[{"data":1,"prerenderedAt":1149},["ShallowReactive",2],{"blog:\u002Fblog\u002Fcontent-calendar-you-can-keep":3,"blog:related:\u002Fblog\u002Fcontent-calendar-you-can-keep":390},{"id":4,"title":5,"author":6,"body":7,"date":376,"description":377,"draft":378,"extension":379,"image":380,"meta":381,"navigation":382,"path":383,"seo":384,"stem":385,"tags":386,"updated":380,"__hash__":389},"blog\u002Fblog\u002Fcontent-calendar-you-can-keep.md","A 4-week content calendar you can actually keep","SelfSM",{"type":8,"value":9,"toc":363},"minimark",[10,19,24,31,34,38,41,80,83,87,90,209,212,216,219,256,259,263,266,270,273,277,280,284,287,319,323,329,339,350,360],[11,12,13,14,18],"p",{},"Most content calendars collapse for one reason: \"post every day\" is a schedule, not a plan. Every morning becomes a decision, and decisions are where consistency goes to die. This is a calendar built to survive that. It decides the ",[15,16,17],"em",{},"kind"," of post ahead of time so that filling any given day is a small task instead of a blank page. You'll get the five buckets it runs on, a filled four-week example you can copy, notes for adapting it per platform, and a template to reuse each month.",[20,21,23],"h2",{"id":22},"what-a-content-calendar-actually-is","What a content calendar actually is",[11,25,26,27,30],{},"A content calendar is a plan that assigns each posting slot a ",[15,28,29],{},"job"," before you write a word. Not the exact caption, and not a rigid script. Just the type of post that goes there: a teaching post, a proof post, a personal one. The value isn't the grid. It's that you never sit down to \"do social\" and instead sit down to \"write today's proof post,\" which is a task small enough to actually finish.",[11,32,33],{},"A good calendar does three things: it stops you repeating the same note two days running, it removes the daily \"what do I post\" decision, and it leaves room to react when something real happens.",[20,35,37],{"id":36},"the-five-buckets","The five buckets",[11,39,40],{},"Rotate every post through five recurring types. Each does a different job for your audience.",[42,43,44,52,58,64,70],"ol",{},[45,46,47,51],"li",{},[48,49,50],"strong",{},"Expertise & frameworks."," Teach one thing you know well. Example: \"The three questions I ask before taking on any client.\" These build authority.",[45,53,54,57],{},[48,55,56],{},"Industry take."," React to something happening in your field. Example: a short, opinionated post on a competitor's pricing change. Timely, and it shows you're paying attention.",[45,59,60,63],{},[48,61,62],{},"Proof."," A result, a case, a before-and-after. Example: \"How we took a client from 200 to 2,000 followers in eight weeks.\" This is what convinces buyers.",[45,65,66,69],{},[48,67,68],{},"Build in public."," What you're working on, what broke, what you learned this week. Example: \"We changed our onboarding and support tickets dropped by half.\" Honest and relatable.",[45,71,72,75,76,79],{},[48,73,74],{},"Human."," A moment, a story, something that isn't strictly business. Example: \"What running taught me about shipping work.\" This is what makes people remember ",[15,77,78],{},"you",", not just your feed.",[11,81,82],{},"Across four weeks you touch all five repeatedly without ever running two of the same in a row.",[20,84,86],{"id":85},"a-filled-example-four-weeks","A filled example: four weeks",[11,88,89],{},"Here's the rotation as a real grid, Monday to Friday. The bucket is the job; the note is a sample angle you'd swap for your own.",[91,92,93,117],"table",{},[94,95,96],"thead",{},[97,98,99,102,105,108,111,114],"tr",{},[100,101],"th",{},[100,103,104],{},"Mon",[100,106,107],{},"Tue",[100,109,110],{},"Wed",[100,112,113],{},"Thu",[100,115,116],{},"Fri",[118,119,120,143,165,187],"tbody",{},[97,121,122,128,131,134,137,140],{},[123,124,125],"td",{},[48,126,127],{},"Week 1",[123,129,130],{},"Expertise: a framework you use",[123,132,133],{},"Industry: react to news",[123,135,136],{},"Proof: a recent result",[123,138,139],{},"Build: what you shipped",[123,141,142],{},"Human: a weekend story",[97,144,145,150,153,156,159,162],{},[123,146,147],{},[48,148,149],{},"Week 2",[123,151,152],{},"Proof: a client before\u002Fafter",[123,154,155],{},"Expertise: a common mistake",[123,157,158],{},"Human: a lesson from outside work",[123,160,161],{},"Industry: a hot take",[123,163,164],{},"Build: a work-in-progress",[97,166,167,172,175,178,181,184],{},[123,168,169],{},[48,170,171],{},"Week 3",[123,173,174],{},"Expertise: answer a FAQ",[123,176,177],{},"Build: a decision you made",[123,179,180],{},"Proof: a testimonial in context",[123,182,183],{},"Industry: a trend you disagree with",[123,185,186],{},"Human: something you changed your mind on",[97,188,189,194,197,200,203,206],{},[123,190,191],{},[48,192,193],{},"Week 4",[123,195,196],{},"Industry: a prediction",[123,198,199],{},"Proof: a small win",[123,201,202],{},"Expertise: a step-by-step",[123,204,205],{},"Human: a personal update",[123,207,208],{},"Build: what's next",[11,210,211],{},"Twenty slots, five buckets, no two the same twice in a row. You can shift the order to fit your week; the point is that every day already has a job.",[20,213,215],{"id":214},"adapt-it-per-platform","Adapt it per platform",[11,217,218],{},"The buckets stay the same across platforms; the format changes.",[220,221,222,234,244,250],"ul",{},[45,223,224,227,228,233],{},[48,225,226],{},"LinkedIn."," Longer posts work. Lean into Expertise, Proof, and Human, and open every one with a strong first line (",[229,230,232],"a",{"href":231},"\u002Fblog\u002Flinkedin-hooks","how to write a hook that earns the click",").",[45,235,236,239,240,233],{},[48,237,238],{},"X (Twitter)."," Shorter and faster. Industry takes and Build-in-public fit best, and longer points become threads (",[229,241,243],{"href":242},"\u002Fblog\u002Fx-thread-structures","thread structures that get read",[45,245,246,249],{},[48,247,248],{},"Instagram."," Visual-first. Proof (before\u002Fafter) and Human (behind-the-scenes) translate best to images and carousels.",[45,251,252,255],{},[48,253,254],{},"Newsletter."," Depth. One Expertise or Proof piece a week, expanded, with the shorter social posts feeding it.",[11,257,258],{},"Run the same five-bucket rotation, and let each platform take the version of the bucket that suits it.",[20,260,262],{"id":261},"map-the-month-in-one-sitting","Map the month in one sitting",[11,264,265],{},"Don't plan day by day. Sit down once, open the grid, and drop bucket labels across four weeks. You're not writing yet, just assigning jobs. That's twenty small decisions made together, when you have context, instead of twenty decisions made cold at 8am.",[20,267,269],{"id":268},"batch-the-writing","Batch the writing",[11,271,272],{},"Once the slots have jobs, write in batches by bucket. Three \"proof\" posts back to back is far faster than one of each, because you stay in the same headspace and reuse the same raw material. Block an hour, pick a bucket, and draft everything of that type for the fortnight.",[20,274,276],{"id":275},"leave-room-to-react","Leave room to react",[11,278,279],{},"Keep one or two slots a week loose. When something real happens (a launch, a news moment, a customer win), it goes there. A calendar that's 100% pre-planned can't respond to the present, and the present is usually your best material. The loose slots are where the calendar meets real life.",[20,281,283],{"id":282},"a-reusable-template","A reusable template",[11,285,286],{},"Copy this and refill it each month:",[220,288,289,295,301,307,313],{},[45,290,291,294],{},[48,292,293],{},"Buckets:"," Expertise, Industry, Proof, Build, Human.",[45,296,297,300],{},[48,298,299],{},"Grid:"," 4 weeks × your posting days, one bucket per slot, no repeats back to back.",[45,302,303,306],{},[48,304,305],{},"Loose slots:"," 1–2 per week, left open for reactive posts.",[45,308,309,312],{},[48,310,311],{},"Batch blocks:"," one writing session per bucket, per fortnight.",[45,314,315,318],{},[48,316,317],{},"Feed loop:"," at month's end, note which posts landed and make more of that bucket next month.",[20,320,322],{"id":321},"frequently-asked-questions","Frequently asked questions",[11,324,325,328],{},[48,326,327],{},"How many times a week should I post?","\nWhatever you can keep up without dropping quality. Five weekdays is a full rotation; three is fine and beats five you can't sustain. Consistency matters more than frequency.",[11,330,331,334,335,338],{},[48,332,333],{},"How far ahead should I plan?","\nPlan the buckets for four weeks; write one to two weeks ahead. Planning the ",[15,336,337],{},"jobs"," far out is cheap and stabilizing; drafting the exact words too far ahead means rewriting when the moment shifts.",[11,340,341,344,345,349],{},[48,342,343],{},"What if I run out of ideas for a bucket?","\nThat's usually a capture problem, not a creativity one. Keep a running note of raw ideas as they hit you, and shaping a week's posts becomes selection instead of invention. See ",[229,346,348],{"href":347},"\u002Fblog\u002Fidea-to-published-pipeline","the idea-to-published pipeline",".",[11,351,352,355,356,359],{},[48,353,354],{},"Do I need a tool for this?","\nA spreadsheet and a notes app run it fine for one brand. Across several clients, each needing its own grid, buckets, and drafts, keeping it all straight becomes its own job. That's where ",[229,357,6],{"href":358},"\u002F"," takes over: it plans the month, drafts each slot in the brand's voice with the AI tells stripped, and ships the posts, with Chloe running the calendar alongside you like a brand manager.",[11,361,362],{},"The goal is simple: never face a blank page cold. Give each day a job, batch the work, and leave a little space for real life to interrupt.",{"title":364,"searchDepth":365,"depth":365,"links":366},"",2,[367,368,369,370,371,372,373,374,375],{"id":22,"depth":365,"text":23},{"id":36,"depth":365,"text":37},{"id":85,"depth":365,"text":86},{"id":214,"depth":365,"text":215},{"id":261,"depth":365,"text":262},{"id":268,"depth":365,"text":269},{"id":275,"depth":365,"text":276},{"id":282,"depth":365,"text":283},{"id":321,"depth":365,"text":322},"2026-07-20","A five-bucket content calendar that keeps you posting without a blank page every morning: what to post, a filled four-week example, per-platform notes, and a reusable template.",false,"md",null,{},true,"\u002Fblog\u002Fcontent-calendar-you-can-keep",{"title":5,"description":377},"blog\u002Fcontent-calendar-you-can-keep",[387,388],"strategy","content calendar","eN-UW7Z51MqdV3_GT3y9Hp5saR1TaIgL7BNCblfsuFs",[391,687,901],{"id":392,"title":393,"author":6,"body":394,"date":678,"description":679,"draft":378,"extension":379,"image":380,"meta":680,"navigation":382,"path":681,"seo":682,"stem":683,"tags":684,"updated":380,"__hash__":686},"blog\u002Fblog\u002Fhow-many-clients-can-one-social-media-manager-handle.md","How many clients can one social media manager actually handle?",{"type":8,"value":395,"toc":670},[396,399,406,410,413,475,478,481,485,488,494,500,506,509,513,516,527,538,544,550,558,562,565,576,582,588,599,605,612,616,619,633,636,640,646,652,658,664],[11,397,398],{},"Ask ten freelancers and you get answers between three and fifteen. Both ends are true, which is why the question is usually answered badly. The number depends almost entirely on things that have nothing to do with how fast you write.",[11,400,401,402,405],{},"The useful version of the question is not \"how many clients\" but \"how many ",[15,403,404],{},"contexts"," can one person hold in a week without quality collapsing.\" That number is smaller, and it is the one that actually breaks.",[20,407,409],{"id":408},"where-the-hours-really-go","Where the hours really go",[11,411,412],{},"Take a typical mid-size client — twelve posts a month across two channels, one strategy call. The visible work is writing. Track it for a month and it comes out closer to this:",[91,414,415,425],{},[94,416,417],{},[97,418,419,422],{},[100,420,421],{},"Activity",[100,423,424],{},"Share of time",[118,426,427,435,443,451,459,467],{},[97,428,429,432],{},[123,430,431],{},"Writing and editing",[123,433,434],{},"25–35%",[97,436,437,440],{},[123,438,439],{},"Re-loading context (what did we agree, what's live, what's their tone)",[123,441,442],{},"15–25%",[97,444,445,448],{},[123,446,447],{},"Chasing and processing approvals",[123,449,450],{},"15–20%",[97,452,453,456],{},[123,454,455],{},"Reporting and account admin",[123,457,458],{},"10–15%",[97,460,461,464],{},[123,462,463],{},"Client comms outside scheduled calls",[123,465,466],{},"10–20%",[97,468,469,472],{},[123,470,471],{},"Scheduling and publishing",[123,473,474],{},"5–10%",[11,476,477],{},"Writing is a quarter of the job. Everything else is overhead, and overhead is what scales badly — because most of it repeats per client rather than per post.",[11,479,480],{},"This is why \"I can write a post in twenty minutes\" is a misleading input for capacity planning. Twelve posts is four hours. The client is not a four-hour client.",[20,482,484],{"id":483},"the-real-ceiling","The real ceiling",[11,486,487],{},"For most people running client work solo, the honest bands look like this:",[11,489,490,493],{},[48,491,492],{},"Three to four clients"," is comfortable, with room for the account that suddenly needs more. Each one gets real thought. This is the band where the work stays good and the money is often disappointing.",[11,495,496,499],{},[48,497,498],{},"Five to seven clients"," is the common working maximum for an experienced freelancer with a repeatable process. Achievable, but only if approvals are batched and context is written down rather than remembered. Without both, this band is where people burn out.",[11,501,502,505],{},[48,503,504],{},"Eight or more"," does not fail on writing capacity. It fails on context-switching and on being the single point of failure for eight relationships. Past this point you are running a small agency whether you call it that or not, and the answer is delegation, not efficiency.",[11,507,508],{},"Anyone selling you a workflow that gets you to fifteen is describing a volume business with templated output. That is a legitimate model. It is not the same job, and it does not command the same rate.",[20,510,512],{"id":511},"the-four-multipliers","The four multipliers",[11,514,515],{},"Two managers with the same number of clients can have wildly different loads. These are the variables that matter more than headcount.",[11,517,518,521,522,526],{},[48,519,520],{},"Approval latency."," A client who approves in a day costs a fraction of one who approves in a week. Slow approvals do not just delay work, they force you to hold the context open — you cannot close the account in your head until it ships. Two slow clients can cost more than four fast ones. This is the single biggest lever, and it is why ",[229,523,525],{"href":524},"\u002Fblog\u002Fsocial-media-approval-process","a designed approval process"," is a capacity intervention, not an admin one.",[11,528,529,532,533,537],{},[48,530,531],{},"Voice distance."," Three B2B SaaS clients are much cheaper to run together than a SaaS, a restaurant and a law firm. It is not the writing, it is the reset — ",[229,534,536],{"href":535},"\u002Fblog\u002Fmanaging-multiple-brand-voices","switching voices"," mid-day carries a real cost and the mistakes it causes are the embarrassing kind.",[11,539,540,543],{},[48,541,542],{},"Fact volatility."," A client whose numbers, pricing and product change monthly needs constant re-verification. A stable client's facts stay true. The volatile one can be twice the work at the same post count.",[11,545,546,549],{},[48,547,548],{},"Number of stakeholders."," One approver is a client. Three approvers who disagree is a committee, and a committee is two clients' worth of work regardless of how many posts it produces.",[11,551,552,553,557],{},"Price against these, not against post count. A five-post client with four stakeholders and weekly pricing changes is not cheaper than a twelve-post client with one decisive approver, and ",[229,554,556],{"href":555},"\u002Fblog\u002Fpricing-social-media-management","pricing by output"," is how people end up resenting their best-paying account.",[20,559,561],{"id":560},"raising-the-number-without-working-more","Raising the number without working more",[11,563,564],{},"Every real gain comes from removing repeat work, not from writing faster.",[11,566,567,570,571,575],{},[48,568,569],{},"Write the context down, once, per client."," Voice rules, permitted claims, forbidden topics, approver, turnaround, publishing days. The reason this raises capacity is not tidiness — it is that re-deriving context from memory is the largest hidden cost in the list above, and it recurs every single session. ",[229,572,574],{"href":573},"\u002Fblog\u002Fsocial-media-client-onboarding","Client onboarding"," is where this gets captured cheaply; after that it gets captured expensively.",[11,577,578,581],{},[48,579,580],{},"Batch by activity, not by client."," All drafting Monday, all approvals sent Tuesday, all scheduling Thursday. Switching costs are paid per switch, so five clients drafted in one block cost far less than five clients handled end-to-end one at a time.",[11,583,584,587],{},[48,585,586],{},"Make approval a fixed weekly slot per client."," Same day, same time, same batch size. It converts your most unpredictable dependency into a scheduled one.",[11,589,590,593,594,598],{},[48,591,592],{},"Standardise reporting."," Same format, same metrics, generated rather than assembled. Reporting is the most compressible line in the table because it is the least client-specific. ",[229,595,597],{"href":596},"\u002Fblog\u002Fclient-reporting-that-gets-read","What clients actually read"," is shorter than what most agencies send.",[11,600,601,604],{},[48,602,603],{},"Kill the client that eats a day."," Most people running five accounts have one that costs as much as the other four. Repricing it or ending it is usually worth more than any tooling change.",[11,606,607,608,611],{},"Tooling helps precisely where it removes repetition: brand context that persists per client, approval links that carry their own state, reports that build themselves. That is the part of the load ",[229,609,6],{"href":610},"\u002Ffor\u002Fsocial-media-managers"," is aimed at — not writing faster, but not re-loading the same five things every Monday.",[20,613,615],{"id":614},"a-number-you-can-actually-use","A number you can actually use",[11,617,618],{},"If you want a rough capacity estimate for yourself, budget per client per month:",[220,620,621,624,627,630],{},[45,622,623],{},"4–6 hours drafting and editing",[45,625,626],{},"2–3 hours approvals and chasing",[45,628,629],{},"1–2 hours reporting",[45,631,632],{},"2–4 hours comms and unplanned requests",[11,634,635],{},"That is 9–15 hours a month for a mid-size client, or roughly 2–4 hours a week. Against 25 genuinely productive client hours a week — not 40, since selling, admin and your own business exist — you land at five to seven accounts. Which is exactly where experienced freelancers report the ceiling to be, arrived at from the other direction.",[20,637,639],{"id":638},"faq","FAQ",[11,641,642,645],{},[48,643,644],{},"How many clients can a freelance social media manager handle?","\nFive to seven is the realistic maximum for an experienced solo manager with batched approvals and written per-client context. Three to four is comfortable. Beyond eight the constraint stops being time and becomes context-switching plus being the only point of failure for eight relationships.",[11,647,648,651],{},[48,649,650],{},"How many hours does one social media client take per month?","\nNine to fifteen hours for a mid-size client at around twelve posts a month, of which writing is roughly a quarter. The rest is approvals, reporting, re-loading context and unscheduled communication.",[11,653,654,657],{},[48,655,656],{},"Why does adding one more client feel like adding two?","\nBecause most of the cost is per client, not per post: a new account adds a whole context, a new approver, a new set of facts and a new relationship. Post-count-based capacity planning consistently underestimates this, which is how people end up over-committed at four clients.",[11,659,660,663],{},[48,661,662],{},"Should I charge per post or per month?","\nMonthly retainers reflect the work better, because post count is a poor proxy for load. Two clients at the same post count can differ twofold based on approval speed, stakeholder count and how often their facts change. Price those, not volume.",[11,665,666,669],{},[48,667,668],{},"What is the fastest way to take on one more client without more hours?","\nBatch by activity rather than by client, and move each client's approvals to a fixed weekly slot. Together those remove most switching cost and most chasing, which are the two largest compressible items in the week.",{"title":364,"searchDepth":365,"depth":365,"links":671},[672,673,674,675,676,677],{"id":408,"depth":365,"text":409},{"id":483,"depth":365,"text":484},{"id":511,"depth":365,"text":512},{"id":560,"depth":365,"text":561},{"id":614,"depth":365,"text":615},{"id":638,"depth":365,"text":639},"2026-08-25","The honest ceiling is lower than the job ads suggest, and writing hours are not what caps it. A breakdown of where the time really goes, the four multipliers, and how to raise the number without working more.",{},"\u002Fblog\u002Fhow-many-clients-can-one-social-media-manager-handle",{"title":393,"description":679},"blog\u002Fhow-many-clients-can-one-social-media-manager-handle",[685,387],"workflow","CPdLCbrPFfyT7JkmU389oHRg2tjfqyJ0RSHAajJIJSQ",{"id":688,"title":689,"author":6,"body":690,"date":892,"description":893,"draft":378,"extension":379,"image":380,"meta":894,"navigation":382,"path":895,"seo":896,"stem":897,"tags":898,"updated":380,"__hash__":900},"blog\u002Fblog\u002Fwhat-to-post-before-launch.md","What to post before you have a product",{"type":8,"value":691,"toc":883},[692,695,698,701,705,708,734,741,745,751,757,763,769,775,781,787,790,794,797,800,806,812,816,819,827,831,834,837,841,844,851,853,859,865,871,877],[11,693,694],{},"The pre-launch founder has the hardest content problem and the best raw material, and almost always uses neither. No customers means no case studies, no numbers, no social proof — so most people fill the gap with industry takes that could have been written by anyone, or with traction language that is not quite true yet.",[11,696,697],{},"Both are worse than posting nothing. The takes are invisible; the inflated claims are remembered.",[11,699,700],{},"What you do have is the thing nobody with a mature product has any more: an unfinished decision, in public, in real time. That is the whole strategy.",[20,702,704],{"id":703},"the-claims-that-cost-you-later","The claims that cost you later",[11,706,707],{},"Start with what not to write, because pre-launch is where founders damage trust fastest and least deliberately.",[220,709,710,716,722,728],{},[45,711,712,715],{},[48,713,714],{},"\"Our customers are seeing…\""," when there are three design partners on a free tier. The plural is the lie, and someone will ask.",[45,717,718,721],{},[48,719,720],{},"\"We're growing fast\""," off a base of 40 signups. Not false, exactly, and everyone reads it as false.",[45,723,724,727],{},[48,725,726],{},"Invented specificity."," A generated post will happily produce \"teams save an average of 6 hours a week\" because it is the shape a post takes. If you did not measure it, that number becomes something you have to defend or quietly delete.",[45,729,730,733],{},[48,731,732],{},"Borrowed authority."," Writing about scaling when you have not scaled anything. Readers who have done it can tell in one paragraph.",[11,735,736,737,740],{},"The rule that survives all of these: ",[48,738,739],{},"never claim a scale you cannot name."," \"One beta customer cut migration from four hours to twenty minutes\" is stronger than \"customers save hours\" precisely because it is narrow. Specific and small beats vague and big, every time.",[20,742,744],{"id":743},"seven-angles-that-work-with-zero-customers","Seven angles that work with zero customers",[11,746,747,750],{},[48,748,749],{},"1. The problem, described better than your buyer can describe it."," If you can articulate the pain more precisely than the person living with it, you have earned attention without claiming anything. This is the single highest-value pre-launch post and most founders skip straight past it to the solution.",[11,752,753,756],{},[48,754,755],{},"2. The decision you are in the middle of."," Not \"here's what we built\" but \"here's the fork we're at and what we're weighing.\" Live decisions get replies from people who have made them. Finished decisions get likes.",[11,758,759,762],{},[48,760,761],{},"3. What you're building and why this way."," The architecture, the constraint you accepted, the thing you deliberately made worse. Technical honesty travels.",[11,764,765,768],{},[48,766,767],{},"4. What you were wrong about."," You have been wrong about something in the last month. That post is more interesting than your roadmap and costs you nothing, because pre-launch nobody expected you to be right yet.",[11,770,771,774],{},[48,772,773],{},"5. Research in public."," You are talking to potential users. Post what you are hearing, anonymised, without conclusions. \"Six of ten people we spoke to had already built a spreadsheet version of this\" is a fact you own.",[11,776,777,780],{},[48,778,779],{},"6. The thing you refuse to build."," Positioning stated as a boundary is memorable and commits you publicly, which is useful discipline.",[11,782,783,786],{},[48,784,785],{},"7. Work in progress with the seams showing."," A screenshot of something half-finished, a migration that broke, a metric that is bad. Build-in-public works when the \"in public\" part includes the parts that are not going well.",[11,788,789],{},"Notice that six of seven require no traction at all. They require you to be doing the work and willing to describe it accurately.",[20,791,793],{"id":792},"build-the-audience-the-launch-lands-on","Build the audience the launch lands on",[11,795,796],{},"The reason to post pre-launch is not the posts. It is that a launch to zero people is a press release into a void, and a launch to 800 people who have watched you reason for six months is an event.",[11,798,799],{},"Two things matter more than volume:",[11,801,802,805],{},[48,803,804],{},"Reply to people, individually, before you need anything."," Pre-launch is the only period where you have time for this, and it is the only distribution strategy that reliably works from a standing start. Twenty real conversations beat two hundred impressions.",[11,807,808,811],{},[48,809,810],{},"Say what you are building, repeatedly, in one sentence."," Not a pitch. A sentence people can repeat to someone else. If your audience cannot describe what you do, they cannot refer you at launch, and referral is most of what a launch is.",[20,813,815],{"id":814},"the-frequency-question","The frequency question",[11,817,818],{},"Lower than you think. Two posts a week for six months beats daily for five weeks and then silence, which is the standard pre-launch pattern: enthusiasm, exhaustion, disappearance a month before the thing you were building for.",[11,820,821,822,826],{},"Pre-launch is a stamina problem, not an intensity problem. The ",[229,823,825],{"href":824},"\u002Fblog\u002Ffounder-content-30-minutes-a-week","30-minutes-a-week method"," exists precisely because the pre-launch period is long and the founder's attention is not.",[20,828,830],{"id":829},"when-the-product-ships","When the product ships",[11,832,833],{},"Do not throw away the pre-launch archive. The decisions, the wrong turns and the research are the raw material for everything you write afterwards — a launch post that references six months of public reasoning is dramatically stronger than one that starts from \"today we're excited to announce.\"",[11,835,836],{},"This is also where most founders lose the plot: the voice that got attention pre-launch was specific, uncertain and technical, and the launch flips it to marketing register overnight. Readers notice the switch immediately. Whatever else changes, keep the register.",[20,838,840],{"id":839},"where-a-tool-helps-and-where-it-does-not","Where a tool helps and where it does not",[11,842,843],{},"No tool solves the pre-launch content problem, because the problem is that only you know what happened this week. What a tool can do is stop the two failure modes: going dark when the build gets hard, and letting a generated draft quietly invent traction you do not have.",[11,845,846,847,850],{},"That second one is why ",[229,848,6],{"href":849},"\u002Ffor\u002Ffounders"," works the way it does — brand memory holds what is actually true, including \"customers: 0\", and a draft that would claim otherwise gets stopped rather than written. Pre-launch is exactly the situation where an AI writer is most dangerous and most tempting, because the gap between what you have and what a post wants is at its widest.",[20,852,639],{"id":638},[11,854,855,858],{},[48,856,857],{},"What should a startup post before launch?","\nThe problem stated more precisely than your buyer can state it, decisions you are currently weighing, what you are building and why this way, things you were wrong about, and anonymised findings from user research. All of these work with zero customers.",[11,860,861,864],{},[48,862,863],{},"How do I build an audience with no product?","\nPost twice a week for months rather than daily for weeks, and spend more time replying to individuals than writing posts. Pre-launch is the only period where you have time for one-to-one conversations, and they convert into a launch audience far better than impressions do.",[11,866,867,870],{},[48,868,869],{},"Should I build in public?","\nIt works when \"in public\" includes the parts going badly. Selective build-in-public that only shows wins reads as marketing and gets treated as such. If you are not willing to post the broken migration, the format is not for you.",[11,872,873,876],{},[48,874,875],{},"Is it bad to post about traction I don't have yet?","\nYes, and the damage is durable. Never claim a scale you cannot name. One named result stated narrowly is more persuasive than a vague plural, and it does not become something you have to defend later.",[11,878,879,882],{},[48,880,881],{},"How often should a pre-launch founder post?","\nTwo to three times a week, sustained. The pre-launch period is typically six to eighteen months, which is a stamina problem — the common failure is intense posting for a month followed by silence right before launch.",{"title":364,"searchDepth":365,"depth":365,"links":884},[885,886,887,888,889,890,891],{"id":703,"depth":365,"text":704},{"id":743,"depth":365,"text":744},{"id":792,"depth":365,"text":793},{"id":814,"depth":365,"text":815},{"id":829,"depth":365,"text":830},{"id":839,"depth":365,"text":840},{"id":638,"depth":365,"text":639},"2026-08-19","Pre-launch content without fake traction: seven angles that work with zero customers, the claims that quietly destroy trust, and how to build an audience the launch can land on.",{},"\u002Fblog\u002Fwhat-to-post-before-launch",{"title":689,"description":893},"blog\u002Fwhat-to-post-before-launch",[899,387],"founders","BSnR1nRUYwbsZ1GSrvhdK1ruP7msgCwyyzC96eDms3M",{"id":902,"title":903,"author":6,"body":904,"date":678,"description":1140,"draft":378,"extension":379,"image":380,"meta":1141,"navigation":382,"path":1142,"seo":1143,"stem":1144,"tags":1145,"updated":380,"__hash__":1148},"blog\u002Fblog\u002Fai-that-sounds-like-you.md","Getting AI to write in your voice (and why \"write like me\" fails)",{"type":8,"value":905,"toc":1131},[906,909,912,916,919,940,943,957,965,969,972,978,984,990,993,997,1000,1006,1017,1023,1029,1033,1036,1039,1042,1050,1054,1057,1065,1071,1075,1078,1089,1095,1097,1103,1109,1115,1125],[11,907,908],{},"You paste three of your best posts into ChatGPT, add \"write in this voice\", and get back something that is unmistakably not you. It has your topic and none of your speech. Add more examples and it gets more average, not more accurate.",[11,910,911],{},"This is not a prompting skill issue. The instruction \"write like me\" asks the model to infer a voice from evidence and then apply it, and the inference step is where it fails. What the model extracts from three posts is not your voice, it is the genre those posts belong to.",[20,913,915],{"id":914},"what-the-model-actually-learns-from-your-examples","What the model actually learns from your examples",[11,917,918],{},"Give a model five LinkedIn posts and it will reliably pick up:",[220,920,921,928,934],{},[45,922,923,924,927],{},"The ",[48,925,926],{},"format",": line breaks, list use, length, whether you open with a one-liner.",[45,929,923,930,933],{},[48,931,932],{},"register",": casual or formal, first person or third.",[45,935,923,936,939],{},[48,937,938],{},"topic domain",": SaaS, recruiting, design.",[11,941,942],{},"Those are the visible layer, and copying them produces something that looks like your posts from a distance. What it will not pick up from five examples:",[220,944,945,948,951,954],{},[45,946,947],{},"Which words you never use.",[45,949,950],{},"Which claims you would refuse to make.",[45,952,953],{},"What you find embarrassing.",[45,955,956],{},"The specific things you have actually done, which are the only source of material that could not have come from anyone else.",[11,958,959,960,964],{},"So the output lands in a predictable place: correct shape, generic substance. It reads like a competent stranger doing an impression, and readers detect it immediately even when they cannot name what is wrong. That gap is ",[229,961,963],{"href":962},"\u002Fblog\u002Fwhy-ai-social-posts-get-ignored","why AI-written posts get ignored"," rather than disliked. Nothing about them is bad. There is just nobody in them.",[20,966,968],{"id":967},"a-voice-is-three-separate-things","A voice is three separate things",[11,970,971],{},"Treating \"voice\" as one setting is what makes it unfixable. It is at least three, and they need different inputs.",[11,973,974,977],{},[48,975,976],{},"Style."," Sentence length, punctuation habits, whether you use questions, how you open and close. This is the layer examples genuinely teach. It is also the least important, because it is the layer readers notice last.",[11,979,980,983],{},[48,981,982],{},"Vocabulary boundaries."," The words that are yours and the words that are disqualifying. Most people cannot list these from memory but recognise them instantly: \"leverage\", \"unlock\", \"in today's landscape\", \"game-changer\". A negative list works better than a positive one, and it is much easier to write.",[11,985,986,989],{},[48,987,988],{},"Substance."," The facts, numbers, opinions and stories you are allowed to draw on. This is the layer that makes a post identifiably yours, and no amount of style transfer produces it. If the model does not have your material, it will invent something plausible in your shape, which is the worst of both.",[11,991,992],{},"Almost everyone tries to solve a substance problem with style instructions. That is why the loop feels endless.",[20,994,996],{"id":995},"what-to-give-a-model-instead","What to give a model instead",[11,998,999],{},"If you want output you would actually publish, give it these four things and stop tuning adjectives.",[11,1001,1002,1005],{},[48,1003,1004],{},"Ten to twenty of your own posts, not three."," Three is enough to average, not enough to characterise. And use ordinary posts, not only your best: your best posts are your most edited, which means they are the least like how you write.",[11,1007,1008,1011,1012,1016],{},[48,1009,1010],{},"A short list of banned words and constructions."," Ten to fifteen entries is plenty. Include the ones that are technically fine but not you. This is the single highest-return input and takes fifteen minutes; if you want a starting list, the ",[229,1013,1015],{"href":1014},"\u002Fblog\u002Fai-writing-tells","common AI tells"," are most of it.",[11,1018,1019,1022],{},[48,1020,1021],{},"A file of facts you are allowed to cite."," Numbers with the date they were true, customer results with permission status, opinions you hold and would defend. This is boring to assemble and it is the entire difference between a post about your company and a post that could be about any company.",[11,1024,1025,1028],{},[48,1026,1027],{},"Two or three examples of what you rejected and why."," \"Too salesy\", \"I would never say hustle\", \"this claims something I cannot prove.\" Rejections carry more signal per word than approvals, because approval only says the output was acceptable while a rejection names a boundary.",[20,1030,1032],{"id":1031},"the-test-that-tells-you-whether-it-worked","The test that tells you whether it worked",[11,1034,1035],{},"Take the draft and remove your name, your company name and the topic. Show it to someone who knows you.",[11,1037,1038],{},"If they can tell it is yours, the voice landed. If they say \"this could be anyone in your industry\", it did not — regardless of how well it matched your line breaks.",[11,1040,1041],{},"A cheaper version you can run alone: read it out loud. The places you stumble or feel faintly embarrassed are almost always sentences you would not have written. Voice mismatches are much easier to hear than to see, which is why they survive so many silent re-reads.",[11,1043,1044,1045,1049],{},"For the mechanical layer, the constructions that mark text as machine-written regardless of whose voice it was aiming at, you can run a draft through our ",[229,1046,1048],{"href":1047},"\u002Fai-check","AI writing check"," and see the specific spans flagged rather than a score.",[20,1051,1053],{"id":1052},"why-a-chat-window-keeps-losing-your-voice","Why a chat window keeps losing your voice",[11,1055,1056],{},"Even when you get a good result in ChatGPT, it does not persist. Next week you open a new conversation, the model has none of it, and you re-explain from scratch or paste a prompt block you now maintain by hand.",[11,1058,1059,1060,1064],{},"This is the actual daily cost, and it is not about output quality at all. It is that the voice lives in a conversation rather than in the account, so every session starts at zero and every improvement has to be manually carried forward. Anyone who has ",[229,1061,1063],{"href":1062},"\u002Fblog\u002Fchatgpt-claude-for-social-media","used ChatGPT or Claude for social content"," for more than a month has a document somewhere that exists only to re-teach the model things it already learned.",[11,1066,1067,1068,1070],{},"Making the voice a stored profile rather than a prompt is most of what ",[229,1069,6],{"href":849}," does: the samples, the banned list, the facts and the rejections accumulate in one place, and every draft starts from them instead of from a blank context. The interesting part is not that drafts get better. It is that a correction you make once stops being a correction you make weekly.",[20,1072,1074],{"id":1073},"what-will-still-not-work","What will still not work",[11,1076,1077],{},"Two honest limits, because the promise gets oversold.",[11,1079,1080,1083,1084,1088],{},[48,1081,1082],{},"Nothing gives you material."," If you have not done anything worth writing about this month, no voice profile produces a post worth reading. The model can shape what you have; it cannot supply it. This is the same reason ",[229,1085,1087],{"href":1086},"\u002Fblog\u002Flinkedin-ghostwriter-cost","hiring a ghostwriter"," does not remove the founder from the loop.",[11,1090,1091,1094],{},[48,1092,1093],{},"Voice is not a substitute for judgment."," A well-tuned model will happily write a sentence in your voice that you should not publish. The last read is still yours, and the useful version of that read is checking claims rather than admiring phrasing.",[20,1096,639],{"id":638},[11,1098,1099,1102],{},[48,1100,1101],{},"Why doesn't ChatGPT sound like me even when I give it examples?","\nBecause from a handful of examples it extracts format and register, the visible layer, not your vocabulary boundaries or your material. You get the right shape with generic substance, which reads as an impression. Fix it by supplying more samples, an explicit banned-word list and a file of facts you can cite, rather than more adjectives.",[11,1104,1105,1108],{},[48,1106,1107],{},"How many writing samples does an AI need to learn a voice?","\nTen to twenty is the useful range for social posts. Three averages rather than characterises. Include ordinary posts, not only your best ones — heavily edited work is the least representative of how you actually write.",[11,1110,1111,1114],{},[48,1112,1113],{},"Is it better to describe my voice or show examples?","\nBoth, for different layers. Examples teach rhythm and structure, which people describe badly. Explicit rules work better for boundaries: what you never say, what you refuse to claim. Descriptions like \"professional but approachable\" carry almost no information and are the most common wasted instruction.",[11,1116,1117,1120,1121,349],{},[48,1118,1119],{},"Can AI-written posts be detected?","\nDetectors measure statistical patterns and are unreliable on short text, so a confident verdict on a 200-word post is worth little either way. Readers are the real detector, and they respond to the absence of specifics rather than to phrasing. See ",[229,1122,1124],{"href":1123},"\u002Fblog\u002Fwhat-ai-detectors-catch","what AI detectors actually catch",[11,1126,1127,1130],{},[48,1128,1129],{},"What is the fastest thing I can do to make AI drafts sound more like me?","\nWrite a list of ten words and phrases you would never use, and put it in the prompt. It takes fifteen minutes, needs no tooling, and removes most of what makes a draft read as machine-written.",{"title":364,"searchDepth":365,"depth":365,"links":1132},[1133,1134,1135,1136,1137,1138,1139],{"id":914,"depth":365,"text":915},{"id":967,"depth":365,"text":968},{"id":995,"depth":365,"text":996},{"id":1031,"depth":365,"text":1032},{"id":1052,"depth":365,"text":1053},{"id":1073,"depth":365,"text":1074},{"id":638,"depth":365,"text":639},"Prompting an AI to sound like you almost never works, and the reason is mechanical. What a voice is actually made of, what you have to give a model, and how to tell whether it landed.",{},"\u002Fblog\u002Fai-that-sounds-like-you",{"title":903,"description":1140},"blog\u002Fai-that-sounds-like-you",[899,1146,1147],"ai","writing","o_PBouvxI0s03S8G-pPaBBw4lqbGchRi3EAnBrrt11c",1788343317046]