The Conditional Case for a Buddy-Based Job Search

The evidence about Overstapbuddy.nl supports a specific answer to the title question. A successful career-switch buddy can change a job search by making an unfamiliar occupational path more legible, more socially credible, and easier to keep open long enough to act on. The intervention did not simply push people to submit more applications, and it did not deliver a guaranteed dose of mentoring. It offered a relationship inside a public employment service, and that offer improved later labor-market outcomes for a defined group. The final judgment is therefore conditional: the human connection is promising when it meets a constrained transition problem and is embedded in a service that can make contact meaningful, inclusive, and accountable.

The change is in the option set, not just the effort level

A job seeker considering a new occupation faces more than a shortage of vacancies. They need to understand what unfamiliar work involves, decide whether past experience can transfer, judge the cost of training or a temporary loss of stability, and present themselves to employers who may not immediately recognize their fit. A vacancy database can provide openings. A professional adviser can organize information and services. A successful switcher can add something different: an example of a person who has crossed the same kind of boundary and can describe the route without reducing it to a slogan.

That difference explains why the buddy model is best understood as socially supported option expansion. The buddy may help a job seeker keep several occupations visible, compare them more realistically, and recognize a next step that was previously too abstract. They may provide practical search knowledge, a more concrete translation of transferable skills, or a relevant contact. They may also make it less costly to admit uncertainty. None of these contributions requires increasing the number of applications. A person can search more intelligently, or avoid narrowing prematurely, without appearing more active in an administrative job-search record.

The study’s measured pattern fits that interpretation better than a simple effort story. The treatment did not significantly raise application counts, motivation to search in new occupations, or measured search intensity. The control group gradually reduced the number and breadth of occupations it considered, while the treatment group remained comparatively stable. The breadth differences were marginally significant, so they are not a proven mediator. But they identify a meaningful possibility: the platform may have helped preserve an option set until a job seeker could act on it.

The employment results make the possibility consequential. For the analytic sample of people who had worked zero hours before the invitation, the average effect was small during the first twelve months and became visible in months 13–18. Treated workers were about 6 percentage points more likely to be employed and earned about €226 more per month in that later period. Over 18 months, cumulative earnings were €2,185 higher. The delay matters. It is compatible with a process in which someone first interprets a transition, investigates alternatives, adjusts a search, and only later reaches an employment outcome. The study did not observe every step well enough to prove that process, but it makes a rapid “motivation boost” explanation incomplete.

The strongest result appeared among people unemployed for more than 17 weeks. Their employment increase began at roughly 9 percentage points, later settling near 6 points; cumulative earnings after 18 months exceeded the control group by more than €5,000. The shorter-term unemployed did not show statistically significant effects. This pattern suggests that the value of a career-switch buddy may rise when a search has had time to narrow and familiar routes have become less productive. It does not prove that the seventeenth week is a causal threshold. Longer-term unemployed participants also differed in age, education, and remaining benefit eligibility, and the study was not large enough to establish a universal moderator.

The connection among design, breadth, and timing is the series’ central model.