Why Incumbents Always Lose the Reshuffle

Kodak, 1996. The top of the world.

90% of the global film market. 145,000 employees. A near-monopoly on a chemistry so specialized that no competitor in 100 years had been able to replicate it.

The company's core skill was film emulsion — the precise chemical formulation of light-sensitive silver halide crystals on plastic. They were the best in the world at it. Nobody was close.

Digital photography arrives. By 2003 it's better than film for most uses. By 2012, Kodak files for bankruptcy.

The interesting thing isn't that Kodak missed digital — they actually invented the digital camera in 1975. The interesting thing is why they couldn't pivot.

Their entire moat was chemistry. The thing that made them dominant for a century became the thing they could not let go of. The dominant skill was the liability.

Blockbuster, 2002. 9,000 stores, $5.9B in revenue. The world's largest video rental chain.

Their skill was the physical store — picking locations, managing inventory across thousands of retail footprints, training staff, designing the late-fee model that printed money.

Netflix in 2002 was a 90-person company mailing DVDs by post.

Streaming arrives. By 2007 it's better than physical for most uses. By 2010, Blockbuster files for bankruptcy.

Again, they saw it coming. They even bought into streaming early. But the cost of their stores — the asset that defined them — couldn't be unwound fast enough. The skill they had built for 20 years became the weight that sank them.

Same pattern.

Sears, 1989. The world's largest retailer. The "Sears, Roebuck and Co. catalog" was the original Amazon — a 1,000-page book of everything you could possibly want, mailed to your house with logistics nobody could match.

Their skill was catalog-scale logistics — warehousing, shipping, returns, customer service at continental scale.

E-commerce arrives. Logistics gets commoditized — UPS and FedEx handle it, then Amazon eats the integration layer. By 2018, Sears files for bankruptcy.

Their dominant skill — the thing they had spent a century perfecting — became a commodity service anyone could buy. The skill wasn't the moat anymore.

Three companies. Three decades. The same shape.

The thing that made you dominant in the old frame is the thing that makes you slow in the new one. Always. Every time.

— From Reshuffle, Chapter 4

Right now, in 2026, an AI reframe is underway. The honest question for any incumbent's board is the one nobody wants to ask:

What is your "chemistry"? Which of your celebrated capabilities — the one your competitors envy, the one your strategy decks all start with — is about to become your Kodak film?

The pattern doesn't care about your scale, your brand, or how recently you were called a leader. It cares about whether your dominant skill is durable across the reframe. Usually it isn't.

Sangeet on this in Chapter 4 ↗