Do Things That Don’t Scale

Paul Graham’s essay Do Things that Don’t Scale is perhaps the most practically useful essay for early-stage founders. It demolishes the myth that great products naturally attract users.

The Core Insight

Successful startups require founders to perform labor-intensive, non-scalable activities initially. You can’t just build something and wait for users to come. You have to go get them—manually, one by one.

Recruit Users Manually

Compound growth creates exponential returns. Acquiring just 10% more users weekly yields millions within two years. But that growth has to start somewhere.

Examples:

• Stripe’s “Collison installation”: The founders would offer to set up Stripe for you right then and there, rather than just giving you a link
• Airbnb: Founders went door-to-door in New York with luggage, personally photographing listings and improving them

Delight Your Customers

Provide exceptional service that seems impossible to scale. Wufoo sent handwritten thank-you notes to users. The goal is to create “insanely great” user experiences, not just products.

This isn’t just a growth hack—it builds company culture permanently. How you treat your first users shapes how you’ll treat all future users.

Start Narrow

Focus on narrow segments to achieve critical mass quickly, then expand. Facebook started at Harvard only. This concentrated focus creates intensity that broad launches can’t match.

What Doesn’t Work

Big launches and partnership deals rarely drive sustainable growth. Success depends on gradually building loyal users through attentiveness, not initial velocity.

My Takeaway

Treat building a startup as a vector with two components: product innovation AND the unglamorous, labor-intensive work required to launch. The second part is often more important early on, yet it’s the part most founders want to skip.

What unscalable things have you done to get started? I’d love to hear at persdre@gmail.com.