Talent mapping in Japan: a complete guide

Step one: define the population, not the job title

Everything downstream depends on this step, and in Japan it does more work than anywhere else we operate. Japanese organisations recruit into the company rather than into the job, rotate people across functions, and grade them on a ladder that is internal to that employer.

The consequence is that a job title is a rank marker. Two people holding the same title at two companies may have nothing in common in scope, budget, team size or decision rights. A population filtered by title shares a label and little else, and it will look complete.

So the definition is written in terms of capability and demonstrated scope: what this person has actually run, at what size, with what accountability. It is agreed in writing, signed off before fieldwork, and reproduced on the front of the deliverable so nobody has to ask afterwards what was counted.

• The capability being counted, in operational terms rather than title terms.
• The employer set, named, including the organisations deliberately excluded.
• The level band, expressed as scope rather than grade.
• Geography — and whether supplier tiers are in or out, which in automotive and precision manufacturing changes the answer substantially.
• Language requirement, where the role genuinely has one.

Step two: frame it against official data

Japan has an unusually strong official statistical apparatus and it should be used for what it is good at. The Statistics Bureau Labour Force Survey frames the working population. The Basic Survey on Wage Structure gives occupational, age, tenure, firm-size and prefecture structure for pay. The Ministry of Health, Labour and Welfare job openings-to-applicants ratio gives monthly supply direction nationally and for every prefecture.

The frame tells you the order of magnitude and shows where the public record stops. It does not tell you who the people are, and the ratio in particular measures the public employment office channel rather than the professional or executive market.

Using the frame honestly means citing the table and reference period, and saying what it cannot reach. That boundary is where the primary research begins.

Step three: build it in Japanese

The Japanese public record is Japanese-language. Company disclosure, professional and academic registers, technical, patent and conference material, industry bodies and trade press are all primarily in Japanese, and the English layer over them is thin, late and selective.

This matters more than it sounds, because the under-representation is not random. An English-only process skews away from older, more senior, more technical and more domestic populations — which is usually exactly the population the brief is about.

In automotive and precision manufacturing there is a second structural point. Engineering capability is distributed across tier-one and tier-two suppliers rather than concentrated inside the headline manufacturer, so mapping a competitor honestly often means mapping its supplier network as well. That belongs in the scope decision, not in a surprise mid-project.

Step four: resolve the names properly

A Japanese personal name written in kanji can carry several legitimate readings. The same individual therefore appears under more than one romanisation across different sources, and entity resolution performed on the romanised string makes two errors at once: it splits one person into several records and merges several people into one.

Neither error is visible in the finished file, which is what makes it dangerous. A map with duplicate and merged records looks cleaner than one that has been reconciled properly.

Corporate names compound it. A group’s Japanese name, its registered English name and its trading name differ, and subsidiary structures are reported inconsistently between them. Reconciliation runs on the Japanese form, with the original characters kept in the dataset so any record can be audited back to its source rather than to a transliteration of it.

Step five: verify, grade and state the limits

Verification means corroborating each material field against a second independent source where one exists, and recording a confidence grade where one does not. It is the entire cost difference between a list and a map.

Corroboration rates in Japan vary sharply by population. Listed-company senior populations corroborate well. Engineering populations inside privately held suppliers corroborate far less well, and a map that reports one confidence level across both has not measured either.

The legal boundary is part of the method rather than a footnote. Research runs under the Act on the Protection of Personal Information: professional facts only, from lawful sources, with the source recorded; no special-care categories recorded or inferred; and no approach to a competitor’s staff under a pretext to extract information about colleagues.

What the finished map contains

The written population definition on the front. The named population with employer, function, demonstrated scope, approximate level, tenure and location, each field carrying its source or its confidence grade, and the Japanese form of every name retained.

Competitor organisation structure reconstructed layer by layer with inferences labelled. Reachability assessed against notice, tenure patterns, recent movement and the conventions that actually govern senior moves in this market. Supply, scarcity and realistic lead time. Aggregate representation at population level only.

And an expected useful life with the fields most likely to decay named — in Japan, scope before employer, because internal rotation changes what someone runs without producing any external signal at all.