The job may not be missing. You may be looking in the wrong channel.
Imagine two job seekers in Japan: a software engineer and a restaurant operations worker. Both check the same major job board every evening. If the results look thin, does that prove opportunities are scarce for both?
An analysis released on October 8 by recruitment data company Frog categorized full-time job ads captured in July 2026. Among the IT engineer ads in its selected media, 51.8% appeared in direct-recruitment channels. In food service, 78.6% were attributed to Indeed PLUS.
Neither figure measures who actually got hired through that channel. They are shares of collected listings. But they do suggest that different occupations may require different starting points for a search.
For IT engineers, searching and being discoverable are different tasks
Within the sampled IT-engineering ads, 51.8% were classified as direct recruiting, 23.1% as recruitment agency listings and 19.3% as Indeed PLUS. Frog classified services such as Mynavi Scouting, AMBI and MIDDLE NO TENSHOKU in the direct category.
Scouting platforms let employers and recruiters discover candidates from their professional profiles. Someone using them might therefore test two things separately: how many relevant postings they can find, and whether their experience description is specific enough to prompt relevant outreach.
A title like 'backend engineer' may be less informative than an account of systems owned, technologies used and measurable work completed. None of this proves that a profile generates offers: listing share is not the same as individual response or hiring probability.
Indeed PLUS is a distribution network, not one more job board
Food-service postings were heavily concentrated in the Indeed PLUS category: 78.6% within the analyzed sample. That could reflect the practical need for employers with multiple locations to distribute many vacancies efficiently.
But Indeed's own documentation describes Indeed PLUS as a network that distributes jobs to participating job sites. A job seeker may encounter its postings on different partner services, rather than on a stand-alone site called Indeed PLUS.
Creating accounts on several connected sites may expose someone to overlapping openings. A more useful check is whether the ad leads to a current vacancy on the employer's own recruitment page and whether the role meets the applicant's conditions.
Sales, administration and manufacturing showed a wider spread
For sales and office roles, Indeed PLUS accounted for 31.9% of sampled ads, direct recruiting 31.1%, and recruitment agencies 24.2%. For manufacturing and factory positions the comparable figures were 32.5%, 25.0%, and 25.3%.
Unlike the IT example, no single category represented the majority of sampled postings. For these occupations it may be more productive to compare genuinely different mechanisms: searchable ads, scouting and agency channels, then the employers' careers pages.
Count distinct opportunities, not screens of results. One company can publish several versions of the same vacancy across multiple platforms. Company, role, location and employment conditions are useful deduplication keys.
Tokyo's channel mix looked very different from regional Japan
In Tokyo, 43.0% of sampled ads fell into direct recruiting, 20.7% into agencies and 5.5% into Hello Work. The Hello Work share was 30.3% in Kyushu/Okinawa, 30.1% in Chugoku/Shikoku and 28.4% in Hokkaido/Tohoku.
For a candidate considering a regional employer, the government-run Hello Work job search can provide a complementary set of openings to compare with commercial platforms. Its official service allows users to search jobs received by Hello Work offices across Japan.
Occupational and geographic figures were calculated separately. It would be invalid to combine them into a made-up statement about how many Tokyo IT engineers get hired via scouting.
A channel's share of ads is not its success rate
The study gathered full-time job ads from a defined set of platforms. It did not represent every Japanese vacancy, and multiple postings by one company were not removed as duplicates.
Historical comparisons require extra care. Indeed PLUS launched in 2024, changing both distribution and parts of the covered platform universe; the 2026 aggregation included some partner media beyond Recruit's own outlets.
Therefore a falling share for Hello Work does not mean the same percentage fall in actual hires through that service. Nor does the 78.6% food-service figure imply that applicants have a 78.6% chance of success on Indeed-connected listings. Exposure, eligibility, application and hiring are different stages.
A seven-day experiment to find your own best channels
Start by specifying the role and geographic area you actually want. Then compare two or three different discovery mechanisms for seven days: for example, a searchable job board, a scouting or agency service, and Hello Work where relevant. This is Banseog's practical suggestion, not a validated formula from Frog.
Give each channel a consistent amount of attention. Record newly discovered companies and distinct roles after deduplication, whether you meet the requirements, whether salary and location are stated, and whether the employer confirms that the opening is still active. With scouting, evaluate relevance rather than counting messages.
At the end, prioritize the sources that generated genuinely new, eligible opportunities for you. That is a more useful personal signal than the total number of ads on any one site.
BANSEOG VIEW | Visibility is not the same as opportunity
Japan's hiring market cannot be understood through one board alone. Job ads are distributed differently for IT engineers, hospitality staff and regional workers.
Yet the research cannot rank channels by placement success. Some services share an underlying distribution network, so more registrations are not necessarily better. Distinct types of openings and the ability to apply are what matter.
Before concluding that your market has no opportunities, ask whether the sources you are checking actually represent the hiring routes for your role and location.
BANSEOG VIEW
Banseog View — where jobs appear versus what you can actually pursue
Percentages such as 51.8% and 78.6% are shares of job listings within selected platforms, not real-world placement success rates.
Indeed PLUS distributes ads to connected partner sites, so multiple platforms can expose the same underlying opportunity.
Comparing distinct, eligible and still-open vacancies across different search mechanisms is more useful than simply maximizing website registrations.
SOURCES
Primary sources and references
- Frog Co., Ltd. — October 2026 hiring-method trends report
Published October 8, 2026. Covers July 2026 full-time job listings by occupation and region, sampling rules, duplicate ads and comparability caveats.
- HRog — Analyzing recruitment channels with Japanese job-listing data
October 8, 2026. The research provider's editorial presentation of profession- and region-level channel differences.
- Indeed Help — What is Indeed PLUS?
Official explanation of the multi-site job distribution network and participating services.
- Japan Ministry of Health, Labour and Welfare — Hello Work job-seeker service
Official service to search opportunities received by Hello Work offices across Japan.
These values describe the composition of full-time listings collected by Frog from selected media in July 2026, with no removal of multiple ads placed by the same company. They do not represent every job in Japan or actual hire and placement rates. Comparisons with 2023 are affected by changes in the distribution system and sample after the launch of Indeed PLUS. The seven-day channel test is an independent Banseog suggestion, not a demonstrated predictor of hiring success.