solve this HR industry problem, which is having a lot of manual work, $300 billion+ industry. Probably we are a strong leader in this market, but still, it's so fragmented. There are a lot of, you know, like small boutique firms, whatever, $300 billion-plus industry with a lot of manual cost. What we need to be responsible for is how can we have actual technology, AI technology to simplify hiring? Because we believe it's definitely better for job seekers and employers if it's easier, simpler to get a job, to make hires. It's better for the society. We really believe if we can have a better technology, we should be able to get more market share.
Maybe, you know, HR industry size itself might be, you know, smaller, but that's a good thing for the society because the cost of making hire or changing job is a cost for the society. We really believe that we need to deliver innovations to improve the world.
Okay. Great. We're going to wrap up with a quick word association, whatever comes to mind. Short answers. U.S. labor market.
Short supply.
Scale.
Two-sided marketplace.
Agentic.