Do Things That Don't Scale

“Do things that don’t scale” means deliberately using manual, high‑touch, inefficient tactics early on to win your first users and learn fast, before you automate and optimize. The goal is to reach product‑market fit quickly by doing whatever it takes—personally—so you know exactly what to later systematize.

Why it works: You get rich user feedback fast, create delight that builds word‑of‑mouth, and uncover the real “jobs to be done” without waiting on code, funnels, or ads.

Classic examples:

• Stripe: Founders personally visited startups and installed the API on their laptops, removing every adoption hurdle.
• Airbnb: Offered free professional photography to hosts to improve listing quality and bookings.
• Superhuman: Founder‑led, 1:1 onboarding calls to tailor setup and observe friction live.
• DoorDash/Uber (early): Manually fulfilled orders/rides in small areas to validate demand and operations.
• Dropbox: Used a simple explainer video to test demand before building complex sync tech.

Concrete ways to apply this now:

• Personally onboard your first 20–50 users; set them up, sit in on first use, and fix issues live.
• Run a “concierge MVP”: deliver the value by hand (spreadsheets, email, manual steps) while you learn the exact workflow to later automate.
• Do founder‑led sales and support: customized demos, tailored proposals, ultra‑responsive help that turns skeptics into champions.
• Offer “unscalable” perks that remove friction: free data migration, custom integrations, hands‑on training.

How to know when to stop: Once patterns repeat (same questions, same fixes), document and replace the manual steps with product, playbooks, and automation; shift from founder‑led to scalable processes without losing the learned quality.

Pitfalls to avoid: Don’t promise bespoke features you can’t sustain; capture learnings in the product quickly; keep scope tight (narrow segment, one city, specific use case) so “unscalable” effort stays focused and high‑impact.