Reverse Talent Maps: Read a Company by Where Leavers Go

The most honest description of a company is a list of where its former employees went. People leaving say polite things. Where they land ends up on a resume or in the new employer's announcement, and it doesn't flatter anyone. I call the picture you get from those destinations a reverse talent map, and in Japan it is often the only clear view you'll get of a company from the outside.

What a reverse talent map shows

An ordinary talent map starts from a role and asks where the people who could fill it work today. A reverse map starts from one company and asks where the people who left it went.

Same data, opposite direction, different questions answered.

The ordinary map helps you find people. The reverse map helps you read a company. How does the market value its people? Where is it blocked inside? Who is hiring from it on purpose? Which industry do its veterans end up in? That last answer is often a preview of where the company itself is heading.

It also fills a gap that exit interviews leave. I wrote about why exit surveys never tell you why people left. A reason given on the way out can be a courtesy. A destination is a fact.

Where the destinations are written down in Japan

For executives at listed companies, almost everything is public.

Every annual securities report (yuho) has a section on officers. The format set by Cabinet Office ordinance requires each director and auditor's name, date of birth, career history, term and shareholding. The career history usually gives the month and year of each move and the title held at the previous employer. Read one company's report and you see where its executives came from. Search many companies' reports for the same former employer and you see where that employer's people went.

When a listed company changes its representative director, the Tokyo Stock Exchange requires timely disclosure. "Change of representative director or representative executive officer" is on the list of decisions that must be announced, and the release usually includes the new person's history.

Below the executive level nothing is mandatory. You work with LinkedIn profiles, the "we welcome Mr. X" releases that hiring companies put out, the personnel columns in trade press, and the resumes of candidates you've interviewed yourself. LinkedIn coverage is thin in Japan, as I wrote in Is LinkedIn Useful in Japan?, so the further you go from the executive suite, the more you rely on your own interview records.

One rule worth keeping: a reverse map looks at aggregate patterns and never builds files on individuals. "Half of the twenty-odd sales managers who left in three years went to the same company" is the unit. What any one person is doing now is not.

Which patterns matter

Two dimensions do most of the work. Where people went, and whether they moved up a level when they got there. Then you lay the timing of the departures on top.

Where leavers went and whether they moved up says something different in each combination

How to read a reverse talent map. Horizontal: destination. Vertical: change in level at the new employer. Cell text is the author's reading; the top right draws on Somaya et al. (2008), a study of patent attorney mobility

Across: to competitors / to clients and partners. Down: moved up / same level or lower

To competitors, a level up

Good at developing people, short on promotions. If one destination dominates, that company is your real talent competitor and the best place to look for people.

To clients and partners, a level up

The industry rates its people. They take relationships with them, and business often flows back to the old firm.

To competitors, same level or lower

Something inside is pushing people out: a boss, a policy, the business itself. Check whether departures cluster after one event.

To clients and partners, same level

Often a group secondment or transfer, or a stopover before leaving the industry. Read it together with timing.

Headcount lost tells you nothing by itself. Watch the people who joined a competitor one level up.

Start with concentration. If a company's leavers scatter evenly across twenty employers, its people are broadly employable. If half of them went to one company, that company isn't getting lucky. It's recruiting from them on purpose. In the two maps I use, that company goes in the talent-overlap column, whether or not its business looks anything like yours. One caution: with fewer than about twenty departures, I treat any pattern as a lead and hold off on conclusions.

Level change tells you about the inside of the company they left. Lots of people stepping up a level on the way out means it trains people well and promotes few of them. Lots of sideways or downward moves means people were pushed rather than pulled.

Moves to clients have research behind them. Deepak Somaya and colleagues published a study in the Academy of Management Journal in 2008 that tracked patent attorneys moving between law firms and Fortune 500 companies. Roughly, when an attorney left a firm for a client company, the firm's business with that client tended to improve. Losing people to competitors was what hurt. Leavers take relationships with them, and sometimes leave them on your doorstep.

Timing points to events. James Walsh, in the Strategic Management Journal in 1988, followed the top managers of acquired companies for five years. Their turnover was clearly above normal rates, and the most senior and visible people left soonest. Acquisitions, a new president, a new HR system: departures bunch up after each. If the months cluster after one event, that event is your lead.

The best-known reverse map

It's probably Fairchild Semiconductor's. Eight people founded it in 1957, and over the following years its employees kept leaving to start their own companies. Intel and AMD both trace back to it. In 2014 Endeavor Insight, the research arm of the nonprofit Endeavor, counted 92 public companies that could be traced to Fairchild's founders and employees, worth about $2.1 trillion together.

Read it with the figure. Destination: startups they founded themselves, clustered in one region and one industry. Level: from employee to founder. Fairchild's output turned out to include a group of people who could build companies from zero. I don't have reliable year-by-year data on when they left, so I won't read the timing.

Japan has its own version in a single word. "Moto-Riku" means a Recruit alumnus, and you hear it constantly in startup and HR circles. When a former employer's name becomes a nickname for its alumni, someone has already drawn the reverse map in their head.

Draw your own company's first

A reverse map pays off in two everyday decisions. When hiring, it tells you which companies to approach: the ones whose leavers routinely move up are valued by the market and have people who are restless. When sizing up competitors, it finds the one taking your people, which is often a company you never put on the list. It can also tell you which of the three rings in A talent map is not a list to work first.

The most useful version, and the least comfortable, is your own. Who left in the last two or three years, where they went, whether they moved up, when they left. If several of your best people went to the same company at a higher level, you can stop reading the "family reasons" on their exit forms.

Where did the last five people you were sorry to lose end up?

---