How to Network in Tech Without Cringing

Most advice about networking in tech assumes you are the kind of person who enjoys working a room, and then quietly blames you when you do not. The result is a lot of people who know they should be doing something, feel a low hum of dread about it, and end up refreshing job boards instead. That is a shame, because the version of networking that actually works in this industry is much closer to careful, unglamorous administration than it is to charisma.

This walkthrough treats networking as a repeatable practice rather than a personality trait. Nine steps take you from the people you already know, through the ask that earns a reply, the cold message, the informational conversation, and public contribution as the version that scales for quiet people, to the referral request and the reciprocity that keeps a network alive after you stop needing it. It sits alongside our walkthroughs on what to do after a tech layoff, which covers the money and paperwork around a search, and writing a tech resume, which covers the document a referral eventually points at. Open the companion and enter your own numbers once as you read.

Why networking in tech feels cringe, and what it actually is

The cringe is real, and it has a cause. Most people first meet the word “networking” attached to an activity that is genuinely unpleasant: standing in a room holding a drink, trying to convert small talk into an advantage, aware that everyone present knows exactly what everyone else is there for. That activity is transactional by design, it rewards a specific extroverted skill set, and it makes reasonable people feel like they are performing a role written for someone else.

What the word actually describes, in a technical career, is much duller. It is the ongoing state of a modest number of people knowing what you are good at and being willing to say so. That state gets built through work, not charm. It gets built by having been useful to a colleague two jobs ago, by having answered a question carefully in a public forum, by having sent a pull request that fixed a real problem, and by having stayed in occasional contact with people you liked.

The reason it matters in this industry more than in some others is structural. Technical roles are hard to assess from a resume, so hiring is expensive and risky, and anything that reduces the risk gets weighted heavily. A person inside the company saying “I have worked with them, they are good” is a genuinely different signal from an application. That is not a hidden trick, it is just how organisations behave when assessment is difficult.

So the reframe that makes this bearable is: you are not selling yourself. You are making it easy for a small number of people to be accurate about you. The steps below are what that looks like in practice, with the drink and the room made optional.

Before you start: what to line up

None of the steps below need special tools, and the honest prerequisite list is short. Most of the difficulty is emotional rather than logistical, so removing the logistical friction first is worth an hour of setup.

• A current one-line description of what you do and what you want next, in plain language, short enough to say out loud without pausing. This is the sentence that goes in every message you send.
• A list of somewhere between fifteen and fifty people who have seen your work directly, assembled from memory, old messaging apps, commit histories, and your own contacts. Do not filter it yet. Building it is Step 1.
• Somewhere to write things down, which can be a spreadsheet or a plain text file. It needs four columns: who, when you last spoke, what you asked, and what you owe them. Nothing more elaborate is required.
• Something to point at, meaning a resume that is current and, ideally, one piece of visible work. Our note on [building a tech portfolio](/articles/how-to-build-a-tech-portfolio/) covers what counts as pointable.
• A realistic weekly time budget, which for most people is two to four hours. Below that the practice stalls, and above it the practice becomes a second job that crowds out the work that makes you worth referring.

Difficulty is low and the whole setup runs to roughly an hour. Time cost across a full twelve-week stretch at three hours a week is around 36 hours, which is the figure the calculator uses as its default. Nothing here needs a subscription, a course, or a personal brand.

The one genuine prerequisite is a decision about honesty. Everything in this walkthrough works because the things you say are true: you did work with that person, you did read the thing you are referencing, you do want the role you say you want. The techniques that feel slimy are the ones that require you to fake one of those, and they are also the ones that fail, because the people you are contacting have seen the faked version many times.

Step 1: Map the people who have already seen you work

Start with the list, not with a message. The most common failure in networking is beginning at the hardest possible entry point, which is a stranger with a public profile and a full inbox, when there is a much warmer entry point sitting unused in your own history.

Open your messaging apps, your email archive, your commit history, and whatever professional network you already have, and write down everyone who has direct evidence of what you are like to work with. Former teammates. Managers you liked, and managers who liked you, which is not always the same set. People on adjacent teams whose projects touched yours. Classmates or bootcamp cohort members who are now working. Maintainers whose projects you have contributed to. Anyone you helped in a way they would remember, and anyone who helped you.

Do not filter for seniority, and do not filter for whether they are at a company you want. Both filters are wrong at this stage. Seniority matters less than knowledge of your work, and people move, so the person at an irrelevant company today is the person at a relevant one next year. A list of 25 names is plenty. An illustrative list of that size, worked at five outreach messages a week, gives you five weeks of warm contacts before you have to go anywhere near a stranger.

Then sort the list into three buckets: people who would recognise your name instantly, people who would need one sentence of reminder, and people you knew of but never worked with directly. The first two buckets are where almost all of the value is. Work them in that order.

Watch out for the instinct to skip this because it feels too easy or too small. It is genuinely the highest-yield step, and skipping it is how people end up sending forty cold messages while a former manager who would have made a call sits unasked in an old chat thread.

Step 2: Decide what you are actually asking for

Before any message goes out, write down the thing you want in one sentence that a stranger could act on. This step takes twenty minutes and determines the reply rate of everything that follows, which is why it sits before the outreach rather than inside it.

The test is simple. Read your sentence and ask whether a busy person could do something about it in under five minutes. “I am looking for opportunities” fails, because there is nothing to do with it. “I am looking to move from backend work into platform or infrastructure engineering at a mid-size company, ideally somewhere that runs its own Kubernetes rather than buying it” passes, because it gives the reader a pattern to match against everything they already know.

Specificity feels risky, and that is the reason people avoid it. Naming a target seems to close doors. In practice the opposite happens: a broad ask is filed under “keep an eye out”, which is a mental category that never gets revisited, while a narrow ask is remembered because it attaches to something concrete. You can always widen later, and nobody is holding you to the sentence.

Write three versions: the role or problem you want, the kind of company or team, and the constraint that actually matters to you, such as location, remote work, domain, or scale. Then compress them into one sentence. If you cannot compress it, the problem is usually that you have not decided yet, and the honest fix is to decide rather than to send a vague message and hope the recipient decides for you.

Watch out for the trap of asking for a job. Almost nobody can give you a job in a message. Ask for the thing they can give: a name, a view, an honest answer about a team, twenty minutes. Jobs come out of those, not out of being asked for directly.

Step 3: Reopen a dormant contact without the awkward preamble

The warm list is only warm if you actually message it, and the block for most people is the years of silence. The fix is to stop treating the silence as a debt you have to settle before you can speak.

The structure that works is three short paragraphs. First, the reminder: where you worked together and what you worked on, in one sentence, given even to people you are fairly sure will remember, because it costs nothing and removes any chance of them straining. Second, the update: what you are doing now and the one-sentence ask from Step 2. Third, the exit: an explicit, genuine statement that no reply is needed if the timing is bad.

What to cut is the apology. A long opening about how you feel bad for only reaching out when you need something forces the reader to manage your feelings before they get to the point, and it plants the idea that this contact is an imposition. Almost nobody keeps a ledger of who contacted whom last. People who liked working with you are generally glad to hear from you and mildly flattered to be asked.

Also cut the false warm-up, meaning the two weeks of liking their posts before messaging. It is visible, it delays the thing you actually want, and it makes the eventual message read as a manoeuvre. Send the message.

Watch out for volume here. Five to eight of these a week is a sustainable rate that leaves you able to reply properly when several land at once. Twenty in an evening produces a burst of replies you handle badly, which is worse than sending fewer.

The two asks: why one gets a reply and the other does not

It is worth seeing the difference laid out, because almost every disappointing outreach result traces back to this and almost nobody notices it in their own drafts.

The vague ask sounds generous: “Let me know if you hear of anything.” It asks the reader to hold an open-ended monitoring task with no criteria, no end date, and no way to know whether they have succeeded. Everyone agrees to it and nobody does it, because there is nothing to do. The reader is not being dishonest; they simply have not been given a task.

The specific ask is smaller and lands better: “If you hear of a platform or infrastructure role at a company under about 500 people in this region, I would like to hear about it.” Now there is a filter running in someone’s head. When a colleague mentions a team is hiring, the pattern fires. That is the entire mechanism, and it is why specificity beats enthusiasm.

The same applies to introductions. “Could you introduce me to anyone useful” is unanswerable. “You mentioned you know someone on the data platform team there. Would you be willing to introduce me, or tell me whether it is worth approaching them directly” gives the reader a decision with two clear options and an easy no.

There is a third ask worth naming, which is the information ask, and it has the highest acceptance rate of the three because it costs the least. “Would you be willing to tell me what the team is actually like” asks for an opinion, which people generally enjoy giving. Start there with anyone who does not know your work, and let the larger asks become available later.

Step 4: Write a cold message short enough to answer on a phone

Cold outreach is the part everyone dreads and it is genuinely the lowest-yield channel, which is why it sits at Step 4 rather than Step 1. It still belongs in the practice, because sometimes the person you need to speak to is a stranger.

The structure is four sentences. One: who you are, compressed to a single clause. Two: why this person specifically, referencing something they actually did, which means you have read it, not skimmed a headline. Three: the ask, phrased as the small information ask rather than the large one. Four: an explicit permission to decline, written so it does not read as a guilt device.

That is the whole message. Do not attach a resume unasked, do not include three paragraphs of background, and do not open with flattery that could have been written before you knew who they were. The reader is deciding, in about four seconds, whether this is a real message from a real person or one of the dozens of templated approaches they receive. Every generic sentence pushes it toward the second category.

The specificity in sentence two is what does the work. “I read your write-up on migrating the ingest pipeline and had a question about how you handled backfills” cannot be sent to anyone else, which is exactly why it gets read. Any sentence you could paste into a hundred messages is a sentence that signals you did.

Watch out for the follow-up reflex on cold messages. One follow-up after a week or two is fine. After that, stop. On an illustrative plan of five messages a week over twelve weeks, that is around 60 messages, and if you are working the warm list first, only a minority of those should ever be cold.

Step 5: Run the informational conversation like a reporter

When someone says yes to twenty minutes, the failure mode is turning up with nothing prepared and letting the conversation drift into a bad interview where you talk about yourself. Run it the other way. You are there to learn, and the person is the source.

Prepare six questions and expect to use four. Good ones cover what the work actually consists of in a normal week, what surprised them when they joined, what they would want a person joining the team to already be good at, what they would learn if they were starting from where you are, and what they think the team will be dealing with in a year. These are questions about their experience, which people answer readily and enjoy answering.

Keep to the time you asked for. If you asked for twenty minutes, start wrapping up at eighteen. Ending early and on time is a small signal of reliability that costs nothing and is remembered more than you would expect. If they want to continue, they will say so.

Close with the question that compounds: “Is there anyone else you think it would be worth me speaking to?” This is how one conversation becomes two, and it is easy to answer because it asks for a name rather than a favour. Then send a short thank you within a day that mentions one specific thing you took from the conversation, which proves you were listening and gives them something pleasant to remember.

Watch out for the covert-job-application version of this. Asking for twenty minutes to learn and then spending it pitching yourself is the move that gives informational conversations a bad name. If you want to be considered for a role, say so directly rather than smuggling it in.

Step 6: Contribute in public instead of collecting contacts

This is the step that changes the whole practice for people who find the rest of it draining, and it is the one most often skipped because its payoff is slow and indirect.

Contributing in public means producing small, useful, findable artefacts: a write-up of a problem you solved and how, a careful answer to a question in a forum where practitioners actually gather, documentation for a tool you use that was missing a page, a small bug fix or test in an open-source project you depend on, a talk at a local meetup that is really just a write-up read aloud. None of these require you to meet anyone.

The mechanism is inversion. Instead of you finding people, people with the same problem find you, which means the conversation starts with shared context rather than with an introduction. It also produces exactly the artefact a referrer needs, because “here is a thing they wrote about the problem you have” is far more persuasive to forward than “here is a person I met”.

Start absurdly small. One post about something you fixed last month. One answer a week in one community. One documentation fix in a repository you already use. The compounding is the point: work published a year ago keeps working while you are asleep, which no amount of attending events can do.

Watch out for the content-treadmill version, where the goal quietly becomes an audience rather than usefulness. Posting daily to build a following is a different activity with different costs, and it is not required here. Three genuinely useful posts a year will do more for your career than three hundred forgettable ones.

Where the weekly hour actually goes

A practice needs a shape, or it collapses into whichever part is least uncomfortable. The split below is an illustrative planning allocation rather than a measured finding, but it encodes the priority order the steps above argue for: warm contacts first, public work second, cold outreach third, events last.

Read it as a diagnostic rather than a prescription. If your actual week is 90 percent cold outreach and nothing else, that explains a practice that feels humiliating and produces little. If it is 100 percent events, you are spending your energy on the highest-cost channel and getting the shallowest contacts from it.

The allocation also shifts with circumstance. In the first fortnight of a search, warm contacts should take almost everything, because that is where the fastest replies live. Once the warm list has been worked, public contribution and cold outreach take over. If you are not searching at all, the whole budget can drop to an hour a week spent almost entirely on the first two slices.

Step 7: Work a meetup or conference without the small talk

Events are the part of networking most people mean when they say they hate networking, so the goal here is to make them survivable and optional rather than central. If you skip this step entirely and do more of Step 6 instead, the practice still works.

If you do go, go with a job. A person with a task has something to do with their hands and their attention, and the dread mostly comes from having neither. Useful jobs include: ask one specific question after a talk, find the two people whose work you already follow and thank them for a specific thing, volunteer to help run the event, or offer to give a five minute lightning talk. Volunteering is the strongest of these, because a role gives you a reason to speak to everyone and a graceful exit from every conversation.

Choose the format deliberately. A small local meetup on a narrow topic is usually a better trade than a large conference, because everyone present shares context, conversations start at the technical level rather than at the introductions level, and the same faces return monthly, which is what turns acquaintances into contacts.

Set a leaving time before you arrive and keep it. Two good conversations and an early exit beats four hours of endurance followed by a week of recovery. Nobody is counting your hours in the room.

Watch out for the business-card reflex, meaning collecting contacts you will never speak to again. Three names you follow up with properly within two days are worth more than thirty you do not. The follow-up is the event; the room is just where it started.

Step 8: Keep a contact record and follow up on a schedule

The reason networks decay is not malice or laziness, it is that human memory is bad at this and there is no natural prompt. A four-column record fixes it, and it takes about a minute per entry.

The columns are: who they are and how you know them, the date you last spoke, what you asked or what they offered, and what you owe them. That last column is the one people leave out and the one that does most of the work, because it converts a vague sense of obligation into a specific action you can complete.

Follow-up then becomes scheduling rather than courage. A reasonable default rhythm is one follow-up on an unanswered message after a week or two, then stopping; a thank-you within a day of any conversation; a short update to anyone who helped once the outcome is known, whatever the outcome was; and a light check-in with your closest contacts once or twice a year with something genuinely useful attached rather than a bare hello.

The stopping rule matters as much as the sending rule. Two messages read as interest and a third reads as pressure, so stop at two and move the name to the back of the record rather than to a permanent exclusion list. Silence is usually about their week, not about you.

Watch out for building an elaborate system instead of doing the practice. A spreadsheet is enough. An afternoon spent configuring a personal relationship database is an afternoon of avoidance dressed up as preparation, and it is one of the most common ways this step gets skipped while feeling completed.

Step 9: Ask for the referral only when the ask is fair

A referral is not a favour like forwarding a link. It is a person putting their own judgment on the line inside their own company, and understanding that is what makes the ask comfortable to send and easy to say yes to.

The fairness test has three parts. Can this person honestly speak to your work, from direct experience or from something substantial they have seen you produce? Is there a specific role, with a link, that you would genuinely accept? Have you given them what they need, meaning the role, your resume, and two or three sentences they can paste that connect what they saw you do to what the role needs? If all three are yes, the ask is fair and you should make it plainly.

Phrase it with an exit. “If you feel comfortable referring me for this role, it would help a lot. If not, no problem at all, and I would still value your read on whether it is worth applying” gives a genuine out. People who decline that gracefully often help in a smaller way instead, and the relationship survives, which matters more than any single role.

If the fairness test fails, ask for the smaller thing. From someone who does not know your work, the honest ask is a conversation or an opinion about the team, not a vouch they cannot support. Our note on preparing for a technical interview covers what happens after a referral lands you in a process, which is where the referral stops helping and your own preparation takes over.

Watch out for asking too early in a new relationship. A referral request in the first message to someone you just met is the move that makes people stop replying, because it asks for the largest possible thing at the moment of least possible evidence.

How outreach volume turns into conversations

Volume is not the point of this walkthrough, but arithmetic helps set expectations, because most people quit at the point where the numbers say almost nothing should have happened yet.

Two things are worth reading out of that chart. The first is that the middle row, five messages a week, is the sustainable one for most people alongside a job, and it still produces roughly 20 conversations over a quarter, which is more than enough to change what you know about your own market. The second is that the top row is not four times better than the second row in any way that matters, because reply quality falls as personalisation falls, and 48 rushed conversations is a worse quarter than 20 careful ones.

Note also what the arithmetic deliberately leaves out. It says nothing about outcomes, because the number of conversations that turn into anything depends on the roles that happen to exist, your fit for them, and timing you do not control. Treat it as a way to set expectations for effort, not as a forecast of results. Run your own version in the calculator.

Reciprocity: what keeps a network alive after the search

A network built entirely during a job search dies the week the search ends, and then has to be rebuilt from cold the next time. That rebuild is the actual cost of treating networking as an emergency activity, and it is why the reciprocity step is not a moral flourish but the mechanism that makes the whole thing cheap the second time.

The maintenance is small. Reply to people who contact you the way you wanted to be replied to, which mostly means promptly and honestly, including a quick no when you cannot help. Refer people when you genuinely can, applying the same fairness test you would want applied to you. Tell people what happened after they helped, because almost nobody closes that loop and the ones who do are remembered. Send the occasional useful thing to someone who would find it useful, with no ask attached.

There is also the specific obligation that comes with being helped. If someone gave you twenty minutes when you had nothing to offer, the repayment is not to that person, it is to the next person in your position who contacts you. That is the whole social contract holding this together, and it is why the industry norm of answering a polite stranger persists at all.

Practically, this is fifteen minutes a fortnight once you are settled. The record from Step 8 tells you who you owe something to. Work the “what you owe them” column down when you have capacity, and the network stays warm at a cost that is close to trivial compared to rebuilding it from zero.

Networking as an introvert, without acting like someone else

The advice to “just put yourself out there” fails a lot of people because it prescribes a delivery mechanism rather than the underlying goal. The goal is that some people know your work. Large rooms are one delivery mechanism and, for many quiet people, one of the worst available.

The channels that suit a lower social battery are the asynchronous ones, and they are not a consolation prize. Writing lets you be as thoughtful as you actually are rather than as quick as a room demands. One-to-one conversations, whether by video or over coffee, produce far more depth per unit of energy than a crowded event. Structured roles at events, such as volunteering or speaking, remove the unstructured mingling that is the actual difficulty. Small communities where the same people return each month build familiarity slowly and without performance.

There is also a real advantage worth naming. People who find this hard tend to prepare, and preparation is exactly what makes an informational conversation good. The person who arrives with six thought-out questions is remembered more warmly than the person who worked the whole room, because they were more interesting to talk to.

The one thing that does not work is waiting until you feel confident. Confidence follows evidence, and the evidence comes from a handful of conversations that went fine. Start with the safest name on your warm list, send one message, and let the practice supply the confidence rather than requiring it up front.

What to do when nobody replies

Silence is the most common experience in the first fortnight, and it is also the point at which most people conclude the whole thing does not work for them. Before concluding that, check four things in order, because there is usually a fixable cause.

First, check the ask. Reread your last five messages and ask whether a busy person could act on any of them in five minutes. If the answer is no, the problem is Step 2 rather than the recipients. Second, check the warmth. If everything you have sent is cold, low replies are expected rather than a signal about you, and the fix is to go back and work the warm list you skipped.

Third, check the length and the shape. If the message is more than about six sentences, or the ask arrives in the last line, many readers will not reach it. Fourth, check timing and volume. Two messages to the same person is the ceiling, but a quarter is not a long time to sample, and a fortnight of silence is genuinely not evidence of anything.

If all four check out and it is still quiet, the honest answer is that reply rates are unpredictable and partly out of your hands. Shift weight toward public contribution, which does not depend on anyone answering you, and keep the outreach running at a lower, sustainable rate. That is also the sensible response to a slow market generally, since published work keeps accumulating value whether or not anyone is hiring this month.

Networking inside the company you already work at

Almost all networking advice assumes you are looking to leave, which quietly wastes the highest-yield network available to most people: the one inside the building where your work is already visible.

Internal networking is mostly a matter of being known for something specific by people outside your immediate team. That comes from the same activities scaled down: writing internal documentation others reference, answering questions in company channels where other teams can see it, presenting a short summary of something your team learned, and volunteering for the cross-team task nobody wants that puts you in front of three groups you would otherwise never work with.

The payoff shows up in three places. Internal moves are often faster and lower risk than external ones. Internal advocates are the people who mention your name in rooms you are not in, which is a mechanism our note on getting promoted in tech covers in more detail. And when people leave, which in this industry they do constantly, an internal colleague becomes an external contact at a new company at no cost to you.

That last point is worth sitting with. Your future external network is mostly made of your current internal colleagues. Treating them well, doing what you said you would do, and being someone people are glad to have worked with is not a networking tactic. It is the entire foundation, and no amount of message-writing substitutes for it.

A worked example: twelve weeks of five messages a week

Take an illustrative case. A backend engineer with about six years of experience wants to move into platform work. She has a demanding job, so she allocates three hours a week for twelve weeks, which is about 36 hours in total. Her warm list, assembled in an evening, comes to 25 names.

She writes her sentence: platform or infrastructure engineering, at a company under about 500 people, in her region or fully remote. Then she works the list at five messages a week. At 25 warm names and five a week, the warm list lasts five weeks, and the remaining seven weeks mix cold outreach with public work. Over twelve weeks that is 60 messages sent.

On her own planning assumption of roughly one reply in three, that is about 20 conversations across the quarter. Against 36 hours invested, that is about 1.8 hours per conversation, counting the research, the writing, the call, and the follow-up. Her hour budget splits to roughly 12.6 hours on people she already knows, 10.8 hours on public work, 7.2 hours on cold outreach, and 5.4 hours on two meetups.

The public work slice becomes three write-ups: one on a migration she ran, one on a debugging technique, one answering a question that kept recurring in a community she reads. Two of her twenty conversations start because someone found one of those posts, which is the inversion from Step 6 doing its job.

By week twelve she has around 20 conversations, two of which came from a single former colleague introducing her onward, a much clearer picture of which local companies actually run the infrastructure she wants to work on, and one referral from a former manager who could honestly speak to her work. Whether that referral becomes an offer depends on the interview process and on timing, neither of which the practice controls. What the practice did control is that she is now known, specifically, by a handful of people in the area she wants to work in. Enter your own numbers in the companion to see your version of the same arithmetic.

Common mistakes that make networking feel like begging

• Asking for a job instead of for something the person can give. Almost nobody in a message can hand you a job. They can give a name, a view, twenty minutes, or an honest answer, and every one of those is a step toward the job.
• Sending the vague ask. "Let me know if you hear of anything" is agreed to by everyone and acted on by nobody, because it contains no filter a person can run.
• Starting with strangers while the warm list sits untouched. Cold outreach is the hardest channel and the lowest-yield one. Working it first is how people conclude that networking does not work.
• Opening with an apology for the silence. It makes the reader manage your feelings before they reach the point, and it frames your message as an imposition when it almost certainly is not.
• Treating a conversation as a covert application. Asking for twenty minutes to learn and then pitching yourself for the whole twenty is the behaviour that makes people stop granting them.
• Disappearing the moment the search ends. A network worked only in emergencies has to be rebuilt from cold every time, which is far more expensive than fifteen minutes a fortnight of maintenance.
• Confusing an audience with a network. A large following of people who have never seen your work cannot refer you. A handful of people who have worked with you can.

Troubleshooting: career switchers, gaps, and no former colleagues

What if you are switching in from another field and have no technical colleagues at all? Then Step 1 has nothing to work with and Step 6 has to carry more of the weight. Build visible work first, however small, then use communities around the specific technology you are learning as your entry point. The people there share a context with you already, which is a warmer start than a cold message ever is. Our walkthrough on switching careers into tech covers what makes that visible work credible.

What if you have a long employment gap? Say it plainly and briefly, then move on. Gaps are common, and the discomfort you carry about it is usually more visible than the gap itself. In a networking conversation nobody is screening you, so a one-clause mention is enough.

What if the person who could refer you has not seen your work in years? Then the fair ask is smaller than a referral. Ask for the conversation first, give them something current to look at, and let a referral become possible if they end up comfortable with it. Forcing the larger ask is what damages the relationship.

What if you are employed and worried about being seen looking? Keep the outreach to individual private messages, be candid with people you trust about confidentiality, and shift the weight of your practice toward public technical work, which is entirely defensible on its own terms and does not signal anything about your intentions.

What if you tried all of this for a quarter and nothing came of it? That is a real outcome and it is usually about the market or the specificity of the target rather than about you. Widen the target sentence by one dimension, keep the public work running, and drop the outreach to a maintenance rate rather than stopping it. Our note on getting an entry-level IT job covers adjacent routes worth considering when a narrow target is not moving.

The low-cringe networking checklist

• Write one sentence describing what you do and what you want next, specific enough that a busy person could act on it in five minutes.
• Build the list of everyone who has seen your work directly, without filtering for seniority or company.
• Sort it into instant-recognition, needs-one-reminder, and never-worked-together, and work it in that order.
• Send five messages a week, three short paragraphs each, with no apology for the silence and an explicit permission to decline.
• For strangers, keep it to four sentences with one line that could only have been written to them.
• Prepare six questions before any conversation, keep to the time you asked for, and close by asking who else is worth speaking to.
• Publish something small and useful on a regular rhythm, even if it is one post a quarter.
• Follow up once after a week or two, then stop, and thank people within a day of any conversation.
• Keep four columns on every contact: who, last spoken, what you asked, what you owe them.
• Apply the fairness test before any referral request: honest knowledge of your work, a specific role, and everything they need to write it.
• Close the loop with everyone who helped once you know the outcome, whatever the outcome was.
• Keep fifteen minutes a fortnight of maintenance running after the search ends.

The bottom line

Networking in tech goes wrong when it is treated as a personality test rather than as a process, because the personality-test version has no steps and therefore no way to start. The process version has nine, and the first several are administrative: write a sentence, build a list, send a short message, prepare some questions. None of that requires charisma, and all of it is more effective than the version with the drink and the crowded room.

The two ideas worth keeping if you keep nothing else are these. Specificity is what converts goodwill into action, because a vague ask gives a willing person nothing to do. And reciprocity is what makes the whole thing cheap the second time, because a network maintained at fifteen minutes a fortnight never has to be rebuilt from cold. On the walkthrough default of five messages a week for twelve weeks at three hours a week, you are looking at roughly 60 messages, about 20 conversations, and around 1.8 hours per conversation. Run your own figures in the companion, then send the first message to the safest name on your list.

CredYard publishes this walkthrough as an explanation of how professional relationships tend to work in technology, and it is educational material rather than career, employment, or recruitment advice tailored to you. Every message count, reply assumption, hour split, and conversation figure here is arithmetic chosen to illustrate a planning method, not a measured statistic about hiring, referrals, or response rates, and no part of it should be read as a prediction of what your own outreach will produce. Hiring practices, referral policies, and the norms of individual communities vary widely between companies, regions, and specialisms, so check the rules that apply where you are before acting on anything described here. The examples are hypothetical illustrations rather than accounts of real people, and any decision about your own career is worth discussing with people who know your specific situation.