Rivalry vs. Competition

Two ways to use the same adversary

Tobias Lutke draws a distinction between using an adversary as competition and using one as rivalry, and argues the same opponent produces opposite outcomes depending on which mode you are in.¹ In competition mode, the goal is to defeat the other party, and the behavior it produces is reactive: tracking what a competitor does and responding to it. Lutke observes that in many companies the most active Slack channel is the competitive-analysis channel, where people post everything a competitor does, and that the result is a company copying features and chasing a rival's roadmap instead of its own.¹

In rivalry mode, the goal is to use the other party to become better, which Lutke frames as positive-sum. His example is Andre Agassi and Pete Sampras: Agassi has said he could not have become Agassi without Sampras there to raise the standard. Lutke notes the detail that Agassi's internal picture of Sampras was partly his own construction, since Sampras was not thinking about Agassi the same way, and that the rivalry still worked as a motivational structure. He cites Michael Jordan's admission in The Last Dance that he may have fabricated a slight to fuel himself, and argues the fabrication did not matter because what mattered was the standard it created.¹

The mimicry trap

Lutke's claim about competition mode is that it produces mimicry masquerading as differentiation. He draws on art-school training, where students copy great works, and observes that your next painting will not match a Van Gogh, that a copy is at best seven tenths of the original, and that copying is not a route to excellence.¹ Applied to startups, the competitive-analysis channel is the institutional form of this: if the clearest signal about what to build comes from what a competitor is building, a company is running a delayed copy of someone else's product strategy. The difference Lutke isolates is the feedback source. Competition takes the competitor's moves as the signal; rivalry takes the competitor's level as calibration while the operative question stays "what can we do that they can't?" He ties this to a general preference for differentiation, invoking James Dyson's line about making something different even if it is worse, which he says only makes sense in rivalry mode.¹

The competitive-analysis channel is the corporate instance of a far older default, and the general theory of that default belongs to Peter Thiel, whose account of imitation as the condition of human sociality, and of the bubbles, credential arms races, and category compression it produces, is set out in the file on mimetic competition. The two answers do not reduce to each other: Thiel's counter is epistemic, seeing what the imitating crowd cannot, while Lutke's is motivational, raising the standard one is chasing until copying stops being worth the trouble.

Rivalry as an outward distribution device

Eric Glyman adds a use Lutke's framing does not cover: rivalry pointed outward as a go-to-market and attention device rather than an internal motivation device.² Glyman notes that people like rivalries, citing Coke against Pepsi and Uber against Lyft, and describes how Ramp, though focused competitively on Amex, was happy to be compared to Brex by the startup world. "We were very happy to drive a contrast and be mentioned every time someone else was mentioned. We got really slingshotted up very quickly."²

Where Lutke's rivalry faces inward, calibrating an internal standard and working even when the picture of the opponent is inaccurate, Glyman's faces outward, deliberately picking a public antagonist so a smaller company rides a larger one's attention. The two are compatible: a public rival can both set an internal bar and place a company in the market conversation. The shared logic in both founders' accounts is that a rival used well beats undifferentiated competition; the lever differs, with Lutke's operating on motivation and Glyman's on salience, and Glyman's version working even when the company is far smaller than the rival it names.

Open questions

The Agassi and Sampras frame works cleanly in individual sports with an unambiguous measure, and is harder to operationalize in markets where success metrics are vague, since rivalry against an unclear standard loses its calibrating function. The fabricated-rivalry point raises a further tension both founders' examples surface: the motivational benefit seems to hold whether or not the picture of the opponent is accurate, but strategic decisions made on an inaccurate read of a competitor carry real risk. Glyman's outward version has its own failure mode, flattering a bigger rival or attaching to a negative brand.