Pharma Bloodbath X: There Is No Secret Door to the Hidden Job Market — There Is a Referral Economy

[A companion to an earlier piece here, Job Hunt Myth Debunking Part II: No, 85% of Jobs Are Not Filled Through Networking. This one goes deeper on where the myth came from and what the DACH data actually shows.]

Dear #MoreThanCareer community,

Let us perform an autopsy on the most beloved corpse in the careers industry: the hidden job market. It has been declared dead many times. As we say in German, totgesagte leben länger (the ones declared dead live longest), and no statistic has clung to life more stubbornly than this one.

Where the famous 80 per cent actually came from

You have heard the number. Seventy per cent. Eighty per cent. Eighty-five per cent if your coach is feeling ambitious that month. The claim is always the same: most jobs are never advertised, they live in a shadow economy, and only those with the secret handshake get in.

The statistic has a source. It is just not the source anyone hoped for. Tracing it back leads to a 1966 article by a career counsellor named Bernard Haldane, citing a Ford Foundation-funded pilot survey of business vacancies conducted in Rochester, New York.² It was later popularised by Richard Bolles in What Color Is Your Parachute? and cemented by a 1980 New York Times interview.² So when your LinkedIn coach quotes “80 per cent” at you in 2026, they are quoting a regional survey of mostly industrial vacancies from the era of the rotary telephone.

The modern “70 to 80 per cent of jobs are hidden” version is worse: it has no traceable source at all.¹ Follow the citations and you get articles citing articles citing a vague “LinkedIn study” that LinkedIn never published.¹ A coach’s estimate became a statistic the way a rumour becomes a fact, by being repeated confidently by people with something to sell.

What is actually true, separated from what is merely profitable

Jobs do get filled through contacts. This is real, well documented, and boringly consistent across decades of peer-reviewed labour economics. The credible range is that roughly 30 to 60 per cent of workers find their jobs through personal contacts, and around 40 to 50 per cent of employers actively use their existing employees’ networks to fill openings.³ Notice the gap between that and 80 per cent. Notice also that “filled through a contact” is not the same as “never advertised.” Most of those roles were posted somewhere. A contact simply got the candidate through the door faster.

For DACH specifically, we have the gold standard: the IAB-Stellenerhebung, the only representative, statistically valid survey of how German vacancies are actually filled. In 2024, employers used personal contacts or their own staff in 53 per cent of successful placements, and that channel was the decisive one in 31 per cent of new hires.⁹ In the freshest release, covering 2025, those numbers fell to 48 per cent used and 27 per cent decisive, because online media are rising fast.⁸ For comparison, in 2022 personal contacts were decisive in 37 per cent of hires.¹⁴ So the honest German number is not 80 per cent. It is somewhere around a quarter to a third of hires where a contact made the difference, and it is trending down.

Then there is the referral data everyone loves to quote: only about 7 per cent of applicants come via referral, yet referrals account for roughly 40 per cent of hires, with a referred candidate’s chance of being hired running near 28 per cent against under 3 per cent for the cold applicant.⁷ One caveat I will not skip: that data comes from Jobvite, a company that sells referral-recruiting software. Treat it as directional, not gospel. A firm selling umbrellas is rarely your most neutral source on the weather.

So what is the “hidden job market,” honestly?

It is mostly two unglamorous things wearing a trench coat.

The first is a referral economy. Not secret jobs, but ordinary jobs where a known name skips the queue. The role exists, it often gets posted, and you still go through the HR process. The referral does not buy you the job. It buys you a reader.

The second is timing. Many roles are filled quickly, sometimes before they are widely advertised, because a hiring manager already had two plausible names in mind before the requisition opened. The job was not hidden. You were simply late, and the people who were early had been talking to that manager for months about nothing in particular.

Which networking works, and which is interpretive theatre

In 2022, a team from MIT, Stanford, Harvard, and LinkedIn published the largest causal study ever run on this question in Science. They used LinkedIn’s “People You May Know” algorithm to run randomised experiments on more than 20 million people over five years, during which roughly 2 billion new connections and 600,000 new jobs were created.⁵

The result overturned the comfortable cliché in a precise way. Weak ties do help you get hired more than strong ties do, confirming Granovetter’s 1973 theory.⁴ But the relationship is not linear. It is an inverted U.⁵ The connections that generate the most job mobility are not your closest colleagues (they know the same jobs you already know) and not the weakest random strangers either. They are the moderately weak ties: people you sort of know, the acquaintance two desks over at a conference, the former colleague you have not spoken to in three years.⁶

What does not work: collecting connections like Panini stickers. A network of 4,000 contacts you have never spoken to is not a network, it is a list of what we Germans charmingly call Karteileichen (index-card corpses), and a corpse cannot refer you anywhere. The “I’d love to connect” request with no message does not work either, because it asks the recipient to do unpaid relationship labour for a stranger.

What does work: reaching the moderately weak ties on purpose, and doing it before you need anything. A specific, non-transactional message about something the person actually wrote or did. A real question inside your shared expertise. The informational conversation that is about their work, not your CV. Being findable, so that when a hiring manager searches, a trail of your actual competence exists. None of this is glamorous. All of it compounds. And critically, almost all of it must happen while you still have a job.

The DACH picture, where Vitamin B is real but mortal

In German professional life, the relevant phrase is Vitamin B, where the B stands for Beziehungen (connections). It is real. But two things deserve precision rather than folklore.

First, Vitamin B is declining as a share of decisive hires, from 37 per cent in 2022 to 27 per cent in 2025, as online and direct sourcing rise.^{8 14} The contact still helps enormously. It is simply no longer doing the job alone.

Second, the most reliable German pharma roles sit in regulated, technical functions (regulatory affairs, pharmacovigilance, QA, clinical, biostatistics) where a referral gets your file read but cannot fabricate the qualification. Vitamin B opens the door. Your GMP knowledge is what stops them closing it again.

A short international tour

UK and US: broadly the same machinery as DACH, with the referral economy a little more openly acknowledged. The weak-ties research is largely Anglo-American in origin.⁵ Be early, be known, be findable.

Rest of EU: the same logic, complicated by language and labour mobility. For a multilingual DACH professional, a moderately weak tie in Basel, Dublin, or Copenhagen is worth cultivating precisely because those markets sit just outside your immediate competition.

China: here the model genuinely differs. Guanxi (personal relationship networks built on reciprocal obligation) channels jobs primarily through strong ties, the opposite of Granovetter’s Western weak-ties result.¹¹ But the same research shows guanxi’s influence shrinks when large multinational firms impose standardised, open recruitment procedures.¹¹

India: a referral-heavy market, but with a sharp evidence-based warning. A controlled study in Kolkata found that people only referred genuinely high-skill candidates when given a performance incentive to do so.¹² A referral network is only as meritocratic as the incentives running through it.

What this means for you, depending on where you are standing

Juniors (internship, traineeship, first role, visa sponsorship): you have almost no strong ties yet, which is partly a hidden asset. The moderately weak tie is precisely the one a junior can build: the speaker after a webinar, the alumnus two years ahead. You are not asking for a job. You are asking an intelligent question and becoming a name worth remembering. For those needing sponsorship, be direct early: a warm referral can persuade a manager to take on the visa paperwork they would never absorb for a cold applicant.

Two to twenty years, up to management: your risk is a magnificent internal network and a dangerously thin external one. Fix this now, while employed, by reactivating five moderately weak ties a quarter with something specific and useful. Dull, undramatic, and the single highest-return thing in this article.

Executives and C-level: at your altitude the genuinely hidden market does exist, because senior and confidential searches really do run quietly through executive search firms and trusted referral. Your equivalent of networking is reputation and a small number of high-trust relationships maintained over years. You cannot start this when you need it.

The 50-plus reader: your network is your moat, and it is the one asset the market cannot accuse you of lacking. Decades of work mean decades of moderately weak ties, which the evidence says are the most useful kind. A referral is the most efficient way to get a human to look past a birth year.

The five measures that actually move the needle

1. Map and reactivate your moderately weak ties, deliberately and while employed. Not your best friends, not strangers. The acquaintances. Five a quarter, each with a specific, non-transactional reason to hear from you.^{5 6}

2. Become findable, so the search comes to you. A trail of real expertise makes you discoverable to the recruiters and AI sourcing tools that now scan digital footprints. From the market’s view, an unfindable professional is a hypothetical one.

3. Have the conversation about the work, never about the vacancy. The message that opens with “are you hiring” closes the door you were trying to prop open.

4. Treat the referral as a key, not a verdict. It gets your file read. It does not get you hired, and in our regulated corner of pharma it cannot paper over a missing qualification.

5. Build it before the announcement, not after. Every channel here compounds over time and collapses under panic. The professional a hiring manager calls is the one who existed in their memory before the role did.

A note for recruiters and hiring managers

Your referral channel is your highest-yield source by conversion, but its quality is a direct function of your incentives, exactly as the Kolkata result warns.¹² A scheme that rewards volume produces a flood of mediocre acquaintances; one that rewards a successful, retained hire produces the candidate an employee will stake their reputation on. Do not confuse the referral economy with a sourcing strategy you can rest on: the German trend is unambiguous, with digital and direct sourcing now carrying more decisive hires than personal contacts.⁸ And the cheapest win most pharma recruiters ignore: reward the second-degree introduction and you fish a larger pond than your competitors, who are all crowded around the same first-degree puddle.

Sources & References

Credible and primary (representative surveys, peer-reviewed research, foundational literature):

[1] Careery — “The Hidden Job Market Is Real, But Not How You Think” (traces the 70–80% claim to no credible source): careery.pro

[2] Learning Curves / Jesse Preston — “Hidden Origins of the Hidden Job Market” (Haldane 1966, Rochester NY pilot survey, Bolles, 1980 NYT interview): learningcurves.org

[3] “A Social Network Analysis of Occupational Segregation,” arXiv (labour-economics range: ~30–60% find jobs via contacts; 40–50% of employers use employee networks; Rees, Granovetter, Holzer, Montgomery, Topa): arxiv.org

[4] Granovetter’s “strength of weak ties” (1973), via the Stanford report on the 2022 replication: news.stanford.edu

[5] Rajkumar, Saint-Jacques, Bojinov, Brynjolfsson & Aral, “A causal test of the strength of weak ties,” Science (2022): science.org

[6] MIT Sloan press release on the same study (the “moderately weak ties” inverted-U result): mitsloan.mit.edu

[8] IAB-Stellenerhebung 1/2026 (2025 data: personal contacts used in 48%, decisive in 27%; online media now decisive in just over half of hires): iab-forum.de

[9] IAB-Stellenerhebung 1/2025 (2024 data: personal contacts used in 53%, decisive in 31%): iab-forum.de

[10] IAB-Forum on age and recruitment channels (2015 data, used with that caveat): iab-forum.de

[11] Huang, “Guanxi networks and job searches in China’s emerging labour market,” Work, Employment & Society (2008): journals.sagepub.com

[12] J-PAL — “Leveraging Social Networks for Job Referrals in India” (Kolkata study: high-skill referrals depend on incentives): povertyactionlab.org

[14] IAB-Stellenerhebung 1/2023 (2022 data: personal contacts decisive in 37% of hires): iab-forum.de

Vendor or self-interested (real pattern, treat exact figures as directional):

[7] Jobvite source-of-hire data via Zippia (referrals ~7% of applicants but ~40% of hires; ~28.5% success vs 2.7%). Jobvite sells referral software: zippia.com

[13] Example of the inflated coaching-industry version (“85% never posted”) with no primary sourcing, included to show the genre: openarc.net

© 23 June 2026 Andreas Schulz. All rights reserved.