Conference Presentation, Keynote
Building Product, Talking to Users, and Growing with Adora Cheung (How to Start a Startup 2014: 4)
- Founders are advised to dedicate one or two full days per week to idea development rather than fragmented daily efforts, while immersing themselves in an industry for one to two months to identify inefficient processes and validate problems before building.
- Validating a problem through direct conversations is considered essential to ensure founder passion and product relevance, with specific recommendations to work in the target service sector (e.g., as a waiter or painter) to gain necessary expertise.
- A minimal viable product (MVP) should represent the smallest feature set required to solve the core problem; skipping the "viable" aspect results in a flat user experience, while delaying launch to perfect the product risks missing valuable user feedback.
- Early-stage growth strategies favor focusing on a single distribution channel for an entire week rather than attempting multiple strategies, with the expectation that channels like Facebook or Google ads require constant iteration and optimization.
- Customer retention data is critical, requiring months to collect, and a successful business is defined by a retention curve that flattens over time with near-zero churn, whereas a low retention curve indicates a fundamentally flawed business model.
- Paid growth and viral mechanics require strict economic discipline; the speaker predicts that waiting longer than 12 months for customer payback is unsafe territory and that businesses risking bankruptcy through unsustainable acquisition costs are not viable.
- Referral programs are most effective when triggered after users have demonstrated high engagement, rather than immediately at sign-up, and mechanics should be tested using formats such as "$10 for $10" or "25 for $25."
- Differentiation from established competitors is best achieved by identifying one or two clearly superior features rather than aggregating many minor improvements, as exemplified by "next day availability" for immediate service needs.
- Relying on purchased users without initial organic feedback leads to temporary traffic that whittles out over time, and waiting for feedback delays the identification of necessary feature changes or removals.
- While imitation is often cheaper and faster than innovation, launching immediately captures early user feedback that cannot be obtained by waiting, unless the startup requires tens of millions of dollars to launch.
- Founders should avoid building features for a potential million users during the growth stage from ten to one hundred users, as premature automation and feature bloat prevent the ability to iterate quickly.
- Performance marketing involves technical creativity to find unique advantages, and as a startup grows and gains capital, it should reconsider revisiting channels like Google Ads that were previously unaffordable.