Founder LinkedIn engine

1. Research the founder off-page, working from the Resources doc: blogs, podcasts, interviews, talks, newsletters, pull recurring stories, strong opinions, and phrases in their literal voice.
2. Draft the Founder avatar & pillars doc: tone of voice, stories, opinions and an initial 3–5 content pillars, grounded in the LinkedIn audit + off-page research.
3. Interview the founder (call or async voice notes) with the best-practice questions: how would you describe the tone you want to convey · what could you talk about for hours · what hot takes do you hold about your industry · which stories come up again and again · who exactly are you trying to reach and what should they think or do · any no-go topics, words or styles · 2–3 posts (yours or others') whose style you'd emulate · what's happening next quarter we should build content around.
4. Fold the answers in; agree the objective, brand awareness vs lead generation, set the tiered OKRs on the scoreboard (outcomes first: conversations started, attributed demos/leads; then engagement rate ≥3%, impressions, followers), and seed the Experiment bank doc.
5. Human approval, the founder answers the open questions and signs off the summary of what we've learned: avatar, pillars, objective. Refresh quarterly.
6. Build an anecdote bank in the avatar doc from the interview: the founder's real stories, each with the moment, the people, and the lesson. Real stories can't be generated, so they are the part of the voice that survives scale, capture them in conversation, not by questionnaire.
7. More than one voice (founder plus a co-founder or execs): before any writing, give each person a distinct topic lane tied to their real role, and a one-page voice sheet (rhythm, words they use and avoid, hard rules, their anecdotes). Note where lanes may deliberately overlap (e.g. a launch week). Revisit the lane map at the quarterly refresh.

1. Sync external-facing call recordings, sales, customer, partner calls, into a shared folder via the call recorder (e.g. Granola); make recording a standing requirement at onboarding.
2. Run a pass over the period's transcripts: pull specific numbers, customer stories, objections, and strong opinions that match the pillars.
3. Sweep Slack for the founder's messages and team updates that would make posts.
4. Score each insight against the pillars and append the best to the Idea bank doc with source quotes, tagged by pillar, aim for 30–50% of the month's ideas from here. Run monthly; weekly once the flow is smooth.

1. Research what's live in the founder's space, popular topics, formats, conversations worth a take (~25% of ideas).
2. Prepare interview questions from the avatar and this month's mined insights, questions, not pre-baked themes.
3. Run a short content call (or async voice notes): the founder talks, you ask; the transcript is the raw material and carries their literal tone of voice (~25% of ideas).
4. Append the strongest angles to the Idea bank doc.
5. When the founder says there's "nothing to post", start from what has already happened rather than what might: a customer's exact words, the real reason a deal was lost or a customer left, one number with its before and after, a decision that was reversed or a feature cut, raw product output, a category belief the founder disagrees with, a loose term made precise. Pick one family per month and hold it, mixing them trains mixed expectations.

1. Compile the Idea bank into a Monthly brief doc: 10–15 post briefs, each with pillar, topic, insight/inspo, source quote, and asset idea; mirror them into the Upcoming content doc as awaiting review.
2. Where a brief needs depth, ask the founder 2–3 targeted questions in Slack, text or voice note, answered whenever suits them, and fold the answers in.
3. Human approval, draft review: the reviewer (account manager or the client themselves) checks the briefs against the pillars, cuts anything off-strategy, and the founder confirms the topics, in the workspace, or share the brief doc to Google Docs if the client prefers commenting there.
4. Every brief names the position or open question the post will put to readers. A neutral summary earns scrolling; a stated view or a question worth answering earns comments, and comments are what carry the post.
5. Several profiles (optional). If more than one profile is in scope (e.g. two founders plus the company page), set a brief quota per profile (e.g. 10 / 10 / 5) and keep one Content Plan doc per profile. On adoption, duplicate Write → Final review → Schedule as one lane per profile so each person signs off their own posts.

1. Draft each post in the founder's voice from the avatar + brief + source quotes: specific hook, one idea per post, short, end with a take or a question, not a CTA.
2. Write like a person, not a brand; keep the real numbers and stories from the mined calls.
3. Write the posts into the monthly Content Plan doc, one doc per profile, one numbered section per post (topic → Approved? → post copy → asset brief), and prepare assets and the schedule plan (3×/week at a consistent time).
4. Keep outbound links out of the post body, the link goes in the first comment. Draft that comment, plus 3–5 follow-up questions for the thread, alongside each post.
5. With several voices on one calendar, never ship the same idea from two accounts in the same week unless it's a planned campaign. Shared proof points get reframed through each person's lane.
6. Two creative variants (optional). Where a post carries a visual, produce two on-brand asset variants sized for LinkedIn and pair both with the copy, so the final review can pick a lead or run an A/B test.

1. Human pass on every post in the Content Plan doc: tone check against the avatar, fact-check numbers and named clients, strategy alignment.
2. Human approval, final review: send the Content Plan Google Doc to the founder (or the account manager reviews on their behalf), they comment inline and mark "Approved? yes" per post; link the signed-off doc to this step.
3. escalate anything sensitive, confidential client detail, legal exposure, controversial takes.
4. Where a post has two asset variants, sign off the copy and assets together and pick the lead variant, or keep both as an A/B test logged in the Experiment bank doc.

1. Schedule the approved posts from the signed-off Content Plan doc through Outstand across LinkedIn (plus Instagram/X where in scope) at the agreed times, Outstand's queue is the schedule of record (max ~30 days ahead; the month plan lives in the workspace and posts enqueue as they come into window). Move them to "scheduled" in the Upcoming content doc with dates.
2. Reply to every comment in the first hour, replies pull the algorithm.
3. Turn the best comment thread into a next-post idea for the Idea bank.
4. Work the launch window (defaults, tune on your own data):

- T−30 min: first comment with the link and the seed questions ready; publish into the hours the audience actually reads.

- T+0–15: the founder answers every comment individually and in substance.

- T+15–60: keep threads open rather than closing them with thanks; add a second question once the first wave settles.

- T+1–2h: keep replying; optionally post to a second surface for a second pulse, never a duplicate.

- Never edit a post after publishing; never post and leave.

1. Several accounts: stagger posting windows and formats across the day, and have each account spend its engagement time outward, in its own lane, where its audience actually is.
2. Boost only proven winners. Default trigger: engagement at ~1.5–2× the account's own trailing baseline within 6–8 hours, with clearly positive sentiment. Stop on rising cost per action, falling organic engagement, or negative comment clusters. Human approval before any spend.

1. Sort DMs and connection notes: prospect / hire / noise.
2. Move real prospects to a quick call, tag the source in the CRM, and log every conversation started, it's the primary outcome metric and no API exposes it.
3. escalate anything sensitive.
4. Treat comments and likes differently. A comment from someone matching a buying persona gets the full pass: identify the person and company, check the CRM for an existing owner, deal or customer status, log the touch, and draft a connection note or reply for approval. A lone like is logged only, unless the account is a priority account or this is their second or later interaction.
5. Ignore engagement from teammates and from roles outside the buying committee. Never message from the founder's account without sign-off.

1. Pull post-level analytics through Outstand, likes, comments, shares, impressions, reach, engagement rate per post, plus account-level followers (and page clicks/engagement where the company page is connected), log them as measurements on the scoreboard; no manual copying. Profile views: monthly screenshot from the founder's native analytics.
2. Post the weekly pulse to Slack: what published, early numbers, what's queued, anything blocked. Refresh the Upcoming content doc (published → recently published).
3. Monthly: write the report in the workspace, outcomes first (conversations, attributed pipeline), then engagement, then reach, rolled up to the OKRs; note which formats and topics worked.
4. Grade last month's experiment in the Experiment bank doc and pick 1–2 for next month; feed winners back into insight mining and the quarterly avatar refresh.
5. Read each account against its own baseline, not the average. If all accounts rise and fall together, look for the shared element, hook shape, format, posting window, production batch, and break it before diagnosing individual posts.
6. Read engagement by who, not how many: a small audience of the right buyers beats a large audience of peers.

Optional: use when the founder's market moves fast enough that a monthly trend pass misses things.

1. Scan the last 7 days across the 2–3 fixed buckets set in the Founder avatar & pillars doc (e.g. category news, competitor moves, regulation or data releases that matter to the audience). B2B only; check figures against the primary source.
2. Note competitor posts and launches that are getting traction, and what angle they took.
3. Append the 3–5 strongest angles to the Idea bank doc, each with the source link and the take the founder could own.
4. Report: what moved this week, and which angles are ready for next month's briefs. A quiet week is a normal result; say so.