Zig When Others Zag: Building HubSpot with Contrarian Conviction
Dharmesh Shah
Co-founder & CTO, HubSpot
LENNY'S PODCAST
The System
LPM: The Science of Being Funny
"I have custom software that I've written that will say, 'Okay, here are the points at which the audience laughed.' Audibly laughed. That's the only way it counts."
Measure what matters: LPM (laughs per minute) is quantifiable and trackable
Standup comedians: 4+ LPM; top TED talks: ~1.5 LPM; business talks: very low baseline
Two ways to improve LPM: add more laughs OR tighten the words between laughs
Record, transcribe, analyze every talk—then iterate
The Craft
Dharmesh's Humor Framework
Lesson 1: End with the punchline
The funny bit must be the literal last words. Pause immediately after. Give the audience half a second to react and permission to laugh. Continue talking = awkward.
Lesson 2: Stack humor in one story
Amortize the setup cost. Spend 75 words building context, then layer multiple punchlines into the same story. One laugh leads to three.
The engineering approach
Functional decomposition: break public speaking into sub-skills (eye contact, slides, pacing). Master one per year. Treat humor as a learnable skill, not a talent.
Talent vs. skill: Talent is slope; skill is achievable for almost anyone with process + measurement
High-stakes speaking: Hold audience attention, because nothing else matters if you lose it
Annual focus: Year 1, master eye contact. Year 2, slide design. Year 3, humor.
The benchmark: Aim for 1.2–1.5 LPM in business talks to be in top decile
SoloWare philosophy: Build tools only for yourself; UI for one person requires zero user testing
Company Building
Zig When They Zag: The HubSpot Contrarian Playbook
The no-reports bet: Never had direct reports at HubSpot (7,000+ people). Lean into strengths; don't become passively okay at things you dislike.
The all-in-one paradox: Launched with SEO, web analytics, blogging, CMS—in year one. Everyone said focus on one thing. We solved the actual customer problem instead.
The SMB moat: SMB was 100x harder than enterprise in early days. No one had succeeded. Once you crack it, reverse gravity stops pulling you upmarket.
Heuristic for depth: If you're top-3 in one category, you're over-indexed. The value is the integration, not dominance in one feature.
The anti-advice principleBest advice: "Focus on one thing and be world-class." Our zig: do the exact opposite. If you can't articulate why you're zigging, don't zig. But always ask the question.
The SMB advantageMillions of customers (no concentration risk) + short feedback loops + control of roadmap + can experiment freely. Enterprise can't iterate like that.
Philosophy
Lean Into Your Strengths
Don't manage if you're not exceptional at it—even with coaching, you'll be passively okay, and that wastes years
Fall in love with the problem, not the solution. This forces you to define what problem you're actually solving.
High-conviction, low-consensus bets: understand the conventional wisdom, then ask if there's an alternate path worth exploring
Measure each metric; course-correct based on data not ego. Did we succeed at what we promised?
On founder decision-makingAsk yourself at every stage: "Have we considered the opposite path?" You don't have to zig, but you should know what the zig would've been.
Contrarian Wisdom
What Dharmesh Got Right (When Others Were Wrong)
✗Startups must focus narrowly on one thingINSTEAD →✓ If the customer problem requires breadth, go broad. Integration is the moat, not specialization.
✗SMB is a dead-end market for startupsINSTEAD →✓ SMB is the best asymmetric bet if you can survive it. Millions of customers, short loops, you own your roadmap.
✗Founders should manage—it's part of the jobINSTEAD →✓ If you're not world-class at management, don't do it. Your comparative advantage lies elsewhere.
✗You need to memorize your talk and wing the humorINSTEAD →✓ Humor is a measurable craft. Record, transcribe, track laughs, iterate. Treat it like product development.