Tip
Casey Winters talks about Grubhub versus DoorDash: “What I’ve learned now that I’m more senior in my career is during existential threats, it’s like when Nassim Taleb says, ‘The only rational reaction is overreaction.’ Unless you have a real viable reason to assume otherwise, you got to assume the disruptor is right and base your strategy on them playing an optimal game. Grubhub assumed the disruptor was wrong and that it would all play out eventually in their favor, and it clearly didn’t.”
Turns out AI competitive threats work the same way.
A startup launches an AI coding assistant with a radically different model: real-time collaborative AI that edits directly in your codebase, learning your patterns. Their unit economics look terrible—10x your inference costs, burning cash on GPUs. Your product team dismisses it: “Unsustainable burn rate. They’ll run out of money.” Your board agrees: “Stick to our profitable model. They’re overextending.”
You’ve seen this movie before. In 2012, enterprise SaaS companies dismissed Slack: “Chat? We have email. Their unit economics don’t work at enterprise scale.” By 2016, Slack had eaten the collaboration market. In 2008, traditional taxis dismissed Uber: “They’re breaking laws, burning billions in subsidies. Unsustainable.” By 2015, Uber had destroyed the taxi medallion market.
Younger executives see weak unit economics and dismiss the threat intellectually. You know better. You’ve watched enough disruptions over two decades to recognize the pattern: if the market keeps rewarding it (funding rounds, user growth), they’ll find a way to make the economics work. DoorDash had negative margins for 7 years. Then the pandemic hit and they turned profitable overnight.
The mistake isn’t that you can’t predict which threats are real. It’s assuming the disruptor is operating irrationally. Talented people are funding them. Smart customers are choosing them. Assume they see something you don’t. Assume they’ll solve the unit economics. Base your strategy on them playing an optimal game—because if they do and you prepared for them to fail, you’ve lost.
That judgment—knowing when to overreact to competitive threats—comes from watching enough disruptions succeed against “impossible” economics. You’ve seen the pattern: markets that seemed structurally unprofitable (streaming video, ride-sharing, food delivery) that found paths to profitability once they reached scale.
Context
Casey Winters was at Grubhub when DoorDash launched with an “impossible” business model—hiring their own delivery drivers, operating at negative margins, delivering from restaurants without partnerships. Grubhub assumed it was unsustainable.
DoorDash raised billions, expanded selection dramatically, and took the market. For experienced executives evaluating AI competitive threats, this pattern recognition is critical—you’ve watched enough disruptions over decades to know the difference between genuine threats (well-funded, growing fast, solving real pain) versus noise.
That wisdom comes from watching both kinds play out repeatedly.