Why Recruiters in Japan Keep Sending the Same Candidates

If three agencies in Japan send you the same few candidates, nobody is cutting corners. They are fishing in the same small pond. Of the 10.23 million people in Japan who say they want to change jobs, only 3.6 million are actually looking. Agencies and job boards mostly see those 3.6 million. The person you want is more likely in the other 6.63 million.

How many people who want a new job are actually looking?

About one in three. Japan's Labour Force Survey, run by the Statistics Bureau of the Ministry of Internal Affairs and Communications, asks employed people whether they want to change jobs. In its detailed tabulation for 2025, 68.2 million people were employed. Of those, 10.23 million wanted to leave their current job for another, or wanted a second job on top of it. The survey calls them people wishing to change jobs.

Of the 10.23 million, 3.60 million were job seekers, meaning they were actually searching. The other 6.63 million were not.

For comparison, 3.30 million people in 2025 had left a job in the previous twelve months and moved to a new one. That number is a flow and the others are annual averages of a stock, so you can't subtract one from the other. The ratio still holds: about 65% of people who want to move are not looking yet.

Two thirds of people in Japan who want a new job aren't looking, and agencies mostly see the other third

People wishing to change jobs and their breakdown, in millions (bar length relative to 10.23 million). Statistics Bureau of Japan, Labour Force Survey (detailed tabulation), 2025 average

Want to change jobs　of 68.2 million employed

• Wish to changeor add a second job
  10.23m
• Not lookingnon-job-seekers
  6.63m
• Lookingjob seekers
  3.60m

For reference　changed jobs in the past year

• Job changers
  3.30m

Shortlists come from people who are looking. The right person is often among those who aren't.

Why the same names keep coming back

Because candidates spread themselves across channels, and agencies are paid to close fast.

Start with the candidates. People who are looking rarely use one route. Japan's Ministry of Health, Labour and Welfare (MHLW) asked recent job changers in its 2025 Survey on Job Changers which methods they used, with multiple answers allowed: job sites and job magazines 45.1%, Hello Work and other public offices 29.6%, personal connections 22.2%, private employment agencies 20.9%, company websites 13.5%. The ministry's annual report on placement businesses even carries a footnote saying its counts of job seekers and job orders include duplicates, because both sides use several agencies at once.

Now the agencies. Most are paid only when a candidate starts. A consultant with limited hours will spend them on people who can close soon: already looking, clear on salary, free to interview next week. That's a reasonable response to the incentive. I wrote about the same mechanism from the candidate's side in Why recruiters in Japan don't reply to you.

Then the tools. Many searches start with keywords in a database: job title, industry, Japanese level, a technology. Similar databases plus similar keywords give similar lists. People with tidy, keyword-rich résumés get found again and again. Someone doing the same job with a profile they haven't touched in five years doesn't come up for anyone.

So three agencies often hand you three copies of one list.

How do you reach people who aren't looking?

You find out where they work first, then someone contacts them directly. Search can't do the first part. A map can.

A talent map starts with companies, not people. Which companies have people doing the job you need, roughly how many, at what level. The first circle is the same job in the same field. The second is adjacent fields where the daily work is similar. The third is further out and needs a development plan. I laid out the method in The two maps: who competes for your customers and people. Once the map exists, the few dozen or few hundred people out of those 6.63 million who matter to you finally have names and employers.

You'll also see how few of them are registered anywhere. Most need a direct approach, and the first answer is usually "not right now." That's fine. A talent map was never meant to be a list of job seekers, which is the point of A talent map is not a list.

What do you say once you reach them?

Answer the reason they might move. Your company pitch comes second.

Someone who isn't looking owes you nothing. The best clue to what might move them is why people who did move left. In the same MHLW survey, 81.3% of job changers had left their previous employer by their own choice. Their top reasons, up to three allowed, were low pay (25.5%), worry about the company's future (25.2%) and unsatisfying work (24.7%).

Each of those has a concrete answer. On pay, give the range and how it is built, not "competitive." On the future, say what the Japan entity plans for the next two years, where the money comes from and who decides. On the work, say what the person does in the first six months and who they answer to. My sense is that an approach message that does this gets more replies than one built on "fast growth" and "big impact." That's my judgment. I have no data on it.

An agency can send the message for you. Only you can supply what goes in it. Without that, the agency has nothing to send but a template.

What the hiring company controls

The brief, first. Most briefs are a job title and a string of requirements. The quickest thing an agency can do with that is type the requirements into a search box. If you want a different result, write down what the role actually screens for, which requirements can give, how long each takes to learn on the job, and what you explicitly don't need. The more your brief reads like keywords, the more your shortlist will look like everyone else's.

Then what you ask for. Ask for the map before the résumés: target companies, estimated headcount at each, who has been approached and what they said. An agency that can produce that is searching. One that can only produce résumés is filtering a database. Both have a use. They shouldn't cost the same, and you shouldn't expect the same from them.

Then where each candidate came from. For every résumé, ask whether the person registered with the agency or was approached, and whether they agreed to this specific role before being put forward. If nine in ten are registrants, you are probably looking at the same pond again. Agree up front what happens when two agencies submit the same person. A simple rule is that the earlier documented consent from the candidate wins.

This costs more. Approaching people who aren't looking takes longer, gets turned down more and usually means a higher fee or a firmer mandate. It isn't worth doing for every role. For routine, high-volume hiring, the 3.6 million who are looking usually include someone good.

So before the next brief goes out, it's worth asking which kind of role it is. If the honest answer is that the best person for it probably isn't looking, a fourth agency won't help. A map will.

---