Conference Presentation, Keynote
Michael Seibel - How to Plan an MVP
- Early user interaction is prioritized over extended research or industry experience, with a predicted build timeline of weeks rather than months for a lean MVP, potentially utilizing minimal tools like a landing page and spreadsheet.
- Most founders are expected to erroneously believe feedback on early iterations is useless, though the outlook suggests that launching a "shitty initial thing" quickly is superior to building a full vision that may ultimately misalign with customer needs.
- While the complete vision typically requires three years, $10 million, and a full team, the strategy advises ignoring all but a small set of initial users' highest-priority problems until later stages.
- Heavy regulatory hurdles in sectors like insurance or banking, as well as complex "hard tech" or biotech projects, are identified as conditions where building a product in a couple of weeks is difficult, though a simple website explaining the offering may still be built in days.
- In scenarios requiring a heavy MVP, there is a prediction that this approach could paradoxically be faster than a lean MVP, whereas standard advice is to time-box specifications to three weeks to prevent scope creep into three months.
- Founders are warned against falling in love with a product and attempting to pivot it to solve other problems when it fails initially, noting that most products never end up looking like their original vision.
- Press launches are predicted to be less valuable than customer interaction, with specific advice to defer press events to focus on building tangible assets for customers to play with, as momentum strengthens significantly once anything is released to the world.
- Examples cited include Airbnb's 2008 launch without payments or map views, Twitch's origin as a low-resolution single-channel stream, and Stripe's initial reliance on manual office integrations, illustrating that early versions often lack the "full thing."
- Spending time on pitch decks is deemed less valuable than creating buildable items for customers, with the assertion that no amount of research can confirm viability until a product is tested in the real world.