The substrate is always rented

Seven years of the same rules, not seven years of genius

Mercedes won every constructors’ title from 2014 to 2020. The standard read is dominance built on engineering brilliance, and the engineering was real. The deeper fact is that the rules barely moved for seven years. The V6 hybrid formula introduced in 2014 stayed structurally intact through the back half of the decade. Mercedes built a development loop, each season’s learning feeding the next, and that loop compounded because nothing underneath it moved.

That is the whole story of compounding advantage, and it has almost nothing to do with genius. A loop compounds only when the surface it runs on holds still. Change the surface and the loop resets to zero, no matter how good it was. The strategic question every operator eventually has to answer is not whether they have a loop. It is how long the ground under the loop stays flat, and what happens on the day it doesn’t.

You don’t own the ground you build on

Mercedes never owned the rules. They rented them from the FIA, which writes and revises the technical regulations on its own schedule for its own reasons. Every advantage the team built was leased against a rulebook someone else controlled.

This is the general condition, not the exception. F1 teams rent the regulations. SEO operators rent ranking behavior from Google. App developers rent distribution and economics from Apple. Fintech companies rent settlement and interchange from the card networks. In each case the operator builds a loop (content that compounds, an install base that grows, a cost structure that improves) on a substrate governed by a party whose incentives are not the operator’s.

Byrne Hobart has written about platform dependency as a structural risk that founders systematically underprice, and the underpricing has a specific shape. Operators model the loop carefully and treat the substrate as a constant. It isn’t a constant. It’s a lease, and the landlord can rewrite the terms. The lease looks free right up until the reset, and then the entire cost shows up at once.

Stability is the actual gift

Here is the part most operators get backward. They treat the substrate as neutral infrastructure and put all their attention on the loop. The substrate’s most valuable property is not its features. It’s how long it holds still.

Long rule cycles create the conditions for depth. When the surface stays flat for years, the operator who invests in second-order and third-order optimization gets to keep collecting on that investment. Mercedes could pour resources into marginal aerodynamic and power-unit gains precisely because those gains stayed relevant season after season. Depth was rewarded because depth had time to pay.

Frequent resets invert the reward. When the substrate changes often, depth gets wiped before it compounds, and the advantage shifts to whoever rebuilds fastest. Rebuild speed is usually a function of cash. The team or company that can throw the most resources at re-solving the problem from scratch wins the transition. Stability rewards intelligence and patience. Instability rewards balance-sheet size. Every operator should know which regime they’re in, because the two demand opposite strategies.

This is why SEO operators are living through a hard decade. Google’s ranking behavior was stable enough through the 2010s that content depth compounded. AI Overviews and the shift toward answer-in-the-SERP changed the substrate, and a lot of carefully built loops reset at once. The content wasn’t worse. The ground moved.

Governance never optimizes for your stability

The uncomfortable truth is that the party controlling your substrate has no reason to keep it stable for you. Governance bodies optimize for their own legitimacy, participation, and narrative. Product quality on top of the platform is downstream of those goals, and often it’s not even a consideration.

The FIA changes regulations to keep the racing competitive and the sport marketable, not to protect the loop any single team has built. If one team dominates for too long, that’s a legitimacy problem for the FIA, and the fix is a reset that destroys the dominant team’s advantage on purpose. Google tunes ranking to protect its own position against the possibility that users get answers elsewhere, not to preserve any publisher’s traffic. Apple revises App Store policy to manage regulatory pressure and its own take rate, not to protect the developer’s margin.

Read that pattern honestly and the conclusion is sharp. The healthier the governing body’s incentive to reset, the more certain your reset becomes. A substrate that has been stable for a long time is not a substrate that will stay stable. It’s often a substrate under mounting pressure to change, because long dominance by anyone is exactly what governance bodies are built to disrupt.

Innovation from reset is not innovation from depth

The strongest objection to all of this is that rule changes drive innovation. F1’s 2026 regulations will force teams to invent, and the sport will get better designs out of the churn. There’s truth in it, and resets do produce innovation. But innovation-from-reset and innovation-from-depth are different things, and conflating them hides the cost. Reset innovation is broad and shallow: everyone re-solves the same freshly opened problem, and the solutions converge fast because the search space is new and unexplored. Depth innovation is narrow and deep: one operator pushes a stable problem further than anyone else can, because they’ve had years the competition hasn’t. The industries that produce durable, compounding excellence run on depth. Frequent resets trade that away for the appearance of dynamism. You get more churn and call it more progress.

The question is how long, and then what

The strategic question is not whether you have a loop. Loops are common and most operators can build one. The question is how long your substrate stays stable, and what your plan is for the day it isn’t.

Answering it means naming the party that governs your substrate and reading their incentives instead of your own. It means estimating the length of the current rule cycle and how far into it you are. And it means building the reset plan before the reset, because the operators who survive substrate changes are the ones who saw the lease for what it was while it still looked free.

Own the loop, yes. But know that you’re renting the ground it runs on, and price the lease honestly, including the balloon payment that comes due on day one of the next regime.