Emily's advantage is building marketing systems in public
Most marketers post opinions about AI. Emily runs the experiment, pulls the data, and hands you the exact system she just built.
Emily Kramer is the co-founder of MKT1, where she writes one of the most-read B2B marketing newsletters, advises marketing leaders and founders, and invests in early-stage startups. Before MKT1 she was the first-ish marketer four times over, building teams from scratch at Asana, Carta, Astro, and Ticketfly. Her LinkedIn is not a stream of takes. It is a working lab notebook: every post teaches one repeatable framework, one proprietary data point, or one hands-on AI workflow a marketer can copy that week, usually while she is still figuring it out herself.
That is the whole engine. Systems-led marketing growth is when you build an audience by teaching the repeatable frameworks, data, and tools you use every day, so your content doubles as the product you sell. Emily runs it with unusual discipline: do the real work, capture the system behind it, hand it over in a chart or a step list, and end with a genuine question that pulls the audience into the room.
Reposts the same 'AI is changing marketing' opinion everyone already agrees with, and leaves the reader with nothing to build.
Runs the experiment, pulls the data, then hands you the chart, the step list, or the skill you can copy this week.
“The best time to build a skill is when you're already doing a repeat task.”
— From her most-reacted post, a Claude Code field guide for marketers (631 reactions)
Five findings that repeated across 100 posts
- The account is a marketing R&D lab in public. Her biggest posts are original research and hands-on AI experiments: a roughly 100-hour Claude Code deep dive (631 reactions), a State of B2B Marketing data pull, a live buildathon.
- Conversation over reach. She averages 109 reactions but 41 comments a post, a 37.6% comment-to-reaction ratio, roughly six times the LinkedIn norm.
- Every post hands you a system. A chart, a step list, a named role, or a template, something a marketer can apply immediately.
- She teaches the tool she is learning. Claude Code, MCP servers, agents, and AEO, narrated in real time with the mistakes left in.
- Weekday discipline. About 4.1 posts a week, Wednesday and Monday heaviest, with just 3 of 100 posts landing on a weekend.
The numbers behind the account
The story here is not raw reach. It is a steady weekday cadence and an unusually high rate of real conversation.
Across the 100 posts we analyzed, Emily published about 4.1 times a week, almost entirely on weekdays. Wednesday and Monday are her heaviest days, and only 3 of 100 posts landed on a weekend. The reach itself is honest and moderate: she averages 109 reactions, a median of 77, and a top post at 631. None of the 100 cleared 1,000 reactions. If you judged her on virality alone you would miss the point entirely, because the signal that matters is buried in the comments.
When she posts
The content-type mix
Where the engagement comes from
That 6.3% 'Interest' share is unusually high, and it fits the account: people react with curiosity to a chart or a workflow they want to try, not just applause. It is the reaction of an audience that is learning, not spectating.
The top posts
Want to see how your own cadence and comment ratio stack up? Run your profile through our free LinkedIn analyzer.
The six content pillars
Every post is one of six repeatable buckets, so a founder writing a newsletter, advising, and investing never runs out of things to teach.
Claude Code, MCP servers, skills, and agents, taught hands-on for a marketing audience.
Proprietary stats from her State of B2B Marketing study of 100 startups, turned into charts.
Hands-on breakdowns of Mutiny, Profound, Framer, and Wistia, framed by the problem they solve.
New roles and org charts: the Gen Marketer, the Marketing Engineer, the silo-busters.
Newsletter milestones, the MCP server, buildathons, and the job board, narrated as they launch.
Family, puppies, self-deprecating confessions, and the messy middle of learning in real time.
Pillar 1: AI workflows for marketers (the volume engine)
Why it works: Her single biggest post is a field guide, not a flex. She puts in the 100 hours, then hands the reader the eight things she wishes she had known, so the effort becomes a shortcut for everyone else. Teaching the tool you are still learning, mistakes included, is her most reliable engine.
Pillar 2: Original data & research (the reach engine)
Why it works: She does not cite someone else's study, she runs her own on 100 startups, then leads with the single most provocative stat. A proprietary number nobody else has is the most defensible reach engine on the platform, and it makes the reader check their own site on the spot.
Pillar 3: Tool & product teardowns (the credibility)
Why it works: She never reviews a tool from the outside. She uses it, names the founder, and frames the launch around the exact marketing problem it fixes. Teardowns like this make her the trusted filter for a noisy category, which is why founders want her to be an early user and investor.
Pillar 4: The changing marketing org (the point of view)
Why it works: She names the question every one of her advisees is already asking, then gives it a label: the Marketing Engineer, alongside her own Gen Marketer. Naming the role people feel but cannot articulate is category-defining, and it makes her the reference point for the future org chart.
Pillar 5: Building MKT1 in public (the momentum)
Why it works: She narrates the business's growth in public, milestone by milestone, with the exact number and a bit of genuine excitement. Sharing the climb, not just the summit, makes the audience feel like part of the story, and every milestone doubles as proof for the method she sells.
Pillar 6: Candid & personal (the human)
Why it works: A warm, funny scene that still teaches (even people in their 70s learn the tool fast). Dropping the expert voice to be simply human is what keeps a systems-heavy account from feeling like a manual, and these posts earn the goodwill the rest of the strategy spends.
The hooks that earned the comment
Her openers are built to pull a reply, not just a scroll-stop. Most are a surprising number, a real question, or a naming take you have to react to.
Lead with a number only you have. '40% of sites in my recent State of B2B Marketing Report had no homepage schema.'
Ask something you truly want answered. 'Have you built a marketing agent that runs regularly?'
Give a confusing thing a plain verdict. 'Claude Code is a confusing name. It kind of alienates marketers.'
Show the messy middle. 'Confession: I'm not as great as I'd like to be at growing my own LinkedIn followers.'
React to a launch as it happens. 'Framer just announced Agents, built right into their website builder.'
Open on a real moment. 'I did a 90 min Zoom with my dad and his friend (both in their 70s...).'
The through-line is that an Emily hook makes a small promise the reader can judge in a second, a number to check, a question to answer, a verdict to argue with. For the mechanics of writing openers like these, our guide to writing LinkedIn hooks goes deeper, and you can pressure-test your own first line in the free hook generator.
Her top hooks, by the numbers
| Hook type | Opening line | Reactions |
|---|---|---|
| Deep-dive receipt | 'I just spent ~100 hours over 2 weeks in Claude Code...' | 631 |
| The naming question | 'Every marketing leader I advise right now is asking the same question:' | 370 |
| Milestone in public | '74,950 subscribers to MKT1 newsletter.' | 359 |
| The naming take | 'Claude Code is a confusing name. It kind of alienates marketers.' | 303 |
A voice that teaches like a sharp colleague, not a guru
It reads like the smartest marketer you know narrating her workday: a number, a chart, one usable system, and a real question at the end.
- Opens on a number, a chart, or a real question. The first line gives the reader a reason to stop.
- Teaches one system. A named role, a step list, a template, or a workflow, not a mood.
- Names names. Tools, collaborators, and sources by name, which adds proof and goodwill.
- Learns out loud. She documents the tool she is still figuring out, hitting the plan limit and all.
- Short lines, playful glyphs. Emoji section markers and generous white space, built for the phone.
- Self-deprecating asides. A PS or a 'wrong answers only' aside keeps expertise from tipping into lecture.
The recognizable devices are small: emoji as section headers to make a dense breakdown skimmable, an open invitation to be corrected ('correct me if I'm wrong in this table'), and self-deprecating closers ('And now I need a nap'). They all signal a real person thinking out loud, which is precisely what earns the replies. She is not performing authority, she is sharing the workings, and the audience rewards the transparency with conversation.
What she does, and doesn't, do
- Lead with a proprietary number or a real screenshot
- Ask a genuine question and mean it
- Name the tools, the people, and the sources
- Hand over a usable system every post
- Show the messy middle, mistakes included
- Post a vague 'AI changes everything' take
- Ask a question just to game the algorithm
- Hide the source or round off the data
- Leave the reader with nothing to apply
- Perform the polished, finished expert
Holding that voice across experiments, data reports, tool teardowns, and candid questions at four-plus posts a week, while also writing a newsletter and advising, is the part almost nobody sustains, and it is exactly the gap CaptureFlow closes. CaptureFlow is an AI content agent that turns your expertise into weeks of on-brand content for every platform. You capture one idea in 5 minutes (a voice note, a screen recording of a workflow, a chart from your latest research), and CaptureFlow, trained on your voice and your past posts, drafts native content for each channel, a LinkedIn post, an X thread, a carousel, a quote image, an infographic, so holding the cadence never costs you authenticity. See how the AI content agent works.
The systems underneath the posts
Two loops quietly turn 100 teaching posts into subscribers, research, and demand for the MKT1 method.
The newsletter and method funnel
The LinkedIn post is the free sample, the newsletter is the main course, and the MCP server and templates are the product. Each layer feeds the next, so teaching and selling are the same motion, never two separate jobs.
The research-to-content engine
- 1Run original researchPull data on 100 B2B startups with Claude Code for the State of B2B Marketing Report.
- 2Package it as a reportTurn the raw data into a multi-part newsletter and a stack of charts.
- 3Slice each finding into a postOne stat, one chart, one takeaway per LinkedIn post.
- 4The comments become researchReplies surface the next questions, objections, and examples.
- 5The next report gets sharperAudience input feeds the following study, one cycle bigger.
Choosing the media
A chart or stat from her own research, the format that reaches furthest.
A screenshot of the tool or homepage she is breaking down, annotated.
A short screen recording or unboxing clip of a workflow actually running.
Plain text, one real question, built to pull replies from the feed.
A simple table or diagram that makes a confusing topic finally click.
A personal scene or confession, no CTA, to keep the account human.
This research-led model is a close cousin of the audience-first approach we mapped in the Amanda Natividad playbook, and it is the template most marketing teams should study: do the work in public, pull the data yourself, and let the frameworks you publish become the product you sell.
Your 30-day challenge
Run the playbook for a month. Turn your own frameworks and research into daily teaching, then start the conversation.
- Days 1-2: List the frameworks, templates, and workflows you already use with clients
- Days 3-4: Teach your best one in full, with a chart or a step list
- Days 5-7: Turn a repeat task you do into a named, reusable system
- Days 8-10: Pull a small dataset only you could gather (your accounts, your tools, your niche)
- Days 11-12: Post one surprising stat from it as a chart
- Days 13-14: Break down a tool or trend you are actually using
- Days 15-17: Post once a day on weekdays, no weekend required
- Days 18-19: Ask one genuine question you truly want answered
- Days 20-21: Share one candid, human moment with no CTA
- Days 22-24: Reply to every comment to feed the ratio
- Days 25-27: Package your best posts into one longer resource
- Days 28-30: Review analytics and double down on the format that reached furthest
Want the cadence without writing every post from scratch? See pricing to start turning your research and frameworks into weeks of content.
The metrics to track weekly
| Metric | Benchmark to aim for |
|---|---|
| Posts per week | 4+ |
| Comments per post | 40+ |
| Comment-to-reaction ratio | 20%+ |
| Reactions per post | 100+ |
| Original data points published | 1+ |
| Named systems or frameworks | 4+ |
The takeaways
- 01Build in public as a systems-builder, not a pundit. Emily's biggest posts are original research and hands-on AI experiments a marketer can copy, not opinions.
- 02Optimize for comments, not just likes. She earns a 37.6% comment-to-reaction ratio, roughly six times the LinkedIn norm.
- 03Teach the tool you are learning. She narrates Claude Code, MCP servers, and AEO in real time, mistakes left in, so the audience learns alongside her.
- 04Hand over a system every post: a chart, a step list, a named role, or a template the reader can use that week.
- 05Post on weekdays, on repeat. About 4.1 posts a week, Wednesday and Monday heaviest, with just 3 of 100 posts on a weekend.
- 06Batch-capture your research and frameworks so a four-a-week cadence survives a heavy client week.
Frequently asked questions
- How did Emily Kramer grow her LinkedIn following?
- By teaching the B2B marketing frameworks, original research, and AI workflows she uses at MKT1, about 4 times a week, and asking real questions that pull replies. Across 100 recent posts she averaged 109 reactions and 41 comments each, and her account passed 51K followers.
- What kind of post performs best for Emily Kramer?
- Hands-on AI experiments and proprietary data. Her top post, a roughly 100-hour Claude Code field guide for marketers, earned 631 reactions, and a 'Marketing Engineer' role breakdown earned 370.
- How often does Emily Kramer post, and when?
- About 4.1 times a week across the 100 posts we analyzed, almost entirely on weekdays, with Wednesday and Monday heaviest and only 3 of 100 posts landing on a weekend.
- How do you apply this playbook without spending hours a week?
- Batch-capture your frameworks and research, then let a content agent draft in your voice. CaptureFlow turns one 5-minute capture into a week of native posts across platforms, so you can hold the cadence without writing every post from scratch.
