Transcript of Earnings Results Call and Sell-side Analyst Follow-up for Q4 FY2025

This is an edited version of the English translation of the Q4 FY2025 earnings call which was

conducted in Japanese. Please note there are differences between the simultaneous English audio

translation during the Q&A session and this version. The Q&A session was translated from Japanese

using an AI tool and subsequently edited for clarity.

Recruit Holdings Q4 FY2025 Earnings Call

May 15, 2026

Recruit Holdings Q4 FY2025 Earnings Call

‧‧‧‧‧‧‧‧‧

Q&A

‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧

Sell-side Analyst Follow-up

‧‧‧‧‧‧‧‧‧‧‧‧‧‧‧

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Recruit Holdings Q4 FY2025 Earnings Call

Slide 01

Shen : Welcome to the Recruit Holdings FY2025 earnings call.

This call is a simultaneous translation of the original call in Japanese and translation is provided for the

convenience of investors only.

I’m Mizuho Shen, manager of Investor Relations and Public Relations.

Earlier at 3:30 PM, we disclosed the Earnings Release, Earnings Summary, and the presentation slides of this

results call on our IR page. The video and transcript of this results call and the transcript of follow-up meeting

with sell-side analysts which will be conducted followed by this results call, will be posted promptly on our IR

website after the session.

Today’s presenters are;

Hisayuki Idekoba, Representative Director, President and CEO;

and Junichi Arai, Executive Vice President and Chief Financial Officer

In the first 25 minutes, Deko and Jun will provide a presentation, followed by a Q&A session.

Now I’ll turn the call over to Deko.

Slide 02

Hello, this is Deko from Recruit Holdings.

Slide 03

First, I would like to start from this page and talk about revenue and profit.

What I remember is that last year, I spoke in a similar way, and at this time last year, I was talking about how

we were doing various things with AI and everyone’s reaction was like, “wait and see attitude.”

And the reason for that was the numbers.

This time, we have been doing quite a lot with AI, and we are now in a situation where actual results are

appearing in the numbers. For the fiscal year ending March 2026, we achieved record-high profits, and as you

can see here, for FY2026 as well, both revenue and profit growth are accelerating even further.

Of course, AI utilization is advancing across all of our businesses, but today, since we receive the most

questions about how we are utilizing AI in the HR Technology business and how we are generating revenue

from it, I would like to spend some time explaining that.

Slide 04

The other day, I bought a tablet device. Because of my job, I often travel long distances, so I wanted to

download various movies. I thought a large storage capacity would be good and while a large screen size

becomes difficult to carry around, if it is too small, it is hard to watch.

As I kept comparing different options, there was also a huge difference in prices between very expensive and

very cheap models. Before I knew it, an hour had already passed. I realized that choosing consumer

electronics is actually surprisingly difficult.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Slide 05

This may sound a little strange, but what if all of these were free?

You may be wondering what I’m talking about. But if all of them were free, then you would just choose the best

one. If everything were free, everything would become simple. There would be no need to spend an hour

agonizing over it.

Slide 06

Then, if you think about what job sites are like when changing jobs, they are actually no different from a free

e-commerce site.

You may think that’s not true, because changing jobs is something people take seriously and think about

carefully. But regardless of what your current salary is, whether the next position pays two, three, or four times

more, or even if you only work two or three days a week, it doesn’t matter — you can apply to as many jobs as

you want for free.

Japanese people are serious-minded, so they do not apply to that many jobs, but globally, people apply like

crazy. They keep pressing the apply button over and over.

In the past, when a competing search company announced that it would start a job search service, people

said that Recruit and Indeed might be disrupted.

And now, people sometimes say that if AI can tell users “this kind of job may be good for you” through various

prompts and conversations, then people may stop coming to Indeed.

But when you think about what kind of data is being used to build those algorithms, whether it is AI or machine

learning actually does not make much difference. You realize that data such as “this job seems good for this

person,” “this person seems to like this job,” or “this job seems popular” does not actually mean very much.

And the reason is that whether an application leads to an actual hire ultimately depends on whether the

company thinks that person is a good fit.

Slide 07

In HR matching, no matter how much data on the user side you collect, it actually does not mean very much.

What matters is whether the company likes that person or not. People often say that hiring is “two-sided

decision-making.” It is a type of matching where both sides must say “yes.”

So even if you introduce AI on one side, or even if you have data about companies’ various expectations, data

such as “this company tends to hire these kinds of people”, that alone still does not easily lead to successful

matching. That is the nature of this model.

It is a very unique matching model.

Slide 08

Therefore, over the past two to three years, our AI development investments have focused on developing AI

tools for the employer side.

Of course, data on the user side, such as what kinds of jobs job seekers like, is also important. However,

what is extremely important for improving matching is having solid data on the company side: data showing

that “this company hires these kinds of people for this position,” and being able to collect hundreds of millions

of such data points around the world. That is why we have been investing in encouraging companies to use

more AI tools.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

This still may be a little difficult to understand, so let me explain today using a simple example.

Slide 09

Please imagine that you are the owner of a restaurant with 10 floor staff and 2 chefs.

If one floor staff member quits, of course the owner would be troubled. But would it reach the point where the

restaurant could no longer operate? Probably not. Depending on the day of the week, it may simply mean

customers have to wait a little longer.

But if one of the two chefs quits, then in many cases what we hear from employers is: “Cost does not really

matter. We just want to hire someone with the skills and experience to cook delicious food as soon as

possible, even one day earlier.” Otherwise, the restaurant cannot function. There are many cases like this.

We also hear similar stories in the U.S. trucking industry, many companies have more trucks than drivers. In

construction as well, many AI data centers are being built right now, but construction workers are hard to find.

There are many examples like this where hiring is directly tied to revenue, and recruitment becomes extremely

important.

Slide 10

In those situations, what companies use are what we often refer to as Sponsored Jobs, paid job ads that

appear in search results when someone searches for a keyword such as “French chef.”

Many of you may think, “Well yes, of course ads like this appear when someone searches for ‘French chef.’”

But AI matching has dramatically improved the quality of those matches.

Slide 11

Let me explain this a bit further. We have not disclosed these figures before, but here is what they mean.

A user performs a keyword search, and in response, a job advertisement appears in the search results. Right

now, only 30% of applications come from that.

With AI, we do things like suggesting jobs, “How about this job for you?” and, based on data from employer AI

tools such as AI sourcing and AI screening, we send emails or show pop-ups saying things like, “This

company seems to want you to apply for this position.” In fact, this accounts for 70%. This is an extremely

important point, and a key area where AI is evolving.

Slide 12

In addition, since around last year, we have been more fully rolling out and selling a product called Premium

Sponsored Jobs. It is sold as something like an even better version of the Standard Sponsored Jobs that I just

described.

And what it is, simply put, is a product that attracts candidates of even higher quality.

In the past, as I mentioned earlier, when companies urgently needed to hire someone and increased their

budget, what often happened was that for a single position they would receive 200, 300, or even 400

applicants. In other words, the focus was on simply increasing volume.

As a result, restaurants would say things like, “Even if 500 people apply, it just becomes difficult for us to

choose. It makes it hard to increase the budget for that.”

Now, however, if companies increase their budget and pay more, the product has evolved through AI

matching so that high-quality applicants, the kind where employers think, “I want to hire this person

immediately,” are the ones who apply.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

As a result, hiring time has been reduced by 50%

This is becoming increasingly effective. This is what it means for AI to strengthen matching.

Slide 13

Now looking at the job seeker side, what are they struggling with?

The number one complaint we hear is: “I applied to so many jobs, but nobody responded to me. Are these

companies really hiring at all?”

That kind of thing is a common complaint heard all over the world.

Slide 14

In fact, solving this issue is also where the Premium Sponsored Jobs and AI matching, I mentioned earlier, are

having a significant impact.

The reason is that when we quickly match employers who urgently want to hire, such as those looking for a

chef immediately, with candidates they would actually want to hire, customers using Sponsored Jobs respond

45% faster.

So from the job seeker’s perspective, speed is increasing. They experience, “I applied and immediately

received a response.”

That is why understanding which companies have strong hiring intent is actually extremely important. Of

course, the job data that Indeed possesses can also be created by crawling other sites and gathering data

from them.

However, it is much more difficult to determine which companies truly have strong hiring intent.

Improving this matching through AI creates benefits not only for employers, but also for people seeking jobs.

Slide 15

As a result, the number of monthly active users in March increased 18% year over year.

As I mentioned earlier, people were saying, “Since everyone is moving to AI services, won’t people stop

coming to Indeed?” But in reality, usage increased 18% year over year. This is a record high.

Slide 16

This shows just how much AI is improving the user experience.

And of course, because we are increasingly matching users with companies that have strong hiring intent, as

a result, US ARPJ — that is, how much we charge relative to the amount of hiring activity — grew 25% year

over year in Q4 in the U.S.

When I explain this, people often ask, “If you raised prices by 25% year over year, aren’t customers

complaining? Wouldn’t they get angry if prices increased that suddenly?”

But I would like you to look a little more closely at the details.

Slide 17

This is about the structure of the job market and the recruiting industry.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

As I have said many times, online advertising, job advertisements, and hiring tools represent only a very small

portion of the total market. Even if we say we hold 60% or 70% share here, when viewed as part of the overall

market, it is still only a small percentage.

What is much larger in terms of market size are offline matching businesses such as placement agencies and

staffing agencies.

When we speak with companies, many tell us: first they try Indeed for free, then they try paying for it, and

even then, if they still cannot hire successfully, they know it is expensive but eventually they turn to retained

search firms or placement agencies.

Slide 18

And the reason is what I mentioned earlier: companies know they can hire more cheaply through Indeed.

They also know offline agencies charge high fees, but because they are struggling to hire, they end up relying

on those services.

So when we steadily improve our matching through AI and become increasingly able to deliver candidates

that companies want to hire immediately, the way we think about it is that we are effectively taking share from

that side of the market.

Slide 19

This is exactly what I have been talking about since becoming CEO: one of our missions, “Simplify Hiring.”

Our mission has been to make job searching so simple through technological advancement that a person

looking for a job can press a button once and immediately find their next job.

In that sense, I believe the fact that AI can automate so many different manual tasks represents a tremendous

opportunity, and I think we have finally reached a stage where this is appearing in the numbers and in actual

business results.

Slide 20

AI is truly transforming so many things. And as we see the numbers are actually coming through, my

confidence as a manager of the company is growing.

Therefore, I certainly believe we can achieve 10% revenue growth. And in the medium term, as everyone here

knows, we are not currently in an environment of particularly strong hiring demand. In fact, some people are

even saying we may soon enter a recession. Yet even under those conditions, we are producing these results.

So when normal hiring demand returns, I believe revenue growth exceeding 20% is fully achievable, and I

also believe EBITDA+S margin can exceed 50%.

Slide 21

Currently, 31 hires are made every minute on our platform.

In other words, we are helping one person get hired every two seconds.

We hope to continue expanding this further around the world so that we can help as many people searching

for jobs as possible.

Thank you very much.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Page 23

Hello everyone, I’m Junichi Arai. I hope you enjoyed Deko’s presentation.

I have been by Deko’s side since he first became CEO, watching him present that 'big circle' diagram a few

pages back and explaining his vision to 'Simplify Hiring.'

Looking at where we are today, it feels as though the world has finally caught up with his vision.

With the accumulation of our unique data on Indeed and the full utilization of AI, I hope you were able to feel

that we have taken another step closer to realizing 'Simplify Hiring'—the world Deko has envisioned and has

been sharing with you since the beginning. Personally, I am incredibly excited about this progress.

Furthermore, as we have been discussing since FY2024, we designated the past two years as 'Year 0'—a

period for preparation and strengthening our foundations. After two years of saying we are building our

strength, I am very pleased to finally be able to report significant results and tangible progress, which are now

clearly reflected in our financial performance.

As you can see from the table of contents for my presentation, like last year, I will present the Executive

Summary from here on stage.

The remaining items will be explained in detail afterward in the meeting with equity research analysts

scheduled for later today.

Page 25

For FY2025, consolidated revenue, EBITDA+S, and basic EPS each exceeded the revised guidance

announced with our Q3 results in February and reached record highs.

For FY2026, as we expect this momentum to further accelerate, we expect revenue and profit growth across

all three segments based on an assumed exchange rate of 154 yen per US dollar and assuming no major

changes in the economic environment.

Driven particularly by progress in business evolution and enhanced efficiency in HR Technology, our

assumption is for consolidated revenue to increase by 9.0% year over year to 4.03 trillion yen.

EBITDA+S is expected to increase by 19.5% year over year to 949 billion yen, with margin expanding to

23.5%.

Basic EPS is expected to increase by 27.8% year over year to 447.00 yen.

The number of employees decreased from approximately 50,000 at the end of FY2024 to approximately

45,000 at the end of FY2025, and we do not currently plan any large-scale hiring and the number of

employees remains flattish in FY2026.

Page 26

With regard to our capital allocation policy, in FY2025, we returned a total of 713.1 billion yen to shareholders

resulting in a total payout ratio of 143.5%.

Net cash at the end of March 2026 was 765.9 billion yen.

In May 2024, we announced a plan to reduce net cash from 1 trillion 135.4 billion yen at the end of March

2024 to 600.0 billion yen over the two-year period ending March 2026. We had already reached that level in

the first half of FY2025.

However, as second-half results exceeded our assumptions, we ended the year above that level.

ROE was 31.0%, significantly above 22.6% in FY2024.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Page 27

As for our capital allocation policy for the next three fiscal years starting this fiscal year,

we are making no changes to the current order of priorities.

We intend to maintain year-end gross cash and cash equivalents at around 750.0 billion yen.

If we execute a strategic acquisition, we intend to fund most of the required amount with debt while taking our

credit ratings into account.

Based on our policy of “stable and continuous dividends”, our dividend outlook for this fiscal year is 13 yen per

share in the first half, 13 yen in the second half, totaling 26 yen for the full year.

Regarding share repurchases, after the 350.0 billion yen share repurchase program launched on April 1 this

year is completed at the end of November, as currently expected, we will make appropriate decisions and act

accordingly while monitoring second-half cash flow generation, capital market conditions, and our share price

level.

Going forward, we believe ROE can be further improved to exceed 31% recorded in FY2025 through growth

in net income and continued shareholder returns.

Page 28

As you are aware, we operate through three segments. Among them, HR Technology’s revenue growth, profit

growth, and margin expansion are substantially higher than those of the other two segments.

In FY2026, HR Technology revenue is expected to reach a new record high of over 10 billion dollars or 1.5

trillion yen.

Although this represents only about 40% of consolidated revenue, it is important to consider the revenue

structure of Staffing. While Staffing, which accounts for 45%, records approximately 1.8 trillion yen in

accounting revenue, the margin excluding wages paid to temporary staff is around 300 billion yen.

Accordingly, HR Technology accounts for approximately 65% of consolidated gross profit, which we believe

better reflects its substantive contribution to consolidated revenue.

Page 29

We expect HR Technology, which accounts for about 70% of consolidated EBITDA+S, to continue to grow as

the core driver of our earnings and value creation.

While we do not expect significant organic growth in Staffing under its current business model, we will

continue to focus on efficiency and strive to generate stable EBITDA+S margins.

Marketing Matching Technologies, or MMT, will continue refining and evolving its business model to further

strengthen its competitive advantages and uniqueness in Japan. By increasing revenue and improving

operating efficiency, we expect EBITDA+S margin of 30% this fiscal year, and aim to increase to

approximately 35% by FY2028.

Page 30

Now, I will discuss each segment.

First, HR Technology.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Q4 results substantially exceeded the outlook disclosed in February in all regions, including the US, Europe &

Others, which covers Canada, and Japan. As a result,

full-year revenue in FY2025 increased by 7.6% year over year to 9.67 billion dollars.

EBITDA+S margin was 37.7%, reflecting progress in efficiency improvements, including lower personnel

expenses, and significantly exceeded the FY2024 pro forma margin of 33.0%, despite FY2025 being the first

year of integration of HR Solutions of Matching & Solutions.

For FY2026, while US revenue showed strong year-over-year growth of 26% in March and 27% in April, our

full-year outlook reflects the potential risk from the uncertain economic environment, including geopolitical

tensions and commodity price volatility.

We anticipate continued monetization development, including further growth of Premium Sponsored Jobs, in

the US and in Europe, Canada, and other markets.

In Japan, we expect steady performance of Indeed PLUS and a recovery of the placement business to

FY2024 levels.

As a result we expect revenue to increase by 11.0% year over year on a dollar basis to more than 10 billion

dollars, while EBITDA+S margin is expected to increase to 41.0%.

As Deko mentioned earlier, over the medium term, it is well within our reach to not only maintain double digit

annual revenue growth, but to achieve 20% or greater when hiring demand recovers. We expect our

EBITDA+S margin to exceed 50%.

Page 31

In the US, in FY2025, amid stagnant hiring demand, and while the total number of US job postings on Indeed

declined by approximately 7% year over year, monetization development resulted in revenue of 5.31 billion

dollars, an 8.8% year over year increase on a dollar basis, with a US ARPJ growth rate of 17%.

In FY2026, based on our assumption for the job market that hiring demand will remain flat after bottoming out

in Q4 of FY2025,

assuming the total number of US job postings on Indeed decline by approximately 4% year over year, our

outlook is for revenue to increase by 13.6% year over year to 6.03 billion dollars, or 929.3 billion yen, and for

US ARPJ to increase by 18%.

We position this fiscal year as “a year to steadily lay the groundwork and pave the way” for further US revenue

growth from FY2027 onwards.

Keeping in mind the Japanese proverb "haste makes waste," we intend to carefully consider the timing and

pricing of new plans and product launches before executing them, while maintaining close communication with

our business clients.

Page 32

As for Staffing, revenue in FY2025 increased by 2.2% year over year to 1 trillion 703.4 billion yen. By region,

revenue in Japan was 846.8 billion yen, and revenue in Europe, US, and Australia was 856.5 billion yen.

EBITDA+S margin was 5.9%, maintaining the FY2024 level.

In FY2026, while we do not expect major revenue growth in Japan, our outlook for Europe, US, and Australia

is for revenue to increase by 5.8% year over year in yen terms, supported by performance bottoming out in

key markets and a gradual recovery trend.

Our outlook is for EBITDA+S margin of 5.6%, broadly in line with FY2025, and EBITDA+S of 100.5 billion yen.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Page 33

Next, MMT.

In FY2025, Lifestyle, led by Beauty, drove revenue growth, which increased by 4.7% year over year to 564.6

billion yen, while EBITDA+S margin was 27.4%.

In FY2026, our outlook assumes continued growth in Lifestyle, and we expect revenue to increase by 7.1%

year over year to 605.0 billion yen.

By smoothing out quarterly seasonality in sales promotion and advertising expenses, our outlook is for

EBITDA+S margin of 31.0% in the first half, 29.0% in the second half, and 30.0% for the full year.

Recruit Holdings will continue to pursue its growth strategy as a global technology company, consistently,

amid major changes in the business environment, while also working to continue to improve operational

efficiency.

We respectfully ask for the continued understanding and support of all stakeholders, including shareholders

and capital market participants.

That concludes my remarks.

Comparison of Premium Sponsored Jobs to jobs posted for free. Time to Hire refers to the number of

days from the creation of a job posting on Indeed to the reporting of the first hire for that posting. U.S.

data from Jan 2026 - Mar 2026, for jobs hosted on the Indeed platform using median.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Q&A

Munakata: My name is Munakata from Goldman Sachs Securities. Thank you for the opportunity to ask a

question. My question is about the macro environment and the outlook for the labor market.

Looking at the presentation materials you provided, I once again felt that you are showing very strong growth

while [adapting to] the environment. At the same time, factors that affect the labor market outlook — such as

structural labor supply shortages and AI disruption — are being discussed quite actively in the stock market.

So I would very much appreciate it if Idekoba-san could share his current view on the medium- to long-term

outlook for the U.S. labor market.Thank you.

Deko: Just this morning, Indeed Hiring Lab released a new report on the outlook for the U.S. labor market

through 2040.

I also attend Davos and various conferences in the United States and lately there have been a lot of

discussions claiming that AI will drive unemployment rates to 20% or even 30%. Those kinds of opinions —

the sort that are likely to generate clicks and headlines — have become very prominent.

And when I talk about this, people sometimes say that what I’m saying is just talking my book.

However, if you look at today’s actual data, I think you can see that there are considerable differences across

segments in terms of how AI will affect them.

For example, the impact of AI on employment for the restaurant workers I mentioned earlier, or for those in

construction—such as electricians or plumbers—will be vastly different from the impact on those working in IT

or finance.

We already have data showing what percentage of people are currently working in each sector today.

So, conversely, when someone says unemployment will reach 20% or 30%, you should look at which sectors

would have to shrink, and by how much, for unemployment to actually rise to 20%.

If you do that, you can clearly see that many of these claims are being made without any supporting data

whatsoever — regardless of whether the people making them are so-called AI experts or not.

On the other hand, if you look only at the IT industry, then in a scenario where AI-driven replacement

advances aggressively, it is possible that employment in the IT sector itself could decline somewhat — for

example, by around 20%.

So there may indeed be situations where conditions become difficult in certain sectors from an unemployment

perspective.

However, as you know, people working in the IT industry account for only a relatively small percentage of the

total workforce.

When you consider all of this, the bigger issue — particularly in the United States — is actually the aging of

the workforce, which is progressing very much like it has in Japan.

That said, among developed countries, the United States is still actually relatively young.

Even so, over the next five years — through around 2030 or 2031 — the number of workers in the United

States is projected to decline.

This is because the baby boomer generation is retiring. And furthermore, labor-force participation among

people over the age of 65 is much lower in the United States than in Japan — only around 18%.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

So the impact of these people leaving the workforce — especially those working in areas such as healthcare,

or the electricians I mentioned earlier — is, when viewed overall, larger than the impact from changes brought

about by AI.

I believe that if you look at the actual data and figures, you will understand this point as well.

And by the way, I think this presents a slight tailwind for us.

Munakata: Thank you very much. Since we are allowed to ask two questions, I would like to ask one more.

I would especially like to ask about “Simplify Hiring.”

In the process of making hiring simpler, I also think technological advances such as AI are, in some ways,

creating additional complexity. For example, receiving too many applications may be one such example.

So from your perspective, Mr. Idekoba, where do you think the current progress of “Simplify Hiring” stands

today? And in order to make hiring even simpler, what is still lacking, what can still be improved, and what

kinds of business opportunities exist there?

Deko: That’s an excellent question.

For example, in the United States right now, there are quite a few startups offering services where, if you pay

a monthly fee, AI will submit 1,000 or even 2,000 job applications on your behalf.

There are actually many startups doing this. If you search in English in the U.S., you’ll find search results for

services that automatically apply to jobs for you.

And actually, companies really dislike this.

I think you can understand why — from the company’s perspective, if they suddenly receive 1,000 or 2,000

applications, their costs increase dramatically.

As a result, many companies are now spending heavily on things like reCAPTCHA systems and “I am not a

robot” verification measures to defend against AI bots.

In addition, what these companies are complaining about is that even when they think, “This candidate looks

good,” the person often never shows up for the interview.

So I think this is actually a very good example of how AI-driven automatic applications are not improving

efficiency at all.

We ourselves are also testing things like automated applications in various ways, but we are being extremely

careful to consider both sides of the marketplace.

We are absolutely not doing anything like taking money and submitting applications indiscriminately. Instead,

we are testing matching systems where we send opportunities only when we think, “This person is probably a

great fit for the company,” and at the same time, “This opportunity is probably good for the job seeker as well.”

So unless AI is used in a way that carefully considers both sides and leverages data from both sides, I think it

actually ends up increasing complexity rather than reducing it.

Munakata: I understand. Thank you very much.

Oum: This is Oum from Nomura Securities. Thank you for calling on me. I have two questions.

You mentioned that top-line growth will accelerate once hiring demand normalizes. So, to put it directly, when

do you think that will happen?

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

With volume down 4% this year, my first question is: what kind of environment do you expect for next year and

the year after that?

Deko : We look at a great deal of labor market data, although I think all of you are probably more

knowledgeable about the economy itself than I am.

As I mentioned earlier, this is fundamentally determined by supply and demand, and as you know, hiring

demand is still not particularly strong. I think the situation remains fairly challenging.

That said, looking at historical economic patterns, hiring demand has already fallen enough that it would not

be surprising to start seeing a rebound before long.

One interesting piece of data I have been watching recently is related to software engineer hiring. In the

United States, over 5 years software engineer hiring had fallen, by around 50% compared with pre-COVID

levels.

That data was often cited by people claiming that AI would eliminate everyone’s jobs.

But over the past six months or so, hiring has actually started increasing again.

And when you look at the details, you can see job titles emerging where AI is creating entirely new kinds of

work — positions that did not previously exist.

So in that sense, rather than AI itself being the primary issue, I think what we are seeing is more that the

post-COVID volatility in the U.S. labor market is beginning to settle down — people were laid off during

COVID, then companies hired aggressively afterward, then said they had overhired and started laying people

off again. That kind of American-style labor market dynamism is calming down somewhat.

Looking purely at labor market conditions, I feel things could improve over the next one to two years.

As for the broader economy, though, I’ll leave that to all of you to think about.

Oum : Thank you very much.

My second question is about US ARPJ.

You mentioned growth of 18%, and I would like to understand how sustainable that is going forward into next

fiscal year and beyond.

As you explained earlier, if hiring directly drives revenue, then even if costs increase by 20%, companies

would still think, “It’s better to hire someone one day sooner.”

I would imagine that story continues to some extent next year and beyond, especially considering the

population trends and aging workforce you mentioned earlier.

But should we think about this as something that can continue at roughly the “18% level,” or is this fiscal year

somewhat boosted by the introduction of Premium Sponsored Jobs?

Could you help us understand the general magnitude and sustainability of this growth level?

Deko:

I think that’s an excellent question.

Personally, that is also the area I most want to improve further — ideally, I would like to grow it by 30% or even

40%.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

At the same time, however, if we move too aggressively, it directly impacts customers’ hiring budgets as well.

Especially for customers already spending budgets of 20 million or 30 million dollars, asking them to increase

their budgets by another 30% is something we have to approach carefully and thoughtfully.

That said, as I mentioned earlier, in reality there are many customers who are reallocating spending from

other hiring-related costs toward us.

As I mentioned earlier, when a customer is already spending 20 million or 30 million dollars, and we ask them

to increase that by 20% or 30%, that means asking them to spend an additional 600 million, 700 million, or

even 800 million yen. In many cases, that kind of additional budget does not simply appear automatically.

However, based on our conversations with customers, particularly in the United States, many companies’

Talent Acquisition teams receive huge volumes of applications, review countless resumes, decide which

candidates to contact, and coordinate interviews — and a very large number of companies outsource these

processes to places like India or the Philippines.

So in many cases, the budget is actually being shifted from those kinds of expenditures toward us.

I mentioned TAM earlier, but structurally speaking, if customers are ultimately able to achieve better hiring

outcomes and lower their overall hiring costs from their perspective, then I think there is actually still

considerable room for growth.

As I was just saying, we have to maintain the right balance regarding our revenue growth. What matters is that

customers are successfully hiring and that their satisfaction remains high.

In that sense, one encouraging point right now is that customers using Premium Sponsored Jobs actually

have retention rates that are about 20%

2 higher than those using Standard Sponsored Jobs.

I think that is quite a confidence inspiring data point. In other words, because the product is more effective,

customers are feeling that the higher cost is justified and worthwhile.

So while we do need to carefully balance things, we would certainly like to continue moving in that direction.

Does that answer your question?

Oum: Thank you very much.

Maruyama : This is Maruyama from Nikkei. Thank you.

In the United States, it is said that layoffs driven by AI, so-called “AI layoffs” are spreading.

However, as mentioned in the previous earnings results, Indeed’s job postings in the U.S. still have a high

proportion of in-person roles.

So should we understand that AI-driven replacement is less likely to impact the types of jobs on Indeed, and

therefore that this becomes a strength for Indeed?

Deko : I think this is another case where we need to look carefully at whether something is merely an

impression or whether it is actually supported by data.

For example, even among the so-called “Magnificent Seven” companies, the ones frequently appearing in the

news for conducting what are described as AI-driven layoffs and letting go tens of thousands of employees, if

you actually look at the underlying numbers, their total employee counts are often not decreasing.

As we often say ourselves, there are people who argue that AI is not a “job killer,” but rather a "job

reorganizer."

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

And what that really means is, and I think this is probably true at your companies as well, there are actually

relatively few cases where an entire person’s job is completely replaced by AI and that person is then laid off.

Rather, what happens more often is that certain processes within that person’s work become automated.

So especially in Tech companies, including ours, I think many things are happening where we need to

fundamentally rethink how work is done, with AI becoming part of the workforce itself.

For example, think about dashboards used to view data — charts showing weekly sales figures and so on.

Those dashboards are necessary because humans are looking at them.

But if you simply hand that work over to AI, then even the act of displaying the dashboard becomes

unnecessary. There is no longer any need for a person to first bring up the dashboard and look at it.

So from my perspective as well, I often say to my team that we should go back to the beginning and rethink

the entire process — specifically, how value is being delivered to customers and how revenue is being

generated.

Rather than saying, “Let’s just automate this one part with AI,” or “Let’s think about restructuring only this

area,” what we are really doing is redesigning the whole workflow from the ground up.

In that sense, when you look at the fact that even the Magnificent Seven companies are not actually reducing

their overall employee counts, I think what is happening is that certain functions are being reorganized, while

at the same time new types of people are being added, and the organization as a whole is being redesigned.

At least, that is how I see it.

Am I saying anything strange?

Maruyama : No, not at all. On the contrary, while I think there will be increases and decreases in certain types

of work in the AI era, I personally believe Indeed’s current job mix is actually a strength.

With that in mind, do you have an intention to further increase job postings that are less susceptible to

replacement by AI?

Deko : Rather than saying that we ourselves want to increase those kinds of jobs, as I mentioned earlier,

Japan is actually a very good example.

People used to often say, “Japan’s population is declining, so there’s no way recruitment companies can

continue growing.” And yet, in reality, revenue has continued to grow steadily.

Ultimately, this comes down to supply and demand. When labor supply is tight, unemployment rates do not

rise easily.

And to go a bit further, as I mentioned earlier, take auto repair technicians as an example — people who have

spent years repairing traditional combustion engines, gasoline engines, are gradually retiring in larger and

larger numbers.

And there is a growing concern that there simply will not be enough people left who can repair those vehicles.

That is the kind of thing happening in the U.S. as well.

As a result, because supply and demand are no longer aligned, more and more job postings are appearing on

Indeed, and budgets for those positions are increasing.

So I would not say that we are aggressively targeting those areas through sales efforts or anything like that.

Rather, I think this is happening naturally as a result of structural labor market conditions.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Nagao :This is Nagao from BofA. Thank you.I have two questions.

First, regarding margins. You mentioned that margins could potentially rise from the 40% range to 50%.

Could you explain again where that confidence comes from? For example, do you believe pricing can still

increase further? Do you think customer retention can continue improving? Or do you see opportunities to win

more large enterprise customers?

I would appreciate it if you could elaborate on the background behind that view.

Deko : Thank you.

As I mentioned earlier, we only really began selling AI tools over the past year or so. And I think part of the

reason is precisely because they are AI tools, but especially with large enterprises, there are lengthy

processes involved — legal departments need to review everything, companies need to establish Master

Service Agreements, and in the U.S. there are many processes like that which take considerable time.

On the other hand, as I mentioned in the earlier restaurant example, if you think about a small restaurant with

maybe a dozen employees, it is quite common to hear things like, “the person in accounting is also handling

hiring.”

In many cases, the business owner themselves is handling hiring responsibilities.

From the perspective of those customers, even if the price increases by 20,000 or 30,000 yen, they are often

willing to use the product because it frees up their own time to focus on other things. Adoption tends to

happen quickly, and after using it, they often say, “This is actually really good,” which leads to strong repeat

usage.

This has been true for us for a long time, but when we release products to small and medium-sized

businesses and see strong retention rates, we take that as a sign that the product itself is genuinely good.

After all, those customers tend to be very budget-conscious and quite disciplined in how they spend money.

Large enterprises, meanwhile, have much longer sales processes. And because these are AI tools, things

take even more time, since regulations differ from state to state in the U.S.

However, we are increasingly feeling that customers are beginning to accept these products.

And as you know, small businesses are more likely to disappear when economic conditions weaken. But with

large enterprises, we are often able to secure formal annual contracts, which gives the revenue base much

greater stability and depth.

As you know, even in terms of sales efficiency — meaning how much revenue a single salesperson can

generate — the increasing share of revenue coming from large enterprise customers is improving that sales

efficiency over time.

And of course, when we look internally across the company, there are still many areas where AI can be

applied. I am constantly thinking, from morning until night, things like: “If we introduce an agent here and

redesign this process this way, wouldn’t it work better?”

So I think part of the reason my confidence is growing is because we continue to see more and more

opportunities for AI utilization internally as well.

Nagao : And from a product perspective as well, with ARPJ growing around 20% or 25%, the idea is that as

long as customers are receiving ROI that exceeds those increases, they will continue to retain the product —

and your confidence is also building around that point, correct?

Deko :Yes, exactly. Although maybe I shouldn’t say too much about this, but for example, take a sales

manager reviewing how a salesperson delivered their sales pitch — listening in on calls, accompanying them

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

on visits, and then giving advice like, “You should explain it this way instead.”We already have systems like

that in place.

Now everything is being recorded, and AI scores the conversations and suggests things like, “You should talk

more like this.”

And when we pull together large amounts of data from salespeople who are actually achieving strong results,

we can increasingly identify patterns — such as what kinds of things they say, or what types of data they show

during conversations.

And in our case, if we can eventually apply those kinds of capabilities directly to self-serve customers as well,

then when I think about all of these possibilities, it really expands the vision of what could be achieved.

That may be why I personally feel that margins could probably reach around 50%. Though maybe I’m wrong.

Nagao : Just one quick follow-up. Looking back at your company’s M&A history, it seems that some of your

acquisitions, including Indeed, were made during weaker economic environments, such as after the financial

crisis, and ultimately turned out to be very successful.

And now, with software valuations declining, it feels like this could potentially be a favorable timing for capital

deployment.

So how are things currently looking with respect to M&A activity?

Deko : Sometimes I feel that investors say rather contradictory things.

There were people who, just a few years ago, were constantly telling us, “Buy, buy, buy.” But now those same

people say things like, “Idekoba-san, it was brilliant that you didn’t make acquisitions over the past two or

three years.”

I think that’s largely because SaaS valuations have collapsed, so now people say we showed great discipline.

On the other hand, though, there are now more than 1,000 unicorn companies out there, many of which are

facing questions about what they are going to do next. I think that is part of why issues like private credit

problems are emerging as well.

We are speaking with many different companies, and because AI-driven change is happening so rapidly, there

are many premium quality companies that have been built over the last seven, eight, or ten years.

We are actively looking at opportunities, and if we find companies that truly have strong businesses, share a

similar vision with us, and where we can align properly on price expectations, then we would absolutely like to

pursue those opportunities.

In fact, I would say we are even more proactive about M&A now than we were before.

Yoneshima : This is Yoneshima from Citigroup Securities. Thank you.

I essentially have one broad question, regarding pricing, specifically ARPJ trends and take rates.

With respect to ARPJ, Premium services are currently gaining adoption, and as you mentioned earlier, AI

recommendation tools are also expanding.

As you look to drive further growth going forward, what do you see as the main drivers that will increase

pricing levels?

For example, with Premium services, the pricing may be something like 50% higher than standard offerings.

But at the same time, one could also think that if adoption continues expanding further and further, eventually

that effect alone would run its course.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

So my first question is: going forward, when you think about further increasing ARPJ, will that mainly come

from pricing power, or can prices continue to rise significantly due to supply-demand dynamics?

And my second question is about take rate.

Ever since Indeed launched its service, the take rate has consistently remained below 1% for more than 10

years. So if we think of the current level as roughly 1%, where are we today in terms of the overall

opportunity?

In other words, is there a possibility that the take rate could genuinely exceed 1% at some point in the future?

Or because the market is so large, is the reality more that even over the next 10 or 20 years, it probably won’t

exceed 1%?

I do not yet have a strong sense for that, so I would appreciate your perspective.

Deko : That’s an excellent question.

Earlier, I talked about Premium Sponsored Jobs, and regarding the ability to raise pricing, this may be similar

to what I mentioned before, but I think that if more customers are actually able to hire successfully within, for

example, two or three weeks, then there is probably still significant room to continue increasing pricing.

And the reason is that, ultimately, this is a matter of supply and demand.

The number of people being hired does not fundamentally change.

Especially in cases where hiring previously took two months, companies were already spending a certain

amount of money to make those hires.

And if we already held 80% or 90% market share across the entire market, then perhaps there would not be

much room left for growth.

But in reality, most customers are still using other offline tools or relying on external recruiting services, so I

think a large part of the opportunity comes from taking share away from those areas.

And as I mentioned earlier, large and medium-sized companies often maintain dedicated teams to handle

these hiring processes and operational workflows.

For example, the other day I happened to be seated next to an executive from one of the largest companies,

not only in the US, but globally, at a dinner.

So I quickly looked it up using AI and checked how many applications had been sent to that company over the

past year, and it was something like 3.5 million applications.

So I said to the CEO, “You received 3.5 million applications last year. Even if it takes only three minutes to

review each resume, your team is spending an enormous amount of time on this.”

And now, with AI screening and related AI tools available, if we could reduce that workload by half, or even

down to one-third, then naturally when you ask, “Would you want to buy that?” the response becomes

something like, “Let’s set up a business meeting immediately.”

So I do not really think about it simply in terms of whether we can or cannot raise prices.

That is part of why my confidence has been increasing.

As for the question about how far we are from the 1% level, it is somewhat difficult because the denominator

itself is also growing, the number of hires facilitated through our platform continues increasing as well.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

So the calculation becomes more complicated.

Of course, the take rate has been rising.

However, the calculation differs depending on what exactly you include in the denominator — whether you

calculate it across the entire market including free job postings, or only based on Sponsored Jobs.

So we look at it from multiple perspectives.

As I mentioned earlier, our matching structure is designed to direct serious job seekers toward customers who

have strong hiring intent and are willing to pay, thereby increasing the total number of successful hires.

So in that sense, there is a difference between looking at take rate from the perspective of the overall platform

versus looking specifically at the take rate among paying customers.

Because of that, I think the interpretation can vary somewhat.

Going forward, we recognize that we need to think more carefully about how we disclose these metrics, and

we are considering ways to provide disclosures that are easier to understand.

Yoneshima : Understood. Thank you very much.

Shen : Thank you. We will now move to the final question.

Yamamura :Thank you for calling on me. This is Yamamura from J.P. Morgan Securities.

I know we are running over time, but if I may take advantage of your generosity, I would like to ask two

questions.

My first question concerns the potential with large enterprise customers, which you have touched on several

times already.

If we think only in terms of “volume multiplied by unit price,” it feels like eventually the discussion runs into the

question of where the ceiling may be. So I have expectations for enterprise expansion not only in terms of

scale, but also in terms of the potential for entirely new monetization opportunities.

At the same time, however, I wonder whether the style of sales required to penetrate large enterprises is quite

different from the style your company has historically used to support small and medium-sized businesses.

So personally, I am curious whether this really matches your company culture and core strengths. That would

be my first question.

Deko: That’s an excellent question.

It’s a very good point, and honestly, especially with Indeed, because we know Recruit so well in Japan, I do

feel that Indeed has historically been weaker when it comes to enterprise sales compared with Recruit in

Japan.

Since returning as CEO of Indeed in June of last year, I’ve approached this almost with the mindset of

rebuilding it from scratch.

Our goal is to become a strategic partner to key customers — not merely a transactional partner, to use the

English expression.

And in that area, we have been making substantial investments and changing the team structure as well.

We’ve been working on a variety of initiatives, and as I look at things now, I increasingly feel that the new AI

tools we are building are starting to have a meaningful impact.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

What I mean by that is that historically, I think we were actually weak in terms of how we presented value to

customers and demonstrated ROI.

I think we’ll probably be able to fully realize this quite soon. We’ve already built many of the individual

components, but for example, we are developing tools where a salesperson can ask AI, “How should I

approach this customer?” and the system will provide detailed guidance on what kind of sales pitch to use.

We’re building tools like that, and I believe this represents a major area of upside potential.

So I think this capability will continue improving going forward.

Yamamura : My second question relates to the job advertising market environment that you touched on earlier.

I understand your point that tech workers represent only a small portion of the overall labor market, and that

structural labor shortages fundamentally outweigh that impact.

However, investors have suddenly become very focused on medium- to long-term scenarios. And in the

medium term, there is concern that if the negative impact broadens significantly, then Indeed could face

meaningful pressure.

I want to dispel this concern..

This may relate to your earlier comments about “reorganization,” but even if the number of tech employees

were to actually decline in the future, I personally do not think people who are laid off would suddenly become

homeless the next day.

If optimization progresses across the overall labor market, or if labor mobility increases and people move

between jobs more actively, then I would think that is not necessarily a bad thing for the job market itself.

And on the employer side, since companies now have access to many different types of talent, there is also a

growing desire to avoid wasting effort and to focus more heavily on quality.

So it seems to me that these two dynamics together are aligning very well with your company’s strategy and

technology, and that this alignment may be one of the reasons behind the kind of standout earnings results

you delivered this time.

How do you see that?

Deko : As always, I prefer to look at actual data rather than rely on impressions.

And if you look at the research data released today by Indeed Hiring Lab, you can clearly see which sectors

tend to have older workforces. Just as you would probably imagine, the tech sector is not filled with people in

their 60s.

So then the question becomes: where are people in their 60s actually working?

For example, there are more older workers in areas like automotive repair, and there are also many in

healthcare.

When you think about what happens as those workers retire, I think the picture becomes quite different.

So rather than viewing this simply as “AI will replace workers across the board,” I think it is important to

analyze things at a much more granular level.

There’s one more point where I feel we still need more data before reaching firm conclusions, so I do not think

we fully understand it yet, but over the past several months, labor force participation in the United States has

actually been declining.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

And I think that is a rather unusual situation.

One hypothesis I have is this: if someone earning a relatively low salary loses their job, they cannot afford to

go without work, so they immediately need to find another job.

However, when people in tech or professional occupations lose their jobs, there are quite a few cases where

they can afford to wait a year before taking their next position.

So I wonder which parts of the labor market are driving such a significant decline in labor force participation.

And because labor force participation is declining, unemployment rates have also been coming down in recent

reports. But if you look at the details, the denominator itself is shrinking.

In other words, the number of people actively looking for work is decreasing.

That is not normally the kind of data pattern you would expect to see during a period when the economy is not

particularly strong, so I think it is a rather unusual development.

Of course, this is something the U.S. government itself probably still needs to investigate more carefully.

However, with respect to the next five years, I believe the data released this morning by Indeed Hiring Lab

provides a fairly accurate picture.

In reality, occupations such as truck drivers and construction workers are already being affected. And

especially because U.S. immigration policy has become considerably stricter, that impact is now becoming

very meaningful.

Sectors with high concentrations of immigrant labor — for example healthcare and others — are being

particularly affected.

When I think about all of these factors together, otherwise it would be difficult to explain why our revenue

performance has remained this strong, I do think there is a meaningful linkage there.

I hope that helps answer your question. I’ll continue thinking about it further as well.

Shen : We have gone slightly over time, but with this, we will conclude today’s financial results presentation.

Thank you very much to everyone for participating.

Deko : Thank you very much.

Arai : Thank you.

Internal data. For U.S. jobs hosted on the Indeed platform and % of business clients spending in the

same package MoM. Six-month average from October 1, 2025 to March 31, 2026.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Sell-side Analyst Follow-up

Page 35

Now, let me explain in more detail what I just discussed.

Page 36

Looking at the 6-year CAGR from FY2020, which was during the COVID-19 pandemic, through FY2026

guidance, revenue, EBITDA+S, and basic EPS will be 10.0%, 25.1%, and 33.3%, respectively, reflecting

steadily advancing our unique monetization and operational efficiency improvements, while simultaneously

improving capital efficiency.

Page 41

I will now explain our approach to capital allocation.

Page 42

Shareholder returns, the total payout ratio, and net cash at the end of March this year are as I explained

earlier.

The total number of shares outstanding, net of treasury stock, decreased by 13.2% since the end of March

2022 due to share repurchases and the cancellation of repurchased shares.

Since the end of March 2024, when we fully repaid the long-term borrowings associated with the acquisition of

the European Staffing business in 2016, as well as our corporate bonds, we have effectively maintained a

debt-free balance sheet. As a result, current net cash is approximately equal to gross cash.

Page 43

As mentioned earlier, we plan to maintain year-end gross cash around 750.0 billion yen through FY2028.

Consistent with historical precedents and considering our current business scale, we will maintain our gross

cash and cash equivalents at approximately 750.0 billion yen at each fiscal year-end as a necessary reserve

to prepare for potential changes in the economic environment.

This is to ensure that even in the event of a sudden downturn in economic conditions leading to stagnating

hiring activities among business clients and a subsequent decline in revenue, we can retain sufficient talent

and sustain operations during the recovery period, thereby preserving our ability to respond once the business

environment improves.

Additionally, the reason we changed from net cash to gross cash is to convey that, depending on the size and

number of future strategic acquisitions, we may take on additional debt at a healthy level. This could result in

net cash falling below 750.0 billion yen, and we could even shift into a net debt position.

Page 48

As I mentioned earlier, HR Technology is the substantive driver of our revenue, profit, and enterprise value.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Page 50

Let me review the results and outlook for HR Technology.

Q4 revenue on a dollar basis increased by 14.5% year over year, substantially above the updated outlook of

8.5% disclosed at the Q3 earnings release in February.

This was mainly driven by continued monetization development in the US and in Europe & Others, and a

recovery in Japan exceeding expectations.

Revenue on a yen basis increased by 17.7% year over year.

Page 51

Revenue, which was slightly below 4.0 billion dollars five years ago in FY2020, which at the time did not

include revenue from Japan job advertising and placement services, increased by about 2.5 times through

FY2025.

In FY2026, we expect revenue to exceed 10 billion dollars and increase by 11.0% year over year to 10.7

billion dollars.

Let me explain the assumptions and background behind this outlook by region.

Page 52

First, the US. In Q4 of FY2025, revenue substantially exceeded the updated outlook disclosed in February

and increased by 19.0% year over year to 1.41 billion dollars.

The US ARPJ growth rate, which has been disclosed since Q2 of FY2025 and reflects our monetization

development, was 25%, and the total number of US job postings declined by 5% year over year .

Page 53

US ARPJ growth rate for FY2026 is expected to be 18%, while it was 17% in FY2025.

In FY2025, full-year revenue increased 8.8% year over year, despite a 7% decline of the total number of US

job postings on Indeed year over year.

For FY2026, we anticipate a higher revenue growth of 13.6% year over year. This outlook is based on the

assumption that the total number of US job postings on Indeed remains broadly flat from April onwards,

resulting in a year-over-year decline of approximately 4%.

Page 54

We have received various perspectives on the possibility that advances in, and broader adoption of, AI could

decrease employment levels and, as a result, decrease the number of job postings on Indeed, which could in

turn decrease revenue in HR Technology.

We have also received many concerns from capital market participants about our relative competitive

advantage and durability versus current and future competitors who leverage AI.

First, addressing the question of structural vulnerability based on the concentration and diversification of our

US revenue by occupation,

Looking at full-year FY2025 US Sponsored Jobs revenue on Indeed by occupation, similar to the results from

Q1 to Q3 of FY2025 shared in February,

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

occupations in the “In-Person Economy,” such as Healthcare, which has the largest share, as well as Repair,

Maintenance & Installation, Transportation, and Food & Beverage, which rely on physical presence, skills, or

specific certifications, account for more than two-thirds.

By contrast, occupations considered more exposed to generative AI, such as Technology, Marketing, and

Finance & Accounting, each represent only low single-digit percentages, and together account for only about

15% of the total.

As this trend has been consistent over the past several years, we expect it to continue for the foreseeable

future.

Regarding concerns about many jobs being replaced by AI over the medium term, Deko mentioned at the

2026 Semafor World Economy Summit held last month in Washington D.C., US, that "although AI will impact

changes in jobs and skills, over the medium term, much of the decline in US employment will be due to

demographic changes, namely, an aging society and a decrease in immigration."

And regarding the feasibility of medium-term growth utilizing AI, as Deko mentioned earlier, in a Two-sided

Decision-making Marketplace, by fully utilizing AI to improve matching accuracy, we believe our revenue can

grow over the medium term.

This is because the virtuous cycle on Indeed sustains through the utilization of AI tools.

When business clients use our AI tools to hire effectively and efficiently, job seekers are more likely to be hired

faster by companies that meet their specific criteria. Therefore, job seekers become significantly more

satisfied with Indeed.

Because of this, the number of job seekers who want to use Indeed increases even further. In turn, this

encourages business clients to further utilize our sourcing products, Sponsored Jobs, and the AI features on

Indeed to hire top candidates. This creates a virtuous cycle.

Higher satisfaction from job seekers leads to companies spending more with us. This means transactions and

traffic on Indeed will increase, leading to employer spending increasing, and we believe this will enable our

medium-term growth.

Page 55

Europe & Others, including Canada.

In Q4 of FY2025, revenue increased by approximately 26% in the UK on a local currency basis and by

approximately 29% in Canada. For the region as a whole, revenue on a dollar basis increased by 29.9% year

over year.

For the full year of FY2025, revenue increased by approximately 14% in the UK and Canada on a local

currency basis. For the region as a whole, revenue on a dollar basis increased by 19.2% year over year.

This revenue growth was mainly driven by continued monetization development, specifically through growth of

Premium Sponsored Jobs in these markets.

In FY2026, the combined revenue of the UK, Canada, and Germany will continue to account for about

two-thirds of Indeed revenue in global markets excluding the US and Japan on a dollar basis. Driven by

further monetization development we expect revenue across the region to increase by 17.1% on a dollar basis

to 2.39 billion dollars.

Page 56

In Japan, in Q4,

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Indeed PLUS performed in line with expectations, and supported by a recovery in placement services, results

exceeded the outlook disclosed in February.

Revenue on a Japanese yen basis decreased by 2.2% year over year.

For the full year, with a focus on the second half recovery in placement services, the effect of the recovery

appeared earlier than expected. As a result, revenue decreased by 4.6% year over year to 348.2 billion yen.

The number of successful placements in our placement services this past March exceeded the results of the

same month last year. While it takes some time for a successful match to be recognized as revenue, the fact

that the number of successful placements itself is growing is very good news for us, and our full-year guidance

for this fiscal year is based on this performance.

Therefore, despite the negative impact on year over year revenue growth from changes in revenue recognition

from gross to net and despite the withdrawal from or downsizing of unprofitable businesses, we expect

revenue to increase by 2.1% to 355.5 billion yen, or 2.3 billion dollars.

Page 57

Q4 EBITDA+S margin increased to 38.7%, substantially above the 34.6% outlook disclosed in February, due

to revenue growth and further improvements in cost efficiency.

Page 58

As a result, full-year EBITDA margin for FY2025 was 33.5%. Share-based payment expense was 408 million

dollars, significantly below the initial plan of 570 million dollars due to the workforce reduction in July of last

year. Consequently, EBITDA+S margin reached a record high of 37.7%.

In FY2026, by continuing to increase revenue and improve operational efficiency, we expect EBITDA+S

margin to increase to 41.0%.

We are not planning a significant increase in headcount, and we expect share-based payment expense of 490

million dollars.

Page 59

In summary, for FY2026, we expect segment revenue of 10.7 billion dollars, an 11.0% year over year

increase, and an EBITDA+S margin of 41.0%.

This is a transformative change from FY2016, when we first began disclosing HR Technology separately and it

had revenue of 1.2 billion dollars and an EBITDA+S margin of 12.6%.

As Deko mentioned earlier, over the medium term, it is well within our reach to not only maintain double digit

annual revenue growth, but to achieve 20% or greater when hiring demand recovers. We expect our

EBITDA+S margin to exceed 50%.

Page 60

Staffing is part of the global HR Matching market, but unlike other HR matching services, it does not end with

simply providing a match.

In accordance with labor laws and employment practices in each country, throughout the period temporary

staff are on assignment, we undertake the functions necessary for employment, including social insurance,

benefits, and payroll. Our revenue is recorded on a continuous basis, consisting of the wages paid to

temporary staff, based on hours worked, and our margin.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

In Japan and Europe, US, and Australia, Staffing provides matching and support services for temporary staff

engaged in a variety of roles, including clerical positions, manufacturing and light-duty positions, and specialist

professional fields such as IT, engineering, and healthcare.

Page 61

In Q4 of FY2025, revenue increased by 10.1% year over year, with Japan increasing by 5.8% and Europe,

US, and Australia increasing by 14.6%, supported in part by foreign exchange rate fluctuations.

Full-year results for FY2025 were as I explained earlier.

In FY2026, in Japan, assuming changes in demand in the clerical staffing area, we do not expect a significant

increase in revenue and therefore expect revenue to increase by 3.1% year over year.

Meanwhile, in Europe, US, and Australia, we expect revenue to increase by 8.5% year over year on a yen

basis, supported by performance bottoming out in key markets, a gradual recovery trend, and the contribution

from large-scale project orders.

Page 62

The FY2025 EBITDA+S margin was at the same level as FY2024. In FY2026 as well, through continued

disciplined business management, we intend to maintain the same level.

Page 63

Lastly, Marketing Matching Technologies, or MMT.

MMT operates one of the largest matching platforms in Japan, connecting an account base of approximately

99 million

3 RECRUIT IDs of individual users and approximately 980,000 business clients

4 across multiple

verticals.

Page 64

MMT consists of Lifestyle, including Beauty, Travel, Dining, and SaaS solutions, Housing & Real Estate, and

Others, including Automobile, Bridal, Education, and other verticals.

Revenue is largest in Housing & Real Estate, followed by Beauty, and EBITDA+S margins in these

businesses are substantially above the other verticals.

Lifestyle, which includes Beauty, accounts for approximately 52% of the revenue of MMT.

Page 65

Let me discuss Housing & Real Estate. As I mentioned earlier, Housing & Real Estate is the largest business

within MMT, with revenue of 156.9 billion yen in FY2025.

It is one of the largest matching platforms in Japan

, providing information across various housing categories,

including new-build, existing homes, rental, custom-built homes, and renovation.

In Custom Homebuilding & Renovation Consulting, one of our divisions in Housing & Real Estate, accounting

for 20% of revenue in Housing & Real Estate, the business model has combined an advertising-based model

with a GMV-Linked Model under which we receive fees based on the contract value of custom-built homes.

From FY2027 onward, we will unify the business model under the GMV-Linked Model.

Page 66

I would like to take a moment to review the core characteristics and strengths of our MMT business model.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

The base of individual users with RECRUIT IDs combined with our points program maintains and increases

the number of actions taken on our platforms.

Our matching platforms are integrated with business support SaaS solutions and provide “fulfillment functions”

that efficiently complete a sequence of processes, from customer acquisition, or matching, to inquiry and

reservation management, individual user ID verification, and even payments.

By leveraging the unique data accumulated through these fulfillment functions, MMT uses AI to propose

improvements in services and pricing tailored to each business client, most of whom are small- and

medium-sized businesses, driving growth in their GMV, or Gross Merchandise Value.

At the same time, by shifting multiple platforms, such as Beauty, which we talked about in February, and

Custom Housebuilding & Renovation Consulting, which I mentioned today, from fixed monthly listing fees to a

“GMV-Linked Model,” we believe we can achieve sustained revenue growth even as AI technology becomes

more widespread and continues to evolve.

To provide more color on this, individual users with RECRUIT IDs use multiple vertical platforms through the

points program, and the cross-use rate

6 exceeds three-fourths, generating a total of approximately 400 million

actions

7 annually.

Approximately 90% of all these actions are linked to points

. This structure, which encourages individual users

to continue to take actions on our matching platforms, in other words driving strong retention among individual

users, provides resilience against changes in user acquisition channels, which are accelerating now and are

expected to continue to do so.

Our “fulfillment functions” function in Beauty and Dining as core operating systems that enable the

digitalization of operations through integrated business support SaaS solutions. In Housing & Real Estate and

Bridal, more than 200 consultation counters nationwide function as offline infrastructure that provides

end-to-end support for closing transactions.

Operational data, such as sales, reservations, service details, staff utilization, and customer attributes, is

accumulated as unique data in the databases of our business support SaaS and consultation counter

services.

By using this unique data to provide AI-driven proposals tailored to each business client, we aim to increase

business clients’ GMV and, in turn, drive our own revenue growth.

Page 67

Let me review the results and outlook for MMT. Q4 revenue in FY2025 was 139.8 billion yen, a decrease of

0.6% year over year, largely in line with the outlook we shared in February.

Full-year results for FY2025 were as I explained earlier.

As Ushida explained in February, in FY2026 we are gradually expanding the GMV-Linked Model in Lifestyle,

centered on Beauty, and we expect revenue in MMT to increase by 7.1% year over year.

In Beauty, in particular, we expect GMV-linked revenue of approximately 12.0 billion yen, and the business

model is shifting toward a structure in which growth in business clients’ revenue and our revenue are more

closely aligned.

In Dining, by expanding the “seat-hold” function that we launched nationwide in January this year and

generating more real-time actions, we will evolve the platform so that it better meets the needs of both

business clients and individual users.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

In Travel, operations required to complete bookings, such as real-time inventory adjustments, payment

processing, cancellation processing, and customer support, require an operating infrastructure that is distinct

from searching and comparing information, and we believe our competitive advantage can be maintained for

the foreseeable future.

On the other hand, as mentioned earlier, revenue growth in Custom Homebuilding & Renovation Consulting is

expected to temporarily slow in FY2026, as FY2026 will serve as a transition period to unify its monetization

model into the GMV-Linked model. Consequently, revenue growth in Housing & Real Estate will be in the low

single digits in FY2026. However, in FY2027, we expect the positive effects of this transition to materialize,

with revenue growth accelerating to the high single digits.

Page 68

Regarding EBITDA+S margin, in Q4 of FY2025, EBITDA+S margin was 11.3% due to a one-time impact from

the accounting system upgrade at fiscal year-end, the historical concentration of sales commission, sales

promotion expenses, and advertising expenses in the second half, and in addition, sales promotion expenses

for new initiatives in multiple fields aimed at increasing revenue from FY2026 onward.

As a result, in FY2025 EBITDA+S margin was 27.4%

In FY2026, by driving revenue growth and appropriately controlling personnel-related and other costs, we

expect EBITDA+S margin to increase to 30%.

There is also no change to our target of approximately 35% by FY2028.

Page 69

Historically, EBITDA+S margin in Q4 has declined significantly compared with Q1 through Q3.

Starting from FY2026, we intend to correct this imbalance by leveling out sales promotion and advertising

expenses throughout the year.

To achieve this we will reallocate quarterly sales promotion and advertising expenses based on the number of

individual user actions, such as bookings, that directly contribute to increasing business clients’ GMV.

At the same time, in the first half of FY2026 we plan to begin the kind of strategic investments we made in

Beauty in Q4 of FY2025 to increase business clients’ GMV and, in turn, our revenue from FY2027 onward,

while balancing the smoothing out and effectiveness of these investments.

That concludes my explanation of consolidated results and guidance, capital allocation, and the results and

outlook for each business.

We will now take questions for as long as time permits.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Q&A

Yoneshima: Your current performance is outperforming the guidance from the previous quarter, and the Q4

results appear quite strong. Could you explain what factors changed significantly between your outlook in

February and the actual results in March?

Arai: Before we discuss February and March of this year, I’d like you to recall the situation exactly one year

ago. At that time, discussions regarding US tariffs emerged, and our clients' hiring activities cooled down

almost instantly. To be honest, compared to that, when we formed our forecast in the last week of January, we

did not anticipate such a strong upward trend in February and March.

As Deko mentioned, we can now feel that our various initiatives are taking effect. Around this time last year,

Deko’s outlook was quite pessimistic, looking at the situation in February and March, it was a very difficult

period. You may recall us stating that growth in the US was effectively zero. This time, however, it is the polar

opposite.

As I touched upon earlier, we have engaged in executive-level outreach, and those efforts are now reflecting

in the numbers. The results are visible, and we are building a forecast that reflects this future visibility.

As noted in the earnings announcement, we have incorporated geopolitical risks into our figures. From my

perspective, there are many potential upsides that I haven't even factored in yet. We are essentially looking at

an extension of current trends. Furthermore, because visibility for the second half is still relatively low, we

have kept our forecast on the conservative side. Therefore, if new initiatives come in on top, or if visibility into

the second half improves, these numbers represent a baseline from which we can target the upside.

So, if you ask whether we have fully baked everything into forecast, I would say not necessarily. Looking at

Q4, we saw 25% US ARPJ growth, and the US saw very strong growth in March and April. In contrast, our

full-year US revenue outlook is 13.6%, which is lower on average because the forecast is structured to

account for risks.

While I believe you already consider our forecasts as strong figures, the picture will become even clearer as

the “ fog lifts. ” As our new strategic initiatives succeed, we may see even stronger numbers. For now, this is

the outlook we are projecting.

Maeda: Could you provide the cost outlook? I’d like to know your direction for personnel and marketing

expenses, and if possible, a breakdown by segment.

Arai: Regarding Stock-Based Compensation, you can assume that 97% of it is allocated to HR Tech, as I

mentioned earlier.

On the personnel side, our basic policy is not to increase headcount. While we will continue to reinforce

specific areas, the forecast is built on the assumption that we will not see a sudden increase in hiring. In the

HR Technology segment, even as revenue grows, we do not plan to increase headcount.

As for marketing, we will continue to invest a significant amount. However, at our current revenue scale, even

1% of revenue represents a massive budget. While we won’t be suppressing our marketing spend, there is a

limit to how much we can effectively deploy relative to our scale.

For those who have followed us for a long time, you might remember when a 30-second Super Bowl ad would

impact our margin by 1% or 2%. Those days are far behind us. Today, our scale is such that even if we were

to run such an ad, its impact would be virtually unnoticeable.

As revenue grows, and this relates to the 50% margin target I mentioned, if you model a 10% annual revenue

growth over three years while keeping costs flat, you can see how the numbers work out. It’s not a matter of

us "trying harder".

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Our business model is not one where we have to pile on costs to generate revenue. Since of course we have

strict cost control in place, whenever revenue increases significantly, it flows directly through to margin

expansion. This is the inherent business structure of HR Technology.

In the Staffing segment, we expect to maintain a stable performance as long as the topline remains steady.

Regarding MMT (Marketing Matching Technologies), while there are variations across different domains, we

are heavily focused on AI utilization and reviewing business operations. Consequently, we are confident in our

ability to achieve a 35% margin under normal operations. We are stating this with a high degree of conviction.

Zhai: Regarding this year’s guidance, I would like to ask specifically about factors that could lead the second

half to factors that could lead to second half differing from expectation.

For example, with initiatives such as Indeed Connect, could you elaborate in more detail on what kinds of

things could become upside drivers?

Arai: One possibility is enhancing the Premium package itself so that it delivers enough additional value to

justify even higher pricing. There are also opportunities to introduce options that are effectively an even more

premium version of Premium.

As Deko mentioned earlier, the key is to provide products that customers feel are valuable enough to pay

more for.

Of course, we also intend to continue expanding adoption of the existing Premium services — encouraging

more customers to use them on an ongoing basis, applying them across a wider variety of job categories, and

increasing usage among medium-sized and larger customers as well.

And if we are then able to layer additional offerings on top of that, it would not only contribute to upside in the

second half of this fiscal year, but could also carry over into next year’s growth as well.

That said, as I mentioned earlier, “haste makes waste.” If we move too quickly or push too aggressively, it may

not lead to sustainable growth next year and beyond.

What matters is letting customers experience the product, think, “This is really good,” and then naturally want

to use it more.

I often use the example of ice cream: you can give someone a free spoonful and say, “It tastes good, right?

You don’t have to pay today — just come back with your family this weekend and buy some.”

We do not necessarily need to monetize everything immediately.

The important thing is how to encourage customers to use the products extensively when the timing is right

and they genuinely see the value. Taking the time to drive adoption carefully and steadily will contribute not

only to the second half of this year, but also to next year and the year after.

What we want is to continue expanding the cycle I mentioned earlier and build that positive momentum larger

and larger over time.

Sekine :In Q2, Q3, and Q4, US ARPJ has consistently come in above expectations.

Is this mainly because existing Premium customers are increasing the number of actions they take, or

because, as Deko mentioned earlier, new customers with larger budgets are joining the platform?

What exactly has been driving these quarter-to-quarter changes?

Arai: Top-level strategic engagement doesn’t necessarily produce immediate results.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

If you look back at fiscal year 2025, the real driver was actually that more small and medium-sized customers

increasingly adopted and used Premium products.

Sekine: So more customers are gradually starting to use the product, initially on a smaller scale, and then

expanding over time?

Arai: And since these are year-over-year figures.

Munakata: Regarding the domestic MMT business. You briefly mentioned a very significant move, converting

"SUUMO Counter" (Custom Homebuilding & Renovation Consulting) to a GMV-linked model. Since the

housing domain inherently has a massive GMV, could you tell us what kind of impact you are anticipating?

Arai: Relationships with clients in the housing domain are highly sensitive, so there are limits to what we can

disclose to everyone. We want to avoid any perception that we are eating into our clients' profits, even though

that is not the case, since any small misunderstanding could damage the business greatly. I must be careful

not to say anything misleading.

However, our plan is to implement this shift even if it leads to a temporary dip in our revenue performance

initially. Once the value of this model is fully understood, we expect to see a significant boost. That is what we

are committed to achieving.

As you are aware, much of the housing domain is still an offline business. Shifting to a GMV-linked model is

only possible where we have digital touchpoints, which is why we are starting with the Custom Homebuilding

& Renovation Consulting sub segment in. Whereas like Beauty, the Lifestyle domain is already mostly

digitized. Since we cannot immediately convert offline businesses to GMV-linked models, we view this year as

a transition period. We are prepared for a temporary decline as we work to transition the overall structure.

This accounts for roughly 20% of our total revenue of Housing & Real Estate, so we are not converting the

entire business to GMV-linked billing all at once; we will consider the next steps as we go. We value our

clients and prioritize maintaining strong relationships with them. We view this as a new strategic challenge for

the business and hope to see the transition progress steadily.

Yamamura: Regarding the outlook for job posting volumes in the U.S., what you were likely expecting to be

down 5–6% in Q4 now appears to have improved to around minus 3%, with the rate of decline narrowing. If

this trend continues into next fiscal year, it looks like it could turn positive sometime in the second half.

Arai : The numbers are somewhat volatile, but we are broadly expecting a flat trend from FY2025 Q4 onward.

Yamamura : Is this also including what you mentioned earlier about geopolitical factors?

Arai : Yes, to some extent, it includes geopolitical factors, but we are not expecting any sudden boost.

Yamamura : Europe and Others is showing stronger growth relative to the U.S. There may be some foreign

exchange impact from the USD/EUR exchange rate, but is there also greater pricing tolerance than in the

U.S.?

Arai : There is roughly a one-year lag compared to the U.S. market, and since the decline prior to that was

quite significant, the weak performance two years ago improved last year, and we expect that trend to

continue this year.

Yamamura : Is there any regional difference in customer demand?

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Arai : Yes, that’s right. If you look at it by country, in Europe and Others, the U.K., Canada, and Germany

together account for about two-thirds of the revenue. If the revenue from Europe and Others is roughly

comparable to that of Japan, then even if the two-thirds portion is evenly divided among the U.K., Canada,

and Germany, each country would account for only about two-ninths of Japan’s revenue.

There are country-specific differences, so it is not possible to describe Europe and Canada as a single group

in terms of their characteristics. In particular, Canada has a different profile and is closer to the U.S. While we

do not disclose them separately at this stage because the revenue is still relatively small, we would like to

break them out individually and discuss them separately as each country’s revenue grows over time.

Nagao: Could I ask two questions? When you mentioned low single-digit growth in FY2026 and high

single-digit growth in FY2027, was that referring only to Housing & Real Estate?

Arai : As a result, the Housing & Real Estate segment overall is expected to grow in the low single digits this

year, but next year we believe growth could reach around 5% to 10%.

Nagao : For FY2026, overall MMT growth guidance is around plus 7%, so does that mean you are planning for

particularly strong growth in the Lifestyle segment this year?

Arai : A major factor is that Beauty is expected to contribute an additional 12 billion yen in revenue.

Oum : On the discussion around Indeed’s pricing, I think analysts are particularly focused on the assumptions

behind US ARPJ, and there seems to be some dispersion in the range. You also commented earlier on the

possibility of disclosing a take-rate-like metric.

Rather than simply saying that US ARPJ is increasing, it would be helpful for modeling purposes if you could

provide a sense of the range, for example, how much of the increase is driven by the gap versus alternative

channels such as retained search or staffing agencies. That would make it easier for us to build our estimates.

Arai : As a premise, I would like to reiterate that we are not in the business of operating on a take rate basis. I

would caution against using a take rate framework for modeling, as it can lead to misleading conclusions.

The reason Deko used that figure was simply to enable comparison with others who do operate on a take rate

basis. It was a calculation to show what the implied number would look like if it were calculated in that way,

only to serve as a point of reference.

To be more concrete, suppose I am the owner of a small business and I spend money on Indeed. If I spend a

total of 10,000 yen and end up hiring one person, then you are simply dividing the total spend by the salary of

the person hired. So it may look like “how much was spent to hire this one person,” but in reality, when posting

a job, you are not necessarily trying to hire just one individual.

Because it is an advertising model, you may be trying to hire multiple candidates of a certain type, rather than

making a one-to-one request such as asking a headhunter to fill a specific role.The nature of our business is

fundamentally different from that kind of headhunting model.

In a sense, currently we have no choice but to use a take-rate-like metric to illustrate how little we are actually

charging. The underlying philosophy is fundamentally different, so it is not really appropriate to compare it

directly with US ARPJ.

Oum: Idekoba-san mentioned the restaurant example. If a chef leaves and the restaurant initiates hiring, and

we assume the owner has a hiring budget of 100, is it possible to think in terms of what percentage Indeed

can capture versus what is spent on other channels?

Arai : I think we should distinguish between entities that do not have a hiring budget and those that do have a

hiring budget.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

If there is a defined hiring budget, then the approach becomes more about saying, “we would like to capture

about 30% of that budget.” For example, if 30% is being spent on advertising, 40% on headhunting, and other

portions allocated elsewhere or to internal hiring efforts, then we can position ourselves as saying, “please

allocate this portion to us.” This is essentially the TAM expansion argument Deko mentioned earlier.

As Deko mentioned earlier, the example where a company says, “If you can review 3.5 million resumes, we

are willing to pay for that,” is essentially an approach focused on reducing internal costs or reallocating spend

from other areas.

If a restaurant wants to hire a chef, it is about moving quickly and saying, “just do this for us.” It is about

directly acting on what the machine recommends and precisely matching the candidate.

Furthermore, other restaurants are also looking for the same type of person. If that candidate is hired by your

restaurant, we can then recommend another candidate instead, effectively increasing our revenue opportunity.

The nature of the action is fundamentally different.

Furthermore, other restaurants are also looking for the same type of candidate. If that person is taken by a

hypothetical Oum-san’s restaurant, we would then recommend another candidate instead, effectively doubling

our revenue opportunity. The nature of the action is fundamentally different.

Oum : Competition arises. For example, if that restaurant is willing to pay $50, then the other side would have

to offer more than that.

Arai : If you’re wondering, “Why can’t I hire anyone?”, the answer is simply that the offer you’re making is

lower than others. That is what the AI recommendation would tell you. That is the Premium service.

So the behavior of small and medium-sized enterprises and the way large enterprises spend money and hire

talent are quite different, and they should be considered separately.

However, we have not been that strong in the large enterprise area so far, so we believe there is room for

upside. On the small and medium-sized enterprise side as well, there is clearly upside, and they are willing to

use the product even at higher price points, as Deko has pointed out. And also he expects these dynamics to

continue playing out, and believes this is what will enable the kind of growth we are showing you.

Tsusaka : When the discussion focuses too heavily on US ARPJ, for example, investors often ask things like,

“How long can you keep raising prices?” but I think it can become an overly simplistic debate.

US ARPJ is simply an output metric. It is just the result you get after dividing one number by another. There

are many different things driving the numerator. And once you divide it by the number of job postings or hiring

activity, you get US ARPJ, that is all it is.

It is not a single product, although there may be certain standout products that contribute significantly more

than others. Revenue is being generated by capturing various kinds of recruiting related spending, and when

you divide that revenue by the relevant activity volume, US ARPJ emerges as the result.

So discussing it purely in terms of “how many percent can you increase it” becomes an oversimplification.

I think investors are viewing it in a way that is too simplistic.

As a company, you probably felt you needed to provide some kind of guidance metric, so you started

disclosing US ARPJ, and now everyone has become very excited about it.

Of course, higher unit pricing is a great thing. But it would be helpful if, each quarter, you could also discuss

the qualitative background in more detail.

That would make the conversation more constructive, rather than simply saying, “Premium Sponsored Jobs

are selling well.”

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

Otherwise, the discussion immediately becomes: “How long can that continue?” and “Eventually there must be

a ceiling,” and so on. So I think it would be useful to provide more context and background around what is

actually driving these trends.

Arai : A year ago, when we mostly spoke qualitatively, people told us, “Please give us actual numbers.”

So I’ll take your comment positively, that things have become easier to understand.

That said, there is still the question of, “Where do we go from here?”

The biggest reason we began disclosing US ARPJ is that, because this is a human resources business, there

is a very strong tendency for people to view it purely through the lens of market conditions. And if that is all

people focus on, then the stock simply becomes macro-driven and gets tossed around by external conditions.

At the same time, however, despite those challenges, during “Year 0” , Deko put many things in place, made

various efforts, and we are now beginning to see the effects.

As a result, even in an environment where market demand is not particularly strong, we are still able to grow

our revenue. For example, if the market declines 10% but we are still able to grow revenue by 5%, I think that

is actually quite remarkable.

But unless we disclose something like US ARPJ, it becomes difficult to demonstrate that point clearly, that

even while the market is declining, revenue is still increasing.

So in that sense, I feel we may have entered the next stage of the business.

Tausaka : Looking at this plan, I believe the HR Technology segment is likely approaching an incremental

margin of nearly 70%. The 50% shown in the presentation would continue to improve if the current marginal

profitability level is sustained.

Is this year’s incremental margin an exceptional outcome, or will this become a normal level?

The labor market and job market have not yet fully recovered, so is the contribution margin approaching

nearly 70% this year because the product has improved and we have been able to operate with relatively low

marketing spend? Or is this actually the normal level? If you plug the numbers into my spreadsheet, it comes

out as 50%. What should we think about this?

Arai: It will show up that way. But not this year. I’m not sure whether it will be exactly 41% or not, but that’s not

the point.

We are already using online marketing to spend as much as we can. We are also saying that we will not

increase headcount.

AI token costs, electricity, and AWS expenses will likely increase, but even when you include personnel costs

and other expenses, the cost growth rate will not exceed the revenue growth rate. That is the nature of our

business model.

Yamamura : You mentioned that you want to increase investment. What has changed behind that shift? Is it

that the direction has become clearer, or that your thinking has changed, including pricing and AI, compared to

before? Has uncertainty been reduced, or has your confidence simply increased?

Arai : Our investment is essentially only in people. So if AI means we don’t need to increase headcount, the

game changes.

We can develop products without having to dramatically increase the number of engineers, which makes us

far more efficient. In that sense, the situation is completely different from what it was two years ago.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.

The total number of RECRUIT IDs (as of March 2026)

The cumulative number of listed stores and business locations as of March, 2026, in Beauty, Travel,

Dining, Housing & Real Estate, Automobile, Bridal, and Education verticals, combined with the total

number of stores and business locations using SaaS solutions (if a single store uses multiple services,

it is counted for each service).

No.1 in the total number of listings across Japan for both sales (condominiums, houses, and land) and

rentals on comprehensive real estate portal sites (excluding sites that only post property information

from other companies), as of July 3, 2025 (based on research by Tokyo Shoko Research, Ltd.).

Calculated for the period from March 2025 to February 2026 by dividing the number of RECRUIT IDs

with a login record on two or more platforms or online services within the applicable verticals by the

total number of RECRUIT IDs with a login record on the Company’s matching platforms or online

services in the Beauty, Travel, Dining, Housing & Real Estate, Automobile, Bridal, Education, and

other verticals.

The annual aggregate total number of user actions such as reservations at salons, restaurants, or

accommodations, etc., or making inquiries or requests for materials regarding listings, services, etc.

(excluding cancellations), in the Beauty, Travel, Dining, Housing & Real Estate, Automobile, Bridal,

and Education verticals (FY2025).

Calculated by dividing the annual aggregate total number of actions associated with the standard

points awarded based on user actions by the annual aggregate total number of user actions such as

reservations at salons, restaurants, or accommodations, etc., or making inquiries or requests for

materials regarding listings, services, etc. (excluding cancellations), in the Beauty, Travel, Dining,

Housing & Real Estate, Automobile, Bridal, and Education verticals (FY2025).

Forward-Looking Statements

This document contains forward-looking statements, which reflect the Company's assumptions and outlook for

the future and estimates based on information available to the Company and the Company's plans and

expectations as of the date of this document or other date indicated. There can be no assurance that the

relevant forecasts and other forward-looking statements will be achieved.

Please note that significant differences between the forecasts and other forward-looking statements and

actual results may arise due to various factors, including changes in economic conditions, changes in

individual users’ preferences and business clients' needs, competition, changes in the legal and regulatory

environment, fluctuations in foreign exchange rates, climate change or other changes in the natural

environment, the occurrence of large-scale natural disasters, and other factors.

Accordingly, readers are cautioned against placing undue reliance on any such forward-looking statements.

The Company has no obligation to update or revise any information contained in this document based on any

subsequent developments except as required by applicable law or stock exchange rules and regulations.

This transcript is provided for the convenience of investors only and this is a translated version of the Japanese call.