The $825M Case for Founder-Led Sales

On My Mind

Hi — Asad here!

Man, building in the age of AI is humbling. To do it right, you have to be willing to do the opposite of what worked before. The opposite of what your brain tells you is right. We spend our careers developing perspective, and now most of it is as useful as month-old milk. I’ve had two of these moments this year, and I have a feeling many more are in-store.

The first was earlier this year, when Kyle Norton kindly informed me that the decentralized approach to AI — let everyone experiment, surface the cool stuff — is a dead end. Nothing gets to production. ROI lives downstream of centralized deployment. I wanted ROI. So I had to flip how we were doing it at our firm.

The second was last week, on our pod, when we had Jordan Crawford, a.k.a., the best GTM engineer I know. Jordan’s new thesis is that the way to build GTM Machines today is the exact opposite of how we built them before. We used to start with strategy, build systems, then pick tools. Now? Start with the tools. WHAT? WHY ARE WE DOING THE OPPOSITE?

Because AI tools are powerful but “jagged” — brilliant at some things, hopeless at others (which, incidentally, is also the perspective of Eric Janssen, whose editorial you’ll read below). You have to know their edges before you can build around them. Start with strategy and you’ll architect systems the tools can’t actually deliver on, whereas you start with the tools and you get what everyone’s chasing: speed and quality. The other way gives you one.

I want both. I’m sure you do, too.

Now, onto Eric Janssen. Enjoy!

The Myth of Graduating Early

Conventional wisdom says hire a sales team as fast as you can. Delegate, scale, graduate out of the sales seat. But that playbook was written for a different era.

AI tooling and evolving GTM models are producing a structural shift, enabling smaller teams — including solo founders — to build and scale more effectively than at any point in history. The founder who stays embedded in the selling function builds a customer intelligence edge that feeds every AI tool, every positioning decision, and every product bet the company will ever make.

In an AI-accelerated market, it may be the last durable edge that can’t be bought or copied.

The $825M Acquisition That Started at a Hotel Pool

At a conference in Hawaii in 2017, Rahul Vohra struck up a poolside conversation with a stranger, who turned out to be Shishir Mehrotra, a former Microsoft Outlook veteran and ex-CPO of YouTube. For the next two hours, Vohra did the same thing he’d done with every Superhuman user in the early days of growing the business: a full concierge onboarding. Shoulder to shoulder, watching Mehrotra send emails, understanding his workflow, demoing the product, and sharing his convictions about AI-native potential. What could have ended as just another user signup became an investment. Years later, when Mehrotra became CEO of Grammarly, he led the $825M acquisition of Superhuman.

This is exactly the kind of conversation that can’t be automated, outsourced, or prompted into existence. Because, it wasn’t the product that turned an afternoon margarita into an $825M outcome. Vohra’s point of view on where AI-powered productivity was headed, and his conviction that he was the one to build it, were the clinchers — something that couldn’t be delivered convincingly by anyone or anything else.

That story is a preview of how the most valuable companies of this decade will be built, and why founders who delegate sales too early surrender the one advantage that no hire and no AI can replicate.

But first, some context on why this matters now.

The Biggest Entrepreneurial Wave in History

Entrepreneurial activity is at an all-time high and official statistics understate it: While EIN filings have run above 5 million annually since 2021 (roughly 60% above pre-pandemic norms), that measurement framework was built for an era when incorporation preceded revenue. Today, the sequence is inverted: founders build first, charge customers second, and file paperwork third, if at all.

The signals that actually matter tell the story faster, especially when you look across four layers that traditional statistics were never designed to capture:

• In the build layer: Cursor grew from $100M to $2B ARR in 14 months. Lovable hit $400M ARR in under a year. Replit grew 2,352% in a single year, and 75% of its users never write a line of code. Most tellingly, roughly 60% of Vercel’s v0 usage happens on nights and weekends, signalling millions of people building potential businesses long before they call themselves founders.

• In the stake-a-claim layer: Verisign reported 386.9 million registered domains by Q4 2025, up 22.7 million in a year, with new generic TLDs surging nearly 30%.

• In the commerce layer: Stripe processed $1.9T in payment volume across 5 million businesses in 2025, up 34% YOY.

• And, in the aspiration layer: Y Combinator doubled their application rates from two years prior, with 27,000+ received for a single batch and an acceptance rate below 1%. Meanwhile, “Founder” is now the ninth fastest-growing job title in the U.S., with LinkedIn additions growing 60% YOY and nearly tripling since 2022.

The lesson: The starting gun fired long before the official count began.

Starting and Staying Intentionally Small

The biggest shift isn’t just how many people are starting companies but more so, how few they’re building them with.

Where founders in 2022 boasted about money raised, headcount, and office space, today’s aspire to be on the Lean AI Leaderboard: companies generating $5M+ in revenue with <50 employees. Solo founder share has doubled over the past decade, jumping from 31% to 36% in 2024–25. Telegram generates $33M in revenue per employee. Midjourney is one of the most-used AI products on the planet, and it has <15 people on staff. And we’ve just witnessed the first one-person $1B company. These aren’t founders who couldn’t afford to hire. They’re evidence of a fundamental shift: companies starting smaller, staying smaller longer, and still producing outsized results.

None of this makes the founder’s job easier. When building is cheap, every competitor can ship a comparable product in a weekend. When distribution is automated, every inbox is flooded with AI-generated outreach that sounds identical. The barriers to starting have never been lower, while the barriers to standing out have never been higher.

The advantage, then, is judgment: the human quality that accumulates through direct customer contact, conversations that don’t scale, and the friction of learning firsthand why people buy or don’t. That judgment can’t be hired or automated. It can only be built, and it’s best built in the sales seat.

5 Reasons Founders Must Stay in the Sales Seat

1. The founder is the only one who can sell what doesn’t yet exist.

At the point of first sale, a startup has no brand, no track record, no reference customers, and often, no finished product. Instead of evaluating the product, the buyer’s evaluating the person asking them to believe in it.

What’s new is that everything around the founder has been commoditized, and the external signals that used to prove a company was real have been devalued to the point of meaninglessness. A polished website, a clean demo, a pitch deck, a case study — any founder can generate all of it with the AI platform de jour. The more synthetic the top of the funnel becomes, the more weight buyers place on the one signal that hasn’t been automated: a direct encounter with the person responsible for the company.

A hired salesperson can demo, explain features, handle objections. What they can’t do is commit the company. When a buyer says “we’d need X for this to work,” a salesperson routes the request and hopes. The founder says “yes” and ships it by Friday.

A salesperson also can’t credibly say “no.” Every AI sales agent is engineered to move buyers toward yes, and buyers have learned to discount it. When a founder says “honestly, we’re not the right fit for you,” it’s the most credible signal a buyer can receive, because no coin-operated rep or bot would ever say it.

Only the founder can answer the questions that actually drive early-stage purchases: Where is this going? What do you believe that your competitors don’t? Will you still be here in 18 months when something breaks?

Authority. Credibility. Knowledge. The founder has all three.

2. Story is the product.

AI has flooded every format with noise — feature lists, case studies, testimonials, comparison charts — all arriving at industrial scale, training buyers to skim past them. What hasn’t been flooded is first-person narrative. You can automate a case study, but you can’t machine-generate a credible founder story. The format itself is proof of humanity.

Features converge. Pricing converges. Positioning converges. Every company in a category now uses the same AI tools to produce the same polished outputs. What doesn’t converge is the specific, first-person account of why the founder built the thing, and what they learned from the first people who used it. AI can generate case studies, optimize messaging, and personalize outreach at scale. It can’t manufacture the signal that triggers trust. A sales rep can recite the founder’s story. Only the founder can be the story.

3. The infrastructure excuse is gone.

Founders once had a legitimate reason to hire salespeople early — running a GTM motion genuinely required a team. In 2015, the standard startup sales stack meant a Head of Sales ($200K), SDR ($80K), marketing person ($120K), and sales ops specialist ($100K) — roughly $500K in comp, plus Salesforce, Marketo, Outreach, LinkedIn Sales Navigator, and a half-dozen data vendors. Total setup cost: north of $700K. Today that entire stack runs a founder about $500 a month. The pyramid that used to sit underneath a professional salesperson has collapsed into a single founder armed with AI.

Seed-stage companies now routinely hit $1M-$2M ARR with the founder as the only salesperson, a copilot like Clay or Attio on the operational layer, and a fractional SDR for specific campaigns. The fundraising market has noticed: investors now explicitly ask founders why they’re hiring a sales leader early, rather than assuming they should.

4. Founder-led sales is not an interim phase.

The old playbook treated founder-led sales as something to endure until you could afford to hire pros. AI has inverted that. A founder in 2026 can run targeting, prospecting, outbound, content, and pipeline with tooling that used to require four hires. The question is no longer whether the founder can do it alone — it’s why any founder would delegate the highest-learning activity in the company to someone who wasn’t there for the first hundred calls, rejections, and road-map shifting buyer conversations.

Maor Shlomo built Base44 to 400,000 users and an $80M acquisition with eight people, and no marketing budget, sales team or outbound sequences. Just a founder posting his journey publicly on LinkedIn as the entire GTM engine. The absence of a sales team was the point: every post was a one-to-many sales call, delivered by the one person the market actually wanted to hear from.

The objection that founders don’t scale is a legacy of the SDR-AE-CSM pyramid — a structure built for a world where human labor was the only way to deliver human attention at scale. AI has collapsed that constraint. The founder’s job isn’t to personally make every call. It’s to set the voice the AI is trained on and the judgment the team is calibrated against. Everything downstream, including the outbound, the content, and the pipeline, gets its signal from whoever is closest to the customer. If that’s the founder, the signal is sharp. If it’s a BDR six months in, the signal is generic.

5. Selling is the learning loop that AI amplifies.

Every sales conversation a founder runs is simultaneously five things: a pipeline event, a product research session, a positioning refinement exercise, a market validation data point, and a contribution to the institutional knowledge that determines whether the company can ever scale beyond them. Vohra understood this at Superhuman. He personally onboarded thousands of early users — not just to close business, but to learn the specific words customers used to describe the pain, the moments of delight and frustration, the real-time response to pricing, and the competitive landmines hiding in every deal. That knowledge became the foundation for Superhuman’s positioning, pricing, product roadmap, and sales playbook. The argument isn’t that founders should run onboarding calls forever, but rather, that they should do it long enough to build what cannot be delegated.

An AI version of a salesperson is only as good as the voice-of-customer data you train it on. Synthetic data in, synthetic insights out. A founder who has run 50 discovery calls has customer insight that an SDR hired two quarters ago simply doesn’t, and no AI can fabricate that dataset after the fact.

Stay in the Seat

For 15 years, founder-led sales was treated as a phase. Scrappy, necessary, temporary — something to be endured until you could afford a professional, then graduate away from. The whole point was to get out of the sales seat.

That chapter is over. In a market where AI has commoditized every execution layer of GTM, the only durable competitive advantage is the one the founder builds personally: a deep, lived understanding of the customer that compounds with every conversation and calibrates every AI tool the company will ever deploy.

The founders who define this decade won’t be the ones who build the biggest teams or raise the most money. They’ll be the ones who stay closest to the customer longest — and use AI to scale that closeness rather than replace it.

Subscribe now

Eric Janssen is a full-time faculty member at the Ivey Business School at Western University, where he serves as director of the Executive Education Sales Program and teaches sales and entrepreneurship. A multi-time founder and early-stage team member in venture-backed B2B and B2C startups, he helps founders and leaders grow revenue and build scalable sales engines through his course, The Founder Sales Sprint.