{"course":{"slug":"ai-careers","title":"Working in AI","url":"https://addaly.com/learn/ai-careers","level":"beginner","track":"work"},"lesson":{"slug":"writing-to-strangers","title":"Writing to strangers, and what to say","url":"https://addaly.com/learn/ai-careers/writing-to-strangers","position":36,"of":85,"minutes":8,"answer":"A short, specific message about their problem, sent to the person who does the work, outperforms fifty applications.","keyIdea":"A short, specific message about their problem, sent to the person who does the work, outperforms fifty applications."},"body":{"format":"markdown","text":"## Why this works when applying does not\n\nAn application joins a pile. A message arrives in an inbox. The person doing the job is rarely contacted, has the problem you would be solving, and, crucially, a hiring manager's internal recommendation moves through a company faster than any application ever will.\n\nThis is the highest-yield and worst-feeling activity in a job search. Most people do not do it, which is precisely why it works.\n\n## Who to write to\n\nThe person doing the work: a data engineer, an analyst, the engineering manager. Not the recruiter, who receives hundreds of these. Not the chief executive, unless the company has nine people.\n\nFind them by looking at who speaks about the company's technical work: conference talks, blog posts, GitHub contributions, LinkedIn posts, a podcast, an answer on a forum.\n\n## The four-sentence message\n\n```\nHi Ananya — I read your write-up on moving the reporting\nstack off nightly CSVs. I built something much smaller with\nthe same problem: a delay tracker for a transport\ncooperative, where three months of GPS timestamps were in\nlocal time and eight in UTC with nothing labelling which\n(https://…).\n\nOne question: when you backfilled, did you reprocess the\nold rows or keep both clocks and correct at read time? I\nwent with reprocessing and I am not sure it was right.\n\nNo ask — I am moving into data engineering and reading\npeople who have done it.\n```\n\n> Figure: One message that gets a reply, and six that do not. What the four sentences do: Prove you read their actual work, by naming it; Show your own work in one clause, with a link; Ask a question only somebody who has tried the problem could ask; Remove the pressure of a request. What kills a message: \"I hope this finds you well\"; A CV nobody asked for; \"Any opportunities in your team?\" as the first contact; Flattery with no specifics; Four hundred words about your journey; Asking for a call before giving anything. Specific messages to the person who does the work run somewhere between one in five and one in three. Templated ones run near zero, and everybody can tell the difference instantly.\n\nWhat that message does: proves you read her actual work, shows your own work in one clause, asks a question only somebody who has tried the problem could ask, and removes the pressure of a request.\n\nReply rates on messages like this run somewhere in the region of one in five to one in three. Templated messages run near zero, and everybody can tell the difference instantly.\n\n## What kills a message\n\n- \"I hope this email finds you well.\"\n- A CV attached that nobody asked for.\n- \"Any opportunities in your team?\" as the first contact.\n- Flattery with no specifics: \"I admire your work\" without naming any of it.\n- Four hundred words about your journey.\n- Asking for a call before you have given anything.\n\n## Volume\n\nFifteen to twenty-five well-targeted messages a month is the right scale. Not two hundred: at that volume you cannot be specific, and non-specific is the same as not sending. Two or three real conversations a month, sustained for six months, is a network.\n\n## The channels, and their real constraints\n\n**LinkedIn.** A connection request carries a note with a hard 300-character limit on the free tier, which is roughly the four sentences above compressed. That constraint is useful discipline. Premium raises message limits and does not raise reply rates; the message does that.\n\n**Email.** Company address formats are guessable and usually consistent: `first.last@`, `first@`, `finitial.last@`. Send one; a bounce costs nothing.\n\n**Communities.** Discord and Slack groups for your language, city or tool; local meetup groups; the forum where people actually answer questions. Here the sequence is reversed: be useful in public for three months first, then a message is from someone they recognise.\n\n**GitHub.** If you have fixed a bug or improved documentation in a project they maintain, you already have a conversation.\n\n## The follow-up rule\n\nOnce, after seven to ten days, adding something new: not \"just bumping this\". If there is still no reply, stop. Silence is usually workload, not judgement, and a third message converts neutral into negative.\n\n## The slow version that compounds\n\nAnswer questions in public for three months. In the community for your city or your tool, help people with the thing you learned last month. You are qualified to help anyone one step behind you, which is a much larger group than you think.\n\nWhat this produces is not followers. It is fifteen to thirty people who have watched you be useful, and who will forward you a job when one appears, which is what \"network\" actually means, stripped of its unpleasant connotations.\n\n## The free path\n\nLinkedIn free, community servers, and a guess at an email format cost nothing. Hunter and similar services have free tiers of a handful of lookups a month. The paid products (LinkedIn Premium, Sales Navigator, Apollo) are built for salespeople sending thousands of messages, which is the opposite of this method.\n\n## Where this does not work\n\nIn markets and employers where hiring is centralised and procedural (large IT services firms with campus intakes, public sector recruitment, graduate schemes) the process is the process, and a direct message will be politely redirected to the portal. Match the method to the employer category. Outreach is for startups, mid-size companies, agencies and remote roles; the application form is for institutions."},"check":{"question":"Meera sends 40 messages a week to people at companies she wants to join. Each opens by praising the company, summarises her background, and asks whether they have openings. She gets almost no replies. What is the correct diagnosis?","options":[{"text":"Her background is not strong enough yet, so she should build more projects before reaching out.","correct":false,"tempting":"Assumes the silence is about her qualifications, when the messages give the reader nothing specific enough to evaluate her by."},{"text":"At forty a week she cannot be specific, and a message that praises the company, recaps her CV and asks for openings gives the reader nothing to reply to; twenty specific messages a month about the recipient's own work would do better.","correct":true},{"text":"She is contacting the wrong channel and should apply through job boards instead.","correct":false,"tempting":"Sends her back to the method with the worst response rate, when the problem is the content of the messages rather than the approach."},{"text":"She should follow up more persistently, since most people miss the first message.","correct":false,"tempting":"Doubles the volume of a message that was not working, and converts non-response into active irritation."}]},"prev":{"slug":"the-profile-people-open","title":"The profile a stranger checks in ninety seconds","url":"https://addaly.com/learn/ai-careers/the-profile-people-open"},"next":{"slug":"writing-in-public","title":"Publishing, and the inbound it creates","url":"https://addaly.com/learn/ai-careers/writing-in-public"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'Writing to strangers, and what to say', in Working in AI. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/ai-careers/writing-to-strangers (Accessed: 13 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"Writing to strangers, and what to say","courseTitle":"Working in AI","module":{"index":4,"slug":"getting-seen","title":"Getting seen: the routes that put a person in front of your work"},"answer":null,"keyIdea":"A short, specific message about their problem, sent to the person who does the work, outperforms fifty applications.","hero":{"kind":"figure","figure":{"t":"figure","label":"Figure","figure":{"left":{"title":"What the four sentences do","items":["Prove you read their actual work, by naming it","Show your own work in one clause, with a link","Ask a question only somebody who has tried the problem could ask","Remove the pressure of a request"]},"type":"compare","right":{"title":"What kills a message","items":["“I hope this finds you well”","A CV nobody asked for","“Any opportunities in your team?” as the first contact","Flattery with no specifics","Four hundred words about your journey","Asking for a call before giving anything"]},"title":"One message that gets a reply, and six that do not","caption":"Specific messages to the person who does the work run somewhere between one in five and one in three. Templated ones run near zero, and everybody can tell the difference instantly."},"alt":"Figure: One message that gets a reply, and six that do not. What the four sentences do: Prove you read their actual work, by naming it; Show your own work in one clause, with a link; Ask a question only somebody who has tried the problem could ask; Remove the pressure of a request. What kills a message: “I hope this finds you well”; A CV nobody asked for; “Any opportunities in your team?” as the first contact; Flattery with no specifics; Four hundred words about your journey; Asking for a call before giving anything. Specific messages to the person who does the work run somewhere between one in five and one in three. Templated ones run near zero, and everybody can tell the difference instantly."}},"blocks":[{"t":"h2","id":"why-this-works-when-applying-does-not","runs":[{"text":"Why this works when applying does not"}]},{"t":"p","runs":[{"text":"An application joins a pile. A message arrives in an inbox. The person doing the job is rarely contacted, has the problem you would be solving, and, crucially, a hiring manager’s internal recommendation moves through a company faster than any application ever will."}]},{"t":"p","runs":[{"text":"This is the highest-yield and worst-feeling activity in a job search. Most people do not do it, which is precisely why it works."}]},{"t":"h2","id":"who-to-write-to","runs":[{"text":"Who to write to"}]},{"t":"p","runs":[{"text":"The person doing the work: a data engineer, an analyst, the engineering manager. Not the recruiter, who receives hundreds of these. Not the chief executive, unless the company has nine people."}]},{"t":"p","runs":[{"text":"Find them by looking at who speaks about the company’s technical work: conference talks, blog posts, GitHub contributions, LinkedIn posts, a podcast, an answer on a forum."}]},{"t":"h2","id":"the-four-sentence-message","runs":[{"text":"The four-sentence message"}]},{"t":"code","lang":"","label":"Text","text":"Hi Ananya — I read your write-up on moving the reporting\nstack off nightly CSVs. I built something much smaller with\nthe same problem: a delay tracker for a transport\ncooperative, where three months of GPS timestamps were in\nlocal time and eight in UTC with nothing labelling which\n(https://…).\n\nOne question: when you backfilled, did you reprocess the\nold rows or keep both clocks and correct at read time? I\nwent with reprocessing and I am not sure it was right.\n\nNo ask — I am moving into data engineering and reading\npeople who have done it."},{"t":"p","runs":[{"text":"What that message does: proves you read her actual work, shows your own work in one clause, asks a question only somebody who has tried the problem could ask, and removes the pressure of a request."}]},{"t":"p","runs":[{"text":"Reply rates on messages like this run somewhere in the region of one in five to one in three. Templated messages run near zero, and everybody can tell the difference instantly."}]},{"t":"h2","id":"what-kills-a-message","runs":[{"text":"What kills a message"}]},{"t":"ul","items":[[{"text":"“I hope this email finds you well.”"}],[{"text":"A CV attached that nobody asked for."}],[{"text":"“Any opportunities in your team?” as the first contact."}],[{"text":"Flattery with no specifics: “I admire your work” without naming any of it."}],[{"text":"Four hundred words about your journey."}],[{"text":"Asking for a call before you have given anything."}]]},{"t":"h2","id":"volume","runs":[{"text":"Volume"}]},{"t":"p","runs":[{"text":"Fifteen to twenty-five well-targeted messages a month is the right scale. Not two hundred: at that volume you cannot be specific, and non-specific is the same as not sending. Two or three real conversations a month, sustained for six months, is a network."}]},{"t":"h2","id":"the-channels-and-their-real-constraints","runs":[{"text":"The channels, and their real constraints"}]},{"t":"p","runs":[{"text":"LinkedIn.","b":true},{"text":" A connection request carries a note with a hard 300-character limit on the free tier, which is roughly the four sentences above compressed. That constraint is useful discipline. Premium raises message limits and does not raise reply rates; the message does that."}]},{"t":"p","runs":[{"text":"Email.","b":true},{"text":" Company address formats are guessable and usually consistent: "},{"text":"first.last@","code":true},{"text":", "},{"text":"first@","code":true},{"text":", "},{"text":"finitial.last@","code":true},{"text":". Send one; a bounce costs nothing."}]},{"t":"p","runs":[{"text":"Communities.","b":true},{"text":" Discord and Slack groups for your language, city or tool; local meetup groups; the forum where people actually answer questions. Here the sequence is reversed: be useful in public for three months first, then a message is from someone they recognise."}]},{"t":"p","runs":[{"text":"GitHub.","b":true},{"text":" If you have fixed a bug or improved documentation in a project they maintain, you already have a conversation."}]},{"t":"h2","id":"the-follow-up-rule","runs":[{"text":"The follow-up rule"}]},{"t":"p","runs":[{"text":"Once, after seven to ten days, adding something new: not “just bumping this”. If there is still no reply, stop. Silence is usually workload, not judgement, and a third message converts neutral into negative."}]},{"t":"h2","id":"the-slow-version-that-compounds","runs":[{"text":"The slow version that compounds"}]},{"t":"p","runs":[{"text":"Answer questions in public for three months. In the community for your city or your tool, help people with the thing you learned last month. You are qualified to help anyone one step behind you, which is a much larger group than you think."}]},{"t":"p","runs":[{"text":"What this produces is not followers. It is fifteen to thirty people who have watched you be useful, and who will forward you a job when one appears, which is what “network” actually means, stripped of its unpleasant connotations."}]},{"t":"h2","id":"the-free-path","runs":[{"text":"The free path"}],"role":"free"},{"t":"p","runs":[{"text":"LinkedIn free, community servers, and a guess at an email format cost nothing. Hunter and similar services have free tiers of a handful of lookups a month. The paid products (LinkedIn Premium, Sales Navigator, Apollo) are built for salespeople sending thousands of messages, which is the opposite of this method."}]},{"t":"h2","id":"where-this-does-not-work","runs":[{"text":"Where this does not work"}]},{"t":"p","runs":[{"text":"In markets and employers where hiring is centralised and procedural (large IT services firms with campus intakes, public sector recruitment, graduate schemes) the process is the process, and a direct message will be politely redirected to the portal. Match the method to the employer category. Outreach is for startups, mid-size companies, agencies and remote roles; the application form is for institutions."}]}],"check":{"question":[{"text":"Meera sends 40 messages a week to people at companies she wants to join. Each opens by praising the company, summarises her background, and asks whether they have openings. She gets almost no replies. What is the correct diagnosis?"}],"options":[{"text":[{"text":"Her background is not strong enough yet, so she should build more projects before reaching out."}],"correct":false,"tempting":[{"text":"Assumes the silence is about her qualifications, when the messages give the reader nothing specific enough to evaluate her by."}]},{"text":[{"text":"At forty a week she cannot be specific, and a message that praises the company, recaps her CV and asks for openings gives the reader nothing to reply to; twenty specific messages a month about the recipient’s own work would do better."}],"correct":true,"tempting":null},{"text":[{"text":"She is contacting the wrong channel and should apply through job boards instead."}],"correct":false,"tempting":[{"text":"Sends her back to the method with the worst response rate, when the problem is the content of the messages rather than the approach."}]},{"text":[{"text":"She should follow up more persistently, since most people miss the first message."}],"correct":false,"tempting":[{"text":"Doubles the volume of a message that was not working, and converts non-response into active irritation."}]}],"order":[2,0,3,1],"why":null},"prev":{"slug":"the-profile-people-open","title":"The profile a stranger checks in ninety seconds","minutes":8},"next":{"slug":"writing-in-public","title":"Publishing, and the inbound it creates","minutes":9}}}