Michel Lieben

I got fired from 4 jobs, failed at 7 startups in a row, lost $40K of my own savings, and then built a $6M business. These are the only 7 lessons I wish I had known earlier 👇 1. You're dumb until you're a genius

I built a $7M LinkedIn content strategy from scratch for a founder. (these are the 5 moves I'd walk you through if we had an hour. same ones coldiq runs every post) watch the full session:

We built 12 Claude Code skills that run our entire paid media ops across Google, Meta, and LinkedIn at ColdIQ (and we're giving the whole pack away). Our head of growth Ivan Falco runs $200K/month in ad spend from a terminal. It's how we doubled client load this year without losing quality. The skills do the work that used to fill our media buyers' calendars: spot creative fatigue, adjust bids, upload audiences, run bulk edits, flag broken campaigns, build reports. Each skill does a specific job: Google Ads: → keyword-analyzer: audits quality scores and finds keyword gaps → negative-keywords: reviews search terms and blocks wasted spend → performance-auditor: compares periods and flags what changed → search-terms: surfaces queries burning budget with zero conversions Meta Ads: → audience-builder: turns CRM lists into custom audiences → creative-fatigue-analyzer: spots declining CTR before the metrics flag it → fatigue-monitor: flags when your audience is saturated → spend-tracker: tracks budget pacing across every campaign LinkedIn Ads: → audience-builder: builds targeting audiences at scale → bid-optimizer: adjusts bids across campaigns in bulk → bulk-editor: mass edits campaigns, ads, and naming in seconds → creative-builder: generates ad creatives from brand specs You drop them into Claude Code, connect your ad accounts, and tell it what you need. It reads the skill, plugs into the platform, executes. 300+ hours of work went into building these. Comment ADS and we'll send all 12 over.

The top 1% of cold email campaigns hit 20-30% reply rates. Everyone else gets 2-3%. Same channel. 10x gap. @InstantlyAI analyzed 1,000,000+ emails to map what separates them. 8 things: > small lists, not broad blasts > hyperenriched data, not name + email > openers that prove you researched them > 3-step sequences, not single sends > value-first asks, never calendar-first > ICP, offer, copy. nail the basics first > deliverability hygiene over volume > the right stack (Instantly, Clay, Prospeo) full breakdown in the cheat sheet:

founders after discovering they can pull 6x roi by deploying the method in this article instead of running product ads

Most people see one part of ColdIQ, usually the LinkedIn content, and assume that's the engine. It's not. The thing that pulls 55,000+ visits a month and supports $7M ARR has four layers: marketing brings traffic in > lead capture turns it into pipeline > sales closes > and delivery scales the work without breaking. 1. Marketing (55,000+ visits/mo) This is where the traffic enters. Five channels carry the load. > LinkedIn (~15,000 visits/mo). The heaviest lifter. We have 20+ teammates publishing 150+ posts a month across outbound, AI, and GTM, with a combined audience of about 300K, designed in-house. Stack: Taplio, Scripe. > SEO (~10,000 visits/mo). Yevhen runs programmatic SEO. Long-form blogs, programmatic landing pages on AirOps, and free micro-tools that Sacha builds in Claude Code. Stack: AirOps, Ahrefs, Claude Code, Webflow. > Paid (~6,000 visits/mo). Ivan runs roughly $10K/mo across LinkedIn, Meta, and Google, mostly to test new channels rather than scale a winner. Stack: Claude Code, Fibbler, Usermaven. > Cold outreach (~4,000 visits/mo). Kenny runs internal campaigns off tight lists of fewer than 5,000 deeply enriched contacts at a time, rotating between value-driven lead magnets, lookalike account targeting, social engagement plays, and inbound-led outbound. Stack: Clay, Claude Code, Apollo, Instantly, lemlist, Prospeo, FullEnrich, CompanyEnrich, PredictLeads. > Newsletter, webinars, and emerging channels (~11,000 visits/mo). Monika ships the beehiiv newsletter. Harry runs the webinars. TikTok, Instagram, and X are newer experiments growing share fast. 2. Lead capture Traffic alone doesn't convert, so lead capture is its own layer sitting on top of marketing. > Midbound and Vector surface anonymous visitors. > Customer .io handles lifecycle emails. > Micro-tools and pop-up lead magnets quietly collect emails. > Everything routes through n8n and gets scored in real time by Clay, OpenAI, and PredictLeads before a salesperson ever sees it. 3. Sales > Sales-led motion, 2 to 3 Google Meet calls, $5K to $10K/month retainers. > The qualifying happens before discovery. Case studies, VSL, and free content do the weeding. > Once a call is booked, Clay enriches the account on funding, headcount, tech stack, tenure, and reports, and the proposal goes out shortly after. > Stack: Attio for CRM, Attention for AI notes, Hyperline for CPQ, Qwilr for proposals, Stripe for payments. 4. Client delivery > n8n runs the kickoff workflow. > GSuite, Supabase, Stripe, and Slack are synced so admin tasks like Slack invites, onboarding links, and domain provisioning happen without anyone in the loop. > Our GTM engineers only touch what moves the needle for clients: custom ICP matrix, creative campaign concepts, copy. (bookmark for later)

how to generate unlimited leads with @openclaw

founders after realizing "anonymous visitors" was the highest-intent pipeline source on their dashboard the whole time

We added $151K in MRR in 87 days with 20+ people at ColdIQ posting on LinkedIn every week. None hired a ghostwriter. None took a writing course. Each one still sounds unmistakably like themselves. We pulled it off by treating content like an engineering problem. Let me explain. Content scales like engineering. Most teams treat it as creative work and the voice converges by month two. We solved this with a per-person system loaded before every generation. The pipeline runs 27 skills across eight stages, all in one terminal. One folder per person. A grading rubric refuses to ship anything below a /38. Layer 1: Foundation (built once per person) Every AI content engine I've audited fails at the same point. By month two, the voice drifts toward whatever the base model considers good business writing. Readers unfollow. One folder per human, loaded before every generation: • A voice profile built from a 25-question spoken conversation • An ICP doc split into three buyer tiers, each with its own language map • A content pillars file locking the person's three to five themes The voice profile is the highest-leverage file. We build it from a recorded call. How does the person describe what they do at a dinner. What words do they refuse to use. What phrases do they reach for when explaining something technical to a non-technical buyer. The transcript becomes ground truth, and every draft is compared against it. The pillars file is the final gate. If a draft doesn't deposit into one of the three to five themes, the system rejects the idea before it becomes a draft. Layer 2: Research (six workers, every week) Six workers run weekly and drop a ranked idea batch into each person's folder by Monday morning. • Apify browses LinkedIn for what's performing in the niche right now. • Reddit mines the communities our buyers live in for raw, unedited audience language. • YouTube feeds through Gemini for multimodal video analysis, turning long-form video into pillar-mapped frameworks. • X surfaces the live ICP debates and the contrarian angle worth taking. • Fireflies transcripts are the highest-signal source. Clients describe their problems in their own language on diagnostic calls, and that language becomes hooks unedited. • Repurposing archive indexes everything we've produced so nothing accidentally repeats. A seventh skill, reverse engineering, runs on demand. It scrapes top niche posts and maps reusable templates back into the hook and copy skills. Each idea arrives pre-tagged with hook, pillar, and source. Nobody argues about what to post. Layer 3: Production line (hook → copy → grader) Four skills, each does one job, hands cleanly to the next. Hook generator. 50 templates organised by emotional trigger across desire, curiosity, and fear. Each idea produces 20+ hook variations. Underperforming templates get deprecated. Copy developer. Takes the winning hook and writes a draft in the person's documented voice, loading voice profile, ICP tier, matched pillar, and research snippets. Visual brief generator. Produces a structured layout plus exact copy placement for the designer. Gemini handles first-pass visuals; Figma and Canva sit above for refinement. Post grader. Scores every draft on five dimensions: hook strength, voice fit, specificity of claims, scannability, pillar relevance. /50 ceiling. /38 floor. Anything below 38 auto-rewrites with the failing dimension flagged. Nothing ships under a 38. AI drafts; the human sharpens. The voice profile is the guardrail. Layer 4: Repurpose (one post, six surfaces) The repurpose skill rebuilds a validated LinkedIn post into an X thread, an X long-form article with a cover placard, a newsletter, an SEO blog post, a YouTube video script, and a carousel of sequential visual slides. Each destination gets a native rebuild with native formatting, opening patterns, and a CTA that fits the surface. Voice profile and ICP tier carry over unchanged. Only the structural grammar of the destination shifts. Layer 5: Refresh + maintain Feedback capture auto-logs every session: what ran, what the human kept verbatim, what got rewritten, what the grader rejected. Pattern recognition reads the learning log every five sessions and outputs recommendations to the client folder. Voice refresh runs monthly against new transcripts, flagging drift in voice, ICP language, pillars, and CTAs. Content audit runs quarterly against Taplio's top and bottom performers. System check runs before every production round across Apify, Fireflies, Gemini, Reddit, YouTube, and Claude. Layer 6: Delivery ClickUp is the client-handoff surface. Every approved draft, visual brief, and grader scorecard lands in the client's ClickUp board, where the human can edit, approve, or kick back. Taplio queues approved posts at the right time slot per person and per pillar. Top performers train the hook templates. Underperformers feed the rubric. The loop closes. What to build first 1. The voice foundation: a 25-question spoken conversation, an ICP doc with three buyer tiers, a pillars doc with three to five themes. 2. The post grader: every draft scored on the five dimensions before it ships, anything under 38 rewritten. 3. The Monday research cron: six workers running Sunday night so each person's folder has ranked ideas waiting. Everything else is scaffolding for these three. If these three are clean, the system is working.

> take a mac mini m4 and install @openclaw > spin up one github repo for your whole linkedin engine because folder sprawl is a skill issue > write your linkedin post skill as a workflow, not a knowledge dump, because if your prompt is a doc you're cooked > reverse-engineer your competitors' sites with claude code's chrome extension because curation beats prompting > pull topics from sales calls, not competitor feeds, because the creators you scrape don't know your buyers > wire a daily cron that feeds yesterday's linkedin performance back into the skill because guessing is for hobbyists > name your agents and give them a 9am standup because if you can't manage a real team you might as well manage AI ones > give your orchestrator its own email and let it spawn sonnet sub-agents to draft your linkedin posts because a CSO should delegate > publish blogs your agents wrote so claude and chatgpt cite them back to your buyers > grow to 30k+ linkedin followers running marketing, sales, and customer success off five agents and one mac mini

@LogikaAgency @openclaw thank you.

@MichLieben @openclaw some good content , thanks guys. All i see mostly is copy pasta. this was really authentic.

how to grow your LinkedIn account on autopilot with @openclaw

every b2b marketer running "book a demo" ads to a cold audience after reading this

We just analyzed 100,000+ LinkedIn DMs sent through @Expandi_io to figure out what the top 1% actually do differently. A few patterns that stood out: > Messages under 150 characters pulled +22% reply rates > 3-5 step sequences outperformed single-shot by +42% > Warm-up before connecting (profile visit + like + follow) lifted acceptance by +30.2% > ICP-personalized copy beat generic copy by +54.7% Full cheat sheet below 👇