Nokia and Kodak: Two Ways a Strategy Fails to Cascade

On July 8, 2015, Microsoft filed the paperwork on the largest write-off in its history: a $7.6 billion impairment on the Nokia devices business it had acquired barely two years earlier for somewhere in the region of $7.2–7.9 billion. Sit with that for a second, because it’s stranger than it sounds. The write-off was bigger than the purchase price. Microsoft didn’t just lose the money it spent; the acquisition destroyed value beyond what it cost. Announced the same day: 7,800 layoffs, most of them in the phone business.

Seventeen hundred kilometers away and three years earlier, a different giant had already finished dying. Kodak — the company that invented the digital camera — filed for bankruptcy on January 19, 2012, disrupted by a technology its own engineer had prototyped in 1975.

These two stories get told as the same story: “incumbent misses the future.” They’re not. Microsoft/Nokia and Kodak died of two completely different strategy diseases, and I keep both autopsies pinned above my desk because between them they cover most of the ways I’ve watched real product strategies fail. One is a goal mistaken for a strategy. The other is a strategy that stopped cascading before it reached the layer where money gets made.

The third ecosystem: a goal wearing a strategy’s clothes

The stated logic of the Nokia deal was clear, and it fit on a slide: iOS and Android had become a duopoly, and the mobile world needed — Microsoft needed — a third ecosystem. Windows Phone would be it. Buying Nokia’s devices business, the argument went, would secure the hardware, the distribution, and the commitment to make that happen.

Here’s the problem, and it’s the exact problem Richard Rumelt spends the first half of Good Strategy Bad Strategy dismantling: “become the third ecosystem” is not a strategy. It’s a goal. It describes a desired end state and says nothing about why Microsoft would get there. Rumelt has a name for this — mistaking goals for strategy — and it sits on his list of bad strategy’s hallmarks right next to fluff and the failure to face the actual problem.

Run it through the test I laid out in the strategy-formation post: a real strategy is a diagnosis, a guiding policy, and coherent action. What was the diagnosis here? The honest one was brutal: developers build for platforms with users, users buy platforms with apps, and Windows Phone had neither in sufficient quantity to start the flywheel. That’s a cold-start problem, and it’s the actual obstacle. “Buy Nokia” doesn’t address it. Nokia’s hardware was, by most accounts, genuinely good — that was never the constraint. Owning the factories that made the phones did nothing to change why people weren’t buying the phones. The acquisition was coherent action in service of no guiding policy: a very large, very expensive way of restating the goal louder.

And this is the part I find most useful to carry into ordinary product work, because none of us are buying phone divisions but all of us are in rooms where this happens. “Become the leader in mid-market,” “win the enterprise segment,” “be the platform of choice” — every one of these is a wish wearing a strategy’s badge. The tell is always the same: you can’t derive a next action from it that a competitor couldn’t derive from their identical wish. When the diagnosis is missing, spending money is what fills the vacuum, because spending money feels like commitment. Microsoft committed roughly $7.2–7.9 billion to a goal, and two years later wrote off $7.6 billion — more than it paid — and let 7,800 people go. The wish never survived contact with the missing diagnosis.

Kodak knew — that’s what makes it worse

Now the second autopsy, and first the myth has to go, because the myth makes the story useless. The popular telling — “Kodak ignored digital, laughed at it, got blindsided” — is false, and fact-checkers like Snopes have called it false directly. The truth is more uncomfortable.

Steve Sasson, a Kodak engineer, built the first digital camera prototype at Kodak in 1975. Kodak recorded a 1.4-megapixel CCD milestone in 1986. Over the following decades the company invested billions of dollars in digital imaging — research, patents, products. And here’s the number that should end the “they ignored it” story forever: by around 2005, Kodak was roughly the number one seller of digital cameras in the United States. Read that again. The company that supposedly missed digital was, seven years before its bankruptcy, at or near the top of the US digital camera market.

So Kodak saw the future, invested in the future, and briefly led the future — and still filed for bankruptcy in January 2012. Any explanation of that has to be more interesting than “they didn’t see it coming.” They saw it coming for thirty-seven years.

Where Kodak’s cascade actually snapped

The explanation I find most convincing — Chunka Mui made a version of this argument in Forbes the week Kodak filed — lives in the bottom half of the strategy cascade, the layers where strategy has to turn into a business model, a roadmap, and shipped economics.

Kodak’s old business was one of the great razor-and-blades machines in industrial history. Cameras were the razor; film, paper, and chemistry were the blades, sold again and again at margins that funded everything else. Every layer of the company — sales, manufacturing, retail relationships, incentives, the definition of what a “good quarter” looked like — was built downstream of film’s economics.

Digital cameras have no blades. You sell the device once, at hardware margins, into a market that commoditized fast, and there is no consumable stream behind it. So when Kodak’s strategy cascaded down from “we will lead in digital imaging,” it hit the business-model layer and found a question nobody answered: what, specifically, replaces film’s margins? Where do we play in a world without consumables, and how do we win there in a way that pays for being Kodak? That where-to-play/how-to-win choice was never made in a form the lower layers could execute. Winning share in digital cameras — which Kodak briefly did! — was answering the wrong layer’s question. It was capability without a cascade: world-class technology and even market leadership at the top, connected to nothing at the bottom that could keep the lights on.

I think of these two failures as mirror images. Microsoft/Nokia had no real strategy at the top — a goal posing as one — so the billions cascading downward were coherent-looking action attached to nothing. Kodak arguably had real strategic assets and real strategic awareness at the top, and a machine underneath still tuned to film, so the strategy died on the way down, at the layer where it had to become a business model. Different diseases. Same autopsy line: the strategy never cascaded into a shippable, self-funding sequence of choices. One corpse is missing a head; the other is missing a spine. From a distance they look identical, which is exactly why the lazy telling lumps them together.

The version of this happening on your roadmap

Neither disease requires a Fortune 500 body to host it. I’ve seen the Nokia disease in startups that “commit to enterprise” by hiring two salespeople and changing the homepage — spend standing in for diagnosis. And I’ve seen the Kodak disease more times than I can count in healthy product organizations: a team builds the genuinely innovative thing, demos it beautifully, even wins early adopters, while every incentive, pricing structure, and quarterly target in the company still points at the old cash cow. The new thing is allowed to exist but never allowed to win, because winning would mean deliberately cannibalizing the margins that pay everyone’s salary, and no layer of the cascade was ever forced to make that choice explicitly.

That’s the discipline both autopsies argue for. Strategy isn’t done when the direction is chosen, and it isn’t done when the capability is built. It’s done when every layer — position, objectives, business model, roadmap, backlog — can trace its choices upward to a diagnosis and downward to something that ships and sustains itself. Kodak proves you can be first, invested, and even briefly winning, and still fail if one layer in the middle never commits. Microsoft/Nokia proves you can commit harder than anyone on earth — commit more than the asset is worth — and still fail if there was never a strategy to commit to.

Put it to work

1. Run the Rumelt test on your current strategy document. Find the sentence that states the ambition, then look for the diagnosis it rests on: what is the specific obstacle, and why does our approach beat it? If the document restates the goal at higher resolution (“win X by growing Y”) without naming the obstacle, you’re holding a third-ecosystem memo. Fix the diagnosis before approving any spend that exists mainly to demonstrate commitment.
2. Trace your most strategic initiative to its money layer. Take the innovative bet your company is proudest of and follow it down the cascade: what’s the business model when it wins, and whose current margins does that model replace? If the honest answer is “it would cannibalize the core, and nobody has authorized that,” you’ve found your film-margins problem — surface it as an explicit where-to-play decision for leadership, not a fight the team discovers at pricing time.
3. Check whether “we invested in it” is doing the work of “we have a strategy for it.” List the top three things your org points to as proof it’s serious about the future. For each, ask what choice was made that a competitor with the same budget couldn’t copy. Investment without a differentiated choice is Kodak’s billions: real capability, real sincerity, no cascade.

Further reading

• Richard Rumelt, Good Strategy Bad Strategy — the vocabulary this whole post leans on: diagnosis, guiding policy, coherent action, and the hallmarks of bad strategy, including goals mistaken for strategy.
• A.G. Lafley & Roger Martin, Playing to Win — the where-to-play/how-to-win cascade that Kodak’s story fails against, layer by layer.
• TechCrunch, “Microsoft writes down $7.6B of its Nokia acquisition, announces 7,800 layoffs” (July 8, 2015) — the day the write-off exceeded the purchase price; Microsoft’s own July 8, 2015 8-K filing is the primary source underneath it.
• Snopes, “Did Kodak Invent the Digital Camera?” — the fact-check that kills the “Kodak ignored digital” myth: Sasson’s 1975 prototype was built at Kodak.
• Chunka Mui, “How Kodak Failed” (Forbes, January 2012) — written as Kodak filed; the business-model argument, from film’s razor-blade economics down.