Steve Blank on How AI Can Improve Business Models
A serial entrepreneur, Blank, who helped popularize the notions of lean startups and minimum viable products, says artificial intelligence is building better business models.
Good Morning!
Here are today’s highlights:
In our latest 21 Hats Podcast episode, the owners talk about their best and worst days as entrepreneurs.
Quince, a “dupe site” has learned that you don’t want to make knock-off products too cheap.
Gunner Roofing is reinventing a blue-collar industry by selling roofs online.
Twenty years after one of the biggest ICE raids ever, meat-packing plants still depend on immigrants.
ARTIFICIAL INTELLIGENCE
Steve Blank changed the way we think about startups. Now AI is changing the way Steve Blank thinks about startups: “In 1999, after founding and running eight startups, serial entrepreneur Steve Blank decided it was time to step back and reflect on what it really takes to pursue an early-stage company. His book The Four Steps to the Epiphany, published in 2003, contained the framework that’s become a pervasive logic in how startups are run and funded today—from the fact that small companies aren’t just mini versions of bigger companies to the idea of a ‘minimum viable product.’ ‘I thought the book was going to be a fad,’ Blank said. ‘It just blew me away about what happened to it.’ I recently caught up with Blank at Stanford, where he’s taught for over 15 years, to review what’s gone right and wrong in how his work has been applied and to peer at the AI era through his lean startup lens.”
“If you think about customer development, it depends on human effort: getting out of the building, talking to people, integrating data, deriving insight. It’s essentially the scientific method we’ve been doing for 500 years: hypothesis testing, running that loop where you design an experiment, get some data, and either validate it or modify the hypothesis.”
“We’ve never automated that, or at least not in business. Starting in April at Stanford, we will have students use AI to generate their first business model canvases. We’re going to force them to create artificial personas instead of using commercial data. It doesn’t mean it’s a replacement for human insight, but imagine I can now run a thousand experiments when humans can run 10 experiments.”
“I’ll give you an example with e-commerce. I could purchase demographic and psychographic data about almost anything. Imagine I have a new consumer product; I can now create artificial personas and run AB tests against those personas. In the old days, if I were making a consumer product, I would hand-generate a website and test some fake products there. It’s now possible to have an AI agent create websites. I could create a thousand websites and run a thousand different tests. It’s basically what quants do for financial services—but for startups.”
“I would teach my students that the minute you take money from someone, their business model becomes yours. If you can’t go up to a whiteboard and draw how your VC makes money, you’re going to be the ex-founding CEO. ... It’s why Facebook ended up like Facebook, not a social product that did good. It’s why Juul didn’t end up being a nicotine-cessation product.” READ MORE
Keep reading with a 7-day free trial
Subscribe to The 21 Hats Morning Report to keep reading this post and get 7 days of free access to the full post archives.