A competitor's career histories are a map of where to source
A search tells you who works at a competitor today. It does not tell you how they got there. The dated work history does: the employer each person left to join, and, for the people who have since gone, the employer that took them.
Read across a few dozen histories and two patterns appear. The feeder companies are the employers the competitor keeps hiring from. People still working at those feeders sit one step earlier on a route the competitor has already proved, and few recruiters think to look there. The destination companies are the ones hiring the competitor's people away. They are rivals for the same talent, and employers with a proven appetite for that function.
Run it on a client's competitor and the same two tables answer a question clients rarely get evidence for: where do they find their people, and who is taking ours?
Dated history is bought per person, so this is a sample by design. The cap is agreed first and printed on the output.
What you get back
Two ranked tables for one company and one function. Feeders: the employer held straight before the competitor, how many of the sample came from it, and the years of those moves. Destinations: the employer joined straight after, with the same columns. Under both sit the sample size, how many leavers were confirmed out of how many were found, the employer names that were grouped together, and the pull date.
Nothing is sent, and no contact details are bought.
Variations worth knowing
One half only. If the question is where to source, skip the leavers. The feeder table is the more reliable half.
Split by seniority. Feeders for junior hires and for leadership hires are rarely the same. Ask for the tables cut by level.
Turn the feeders into a search. The top feeders become the source list for source from named competitors with your off-limits list enforced. For the wider market, see map a whole talent pool.
Where this goes wrong
Reading a sample as a census. A few dozen people out of several hundred show where the clusters are, not the full distribution. An employer that appears once is an anecdote. Quote the sample size with the table.
Leavers are harder to find than joiners. They come from a web search of public profiles, not from a database filter, and only count once dated history confirms them. That half is smaller and noisier.
Profiles lag and omit. A history holds what the person chose to write. Short stints and recent moves go missing.
Employer names do not match themselves. A group, its subsidiaries and an old trading name can be one feeder split three ways. The grouping is shown so you can correct it.
Small companies return little. A boutique competitor may have too few indexed people for a pattern to be real, and the output says so.
