Connection note reply rates on LinkedIn fell to 2.2% by early 2026, according to Expandi's benchmark data. Most reps responded by sending more messages. The founders still booking meetings took a different approach: they made sure buyers already recognized their name before any message arrived.
Finn Thormeier has been building that system for B2B companies for five years. His agency, Project 33, runs LinkedIn content programs for more than 50 B2B tech companies, and he hosts the Executive Brand podcast (140+ interviews with operators). He joined Extrovert to walk through the full playbook: who should post, how to extract content from an executive, and how to turn attention into revenue.
You can watch the full session replay or read the summary below.
Why does founder-led marketing outperform branded content?
A brand page publishing the same content as the founder's profile will lose on reach every time. LinkedIn's algorithm heavily weights personal profiles over company pages, so distribution alone favors the individual.
But reach is not the main point. A CEO talking to another senior decision-maker reads like a peer conversation, not a pitch. Buyers skim past company updates; they stop on posts from people who clearly know what they are talking about from experience. In a feed full of AI-generated summaries, someone with a real opinion and a specific example is genuinely unusual.
Who should actually post? (It is not always the founder)
Finn's rule for choosing who owns the content program has two requirements, and both are non-negotiable:
1. They genuinely want to do it. Not "willing to try" or "open to it if the team sets everything up", they actually want to share their knowledge publicly. You can lead a horse to the water, but you cannot make them drink. Finn has seen enough VP-of-Marketing-driven programs where the CEO was skeptical to know that an unwilling executive will kill the program before it has a chance to work.
2. They have insights the ICP cares about. This is why it is not automatically the founder.
At Siteimprove, the founder had left the company, so Project 33 worked with the CMO, who sells to other CMOs, making his content naturally resonant with the audience. At Safetywing (remote insurance), the head of revenue was a digital nomad constantly traveling and working across time zones, which made her a power user of the very product they sold. Her content had built-in credibility with the ICP even though she was not in the typical "founder" role.
For enterprise deals with large buying committees, CFO approval, legal review, procurement sign-off, IT evaluation, Finn's approach expands to match. Having your CFO, head of legal, and procurement leader also creating content means you are building relationships with the exact roles who block or greenlight deals.
For smaller companies: start with whoever genuinely wants to do this and has something worth saying. If that is the founder, great. If an employee volunteers, support them. Never force it.
The extraction interview: how Project 33 builds the content engine
The core of Finn's system is a biweekly interview, roughly 45-60 minutes, that runs like an internal podcast. The critical difference from a regular podcast: the audience is not the interviewer. It is the ICP.
The person running the interview needs three things before sitting down:
- Context on the executive (their background, what they have built, what they care about)
- Context on the company (what you sell, how it is different, where you are winning and losing)
- Context on the ICP (what they worry about, what they are trying to get done, what would make them push back)
Armed with that context, the interviewer pulls for content that would be interesting to a buyer, not just content that the executive finds interesting to talk about.
When the interviewer does not know the ICP deeply, which happens with niche markets like supply chain at large manufacturers, Finn uses a technique he calls the mirror: after the executive answers, follow up with "what would someone in your ICP's position disagree with here?" That question shifts the executive's frame from talking-to-Finn to talking-to-a-supply-chain-director. The answer gets sharper.
Numbers matter on LinkedIn. Finn extracts them on the call wherever possible. When the executive does not have a specific number ready, they get tagged in the Google doc draft: "Can you provide the stat here before we publish?"
Finding the right content angles
The most reliable sources for content ideas that actually resonate with the ICP:
Recent company events. An acquisition, a product launch, a major customer win, a market shift. If a customer just acquired another company, interview the CEO about why, what the vision is, where the synergies go. That is one 45-minute conversation that generates two or three posts of genuine news, the kind of content competitors cannot replicate because they do not have the same story.
What competitors are saying. Finn asks this directly in the interview: "A competitor just said X publicly. Do you agree? Where would you push back?" If a competitor's post performed well, use it as a springboard to get the executive's real reaction. Disagreement generates engagement; consensus generates nothing.
Customer conversation patterns. What objection keeps coming up? What question do prospects ask in every first call? What do customers say after they have been using the product for six months? These patterns live in the executive's head and are invisible to anyone writing generic LinkedIn content from the outside.
The ICP's active problems. Not the executive's strengths, the buyer's current struggles. The content that lands is content that makes the ICP think "this is exactly what I am dealing with right now."
When you do not know the ICP deeply, ask the executive: "What would someone like your ideal customer disagree with in what you just said?" That single follow-up produces more distinctive content than any research report.
Turning content into pipeline: the signal-to-outbound loop
The content part is what everyone talks about. What converts it into pipeline comes after.
Track organic engagement with ICP precision. Project 33 uses Jungler (jungler.ai) to pull organic post engagers via API. Those engagements go into a Clay table, where each profile gets enriched and qualified. The table tracks every company and person who has engaged with the executive's content over time, including how many times and on what date.
Layer in thought leader ad signals. For clients running LinkedIn thought leadership ads, Fibler captures company-level engagement data on the ad side. Because ads run with targeted job titles and account lists, every engagement there is presumed to come from within the buying committee.
Merge into one place and act. Clay deduplicates and scores. When an ICP engages, it pushes a notification to a shared Slack channel: "this CISO just liked your CEO's post." The minimum action is a connection request from the executive: not a pitch, just a connection. If the account is a named target, that signal flows into HubSpot or Salesforce and becomes one input alongside webinar attendance, website visits, and event activity.
Act on the compound score. A target account with seven different people engaging with executive content fifteen times in the last 30 days is hot. That triggers an SDR or BDR to prioritize the account, not because a sequence timer fired but because real buying signals accumulated.
Without the loop, content is just an audience-building exercise.
The engagement metric that actually matters
Finn is direct about the number most people track: total likes tell you nothing useful.
If you sell into chief information security officers and a post gets 30 likes, seven of which are CISOs, that post outperforms one with 60 likes and one CISO. The question is never "how many liked it?" but "did the right people see it?"
| Post | Total likes | ICP likes | Verdict |
|---|---|---|---|
| Post A | 30 | 7 CISOs | Strong signal |
| Post B | 60 | 1 CISO | Weak signal |
This framing changes what you do next. When a post gets 10 likes, most founders feel discouraged. Finn's advice: check who those 10 people are. If two of them are actual prospects, reach out to those two. Amir Utley, co-founder of HockeyStack, built early pipeline this way, posting daily, getting 5-10 likes per post, and manually reaching out to every ICP who engaged. The posts looked low-performing. The pipeline was real.
For founders building this on LinkedIn from scratch, the how to prospect on LinkedIn guide covers the engagement-first mechanics in detail.
Vanity metrics are a distraction. One CISO liking your post and receiving a connection request is worth more than 50 generic likes and no follow-through.
Distribution and consistency: the outlier compounding loop
Some posts will significantly outperform your average. Project 33 calls these outliers. Finn defines the threshold relative to each account's own baseline, not against any absolute number. When a post qualifies and it is evergreen (not a funding announcement, not a dated news reference), repost it verbatim every three months. Keep reposting until it starts performing at average instead of above average. Some posts run on a quarterly cycle for nine months or a year.
| Action | Frequency | Logic |
|---|---|---|
| New content from interview | Biweekly | Fresh signal, new angles |
| Evergreen outlier repost | Every 3 months | Audience turns over; insight stays valid |
| Competitive reaction post | As relevant | Engagement from disagreement |
| Repurpose outlier into video | Quarterly | Reach format that does not see the text version |
Most founders stop reposting something once it feels like they have already said it. That is the wrong instinct: do not stop milking a working piece of content just because you already posted it. Your audience turns over. New people see it. The insight is still valid. Repost it.
Extrovert tracks your prospects and topics, then suggests on-brand comments and DMs from your playbook. Warm, human, ban-safe.
See how it worksAI in the content machine: what works and what produces slop
Finn uses AI. Project 33 has a custom N8N agent built for the drafting step. But the architecture matters.
Raw signal comes first. The interview is the input. Taking the interview transcript and feeding it to an agent with a clear brief, the angle, the hook direction, what to include and exclude from that session, produces output that sounds like the executive. Feeding a prompt like "write an interesting LinkedIn post for a head of supply chain" produces generic content, because the model has nothing real to work with.
One specific technique that changes the output quality: never tell the model to write a LinkedIn post. Tell it to write a mini essay. The "LinkedIn post" frame activates preconceptions: line breaks between every sentence, a hook that asks a question, a P.S. that adds nothing. "Mini essay" skips all of that and the output sounds like a person wrote it.
Human review stays in the loop at every step. The content director runs the interview, gives the agent a brief with the angle and the key excerpts from the transcript, reviews the output, sends it back if needed, and approves before the client sees it. The agent replaces the ghostwriter. The content director still makes every meaningful decision.
The minimum viable founder-led marketing system
If you are a small founder and want to start next week, Finn's checklist:
- Post 2-3 times a week. Block the time, either batch it in a four-hour session once a week or block an hour three days a week. Both work; pick one and keep it.
- Spend 30 minutes a day active on the platform: comment on your prospect list's content, answer your own comments, answer DMs. For SDRs running the same system, LinkedIn for SDRs covers the full workflow.
- If you have any ad budget, run thought leader ads against a narrow list of 200-2,000 target accounts.
- Connect something like Jungler to a Clay table and set up a Slack notification when an ICP engages with your content.
- Act on those notifications: send a connection request or a short message when a named target shows up in your engagers.
Steps 1 and 2 build the presence. Steps 3-5 convert presence into pipeline. Most founders stop at step 2 and wonder why the pipeline is not there.
For founders in more competitive categories, martech, sales tech, Finn is direct: more competition means more effort. You cannot prompt an AI for five minutes and expect to build a meaningful brand in three months in a category where 20 other founders are already established on LinkedIn. In less crowded verticals, even minimal consistent output gives you the field to yourself.
Frequently asked questions about founder-led marketing
What is founder-led marketing?
Founder-led marketing is a go-to-market approach where a founder or senior executive builds a personal brand and audience on LinkedIn to drive awareness, trust, and pipeline. It differs from brand-led marketing in that content comes from a real person with real opinions, and distribution happens through personal network effects rather than company page algorithms.
Does the founder have to be the one posting?
Not necessarily. The right person to own founder-led marketing is whoever genuinely wants to create content and has insights directly relevant to the ICP. At some companies that is the CMO, the head of revenue, or a functional leader whose expertise maps to the buying committee. For early-stage companies, the founder is usually the right default, but only if they are actually willing to commit to it.
How often should an executive post on LinkedIn for founder-led marketing?
Two to three times a week is a workable minimum for building consistent visibility. Consistency matters more than volume: a founder who posts twice a week every week for six months will build more cumulative pipeline than one who posts daily for a month and stops.
How do you turn LinkedIn engagement into sales pipeline?
Track who engages with your content using a tool like Jungler, push that data into Clay, enrich and qualify it, and alert your team when a named target account engages. The minimum action is a connection request from the executive. For more mature programs, LinkedIn engagement becomes one signal in a CRM lead score alongside website visits, webinar attendance, and event activity.
How long does founder-led marketing take to show results?
Most programs start to show qualitative signals within 60-90 days of consistent posting: people mentioning your content at conferences, inbound DMs from prospects. Pipeline attribution typically takes 90-180 days because LinkedIn familiarity has to compound before it converts. The 90-day window is not a bug. It is the moat. Competitors who want quick ROI will quit before it kicks in.
Should you use AI to write executive LinkedIn content?
AI can draft content efficiently when it has real signal to work from: an interview with the executive where they share specific insights, numbers, and opinions. AI without that signal produces generic content that sounds like everyone else. The other key prompt detail: tell the model to write a "mini essay," not a "LinkedIn post." The "LinkedIn post" prompt activates preconceptions about formatting that produce the line-break-heavy output that looks like AI.
What is the most common mistake in founder-led marketing?
Stopping at content creation. Most founders post, see some engagement, and wonder why nothing converts. The conversion mechanism, tracking who engaged, acting on ICP signals, combining content visibility with direct outreach, is what turns awareness into booked meetings. Content without follow-through is a branding exercise, not a pipeline system.
Sources: Expandi: LinkedIn Outreach Benchmarks 2026, Finn Thormeier x Extrovert webinar (2026-06-11), Executive Brand podcast and newsletter
