Do things that don't scale

Executive summary

Unscalable tactics are a startup advantage: personal recruiting, founder-delivered service, duct-tape ops. Use them to learn and to earn champions — give them up only when ripped away, not on a funding milestone.

When to use this page

• Zero to first users.
• Finding product–market fit.
• Turning early customers into advocates.

Test → launch → hack

DoorDash: Interview merchants → afternoon landing page (PDF menus + phone) → founders deliver Pad Thai → Google Docs + Square + Find My Friends. Demand showed up despite ugly UX. Later automate dispatch.

Teespring: Days of free design for tiny non-profit orders — looked terrible from outside; boulder uphill then rolls. Delight with support (founders on tickets); make wrongs right; speed over clean architecture for next order of magnitude only.

Three jobs for unscalable work

1. Find first users — no silver-bullet PPC; founders’ personal time (emails, calls, conferences, street fairs, water bottles at fairs). Prefer paid validation over free forever.
2. Turn users into champions — exceptional experiences; founders on support; proactive outreach to churned users; own social reputation.
3. Find product–market fit — duplicate codebases if needed; crash nightly if necessary; worry about 10→100 not 10→1,000,000.

Manual learning

Doing the work (cleaning, delivering, photographing Airbnb homes) teaches algorithms and product. Email every new customer personally. Capacity constraints (ice cream vs delivery spike) motivate scaling.

When to stop

As long as net-positive, keep doing unscalable things. Smaller competitors can still out-love you if you abandon the advantage too early.

Key takeaways

• Launch experiments fast; founders are the ops stack.
• Handhold early users; optimise for learning speed.
• Scale systems after demand is undeniable.

What to do next

1. Great products for love metrics.
2. Growth and distribution once retention works.