A Smarter Job Search Strategy: Where Jobs Actually Come From

The short answer

Spend less of your week on job boards and more of it on people who are one step outside your close circle. Large-scale research on weak ties finds that acquaintances, not close friends, produce a disproportionate share of new jobs, because they know about openings your own network does not. Apply selectively, track what you send, and treat the board as one channel rather than the whole search.

A job search organizes itself around whatever is most visible, and what is most visible is the job board. It has a search box, a count of new postings, and a submit button. It feels like the search, so it becomes the search, and a bad week gets diagnosed as not enough applications.

The research on how people actually find work points somewhere less convenient.

Where do jobs actually come from?

Disproportionately from people you barely know. The sociologist Mark Granovetter argued this in 1973 in The Strength of Weak Ties: people found jobs more often through acquaintances than through close friends. The reason is structural rather than social. Your close friends move in the same circles you do and hear about the same openings. An acquaintance sits in a different pocket of the world and knows things you do not.

That was a small study for a long time, which is why the modern version matters. Researchers ran large-scale experiments across more than 20 million LinkedIn members over five years, published in Science, and found that weaker ties did causally increase job mobility. It is one of the rare pieces of career advice with an actual experiment behind it rather than a survey.

One refinement worth keeping, because it stops this becoming an argument for collecting strangers. The relationship is not simply weaker equals better. The study found an inverted U: moderately weak ties were the most productive, not the weakest possible ones. People who know you slightly can vouch for you. People who do not know you at all cannot.

What does that change about your week?

Mostly the allocation. If applications are the only thing on your list, you are working exclusively in the most crowded channel, competing against everyone else who found the same posting the same way.

• Reconnect rather than cold-network. The highest-yield contacts are people you already know slightly: former colleagues, people from a previous team, someone you worked with once two jobs ago. That is the moderately-weak band the research points at.
• Be specific in what you ask for. "Let me know if you hear of anything" is unactionable. "I am looking at data roles at mid-size healthcare companies, do you know anyone at X" gives someone something they can do in thirty seconds.
• Apply to fewer roles, properly. A tailored application to a role you match beats twenty generic ones, because relevance is what gets you retrieved from a recruiter's search.
• Keep a record of what you sent where and what came back. Without it you cannot tell which channel is producing replies, so you cannot reallocate anything.

Why does tracking matter more than it sounds?

Because a job search runs for months and human memory does not. Without a record you cannot answer basic questions: which version of the resume earned replies, whether referrals actually outperform applications for you, which companies went quiet after a first call. Those answers are the difference between adjusting your approach and repeating it.

This is the least glamorous part of a search and the one that most reliably gets skipped, because it produces no immediate feeling of progress. It is also the only way to know whether anything you are doing works.

What about the advice you should ignore?

Be skeptical of any confident number in this space. Job-search content is full of statistics with no source behind them: precise claims about how many applications produce an interview, what share of jobs are never advertised, how many resumes software rejects. Most trace back to a vendor's marketing rather than a study, and the numbers drift between retellings, which is the clearest sign there was never an original.

A reasonable filter: if a claim comes with a number but no link to a study, treat it as a story rather than evidence. That includes numbers on this blog. Ours link out.

Who this is not for

If you need work urgently, the slower relationship-building side of this is bad advice for your situation, and volume applying to roles you clearly match is the rational move. If you are in a field that hires almost entirely through formal processes, such as much of the public sector, the posting genuinely is the front door. And if you are early career with a thin network, you have fewer weak ties to activate, which makes the applications channel more important for you than this research implies.

Where our tools fit: Mbason AI tailors a resume against a specific posting, and the job application tracker keeps the record described above, with each application getting a forwarding email that updates its status. That covers the mechanical half. The conversations are yours.

How much of my job search should be networking versus applying?

There is no validated split, and anyone quoting one is guessing. A practical approach is to protect some time each week for reconnecting with people you already know slightly, rather than letting applications absorb everything. Then track which channel actually produces replies for you and shift toward it.

What do you say when reconnecting with someone you barely know?

Be direct and make it easy. Say what you are looking for specifically, acknowledge that you have not been in touch, and ask one concrete question they can answer quickly. Vague requests to keep an ear out put the work on them and usually produce nothing, however well intentioned.

Are referrals really better than applying directly?

A referral generally gets your application looked at by a person rather than found through a search, which is a meaningful advantage at the first stage. It does not carry you past an interview. Treat it as a way in rather than a shortcut to an offer.

How long should a job search take?

It varies enough by field, level, and market that any single figure is misleading. What is more useful is watching your own funnel: applications, replies, first calls, and later stages. If applications produce no replies, the problem is upstream and sending more will not fix it.