I’ve watched this movie enough times to recite it from memory.
A leadership team spends months developing a new strategy. Real work goes into it — research, debate, consensus-building, several near-death experiences in conference rooms. It culminates in an artifact, usually a genuinely beautiful deck, that articulates a smart and cogent shift in direction. Everyone important nods. There’s a palpable sense of relief: the thinking is done; now we just have to implement it. And then, over the following eighteen to thirty-six months, almost nothing changes. The strategy doesn’t fail loudly. It just fails to manifest — until eventually someone concludes that what’s needed is, of course, a new strategy. Roll credits. Cue sequel.
The standard diagnosis is that the strategy was wrong, or the communication was poor, or the market moved. Sometimes. But after years of being called in to examine the wreckage, I’ve come to a more uncomfortable conclusion: the strategy is usually fine. The strategy was never the problem. The problem is that organizations keep writing software for an operating system they don’t run.
The app store problem
A strategy is, functionally, a piece of software. It’s a set of instructions for how the organization should now allocate its energy: pursue this market, deprioritize that product, behave in these new ways. And like any software, it doesn’t execute itself. It runs — or fails to run — on an operating system: the assumptions, structures, habits, incentives, decision rights, and unstated rules that determine how things actually happen in your organization.
Here’s the bind, and it’s worth sitting with because it’s genuinely cruel: your current operating system is perfectly optimized to produce the results you’re currently getting. Those results are presumably why you needed a new strategy. Which means the new strategy must be executed by the very system whose outputs made it necessary — a system that will process the new instructions exactly the way it processed the old ones, because nothing about the system has changed. You’ve written a brilliant new app for hardware that can’t run it. The deck was never going to survive contact with the organization, never mind the market.
This is why “implementation” is the wrong word entirely. Implementation implies a discrete activity with an end state — install the strategy, reboot, done. What’s actually required is implementing: an ongoing, never-finished negotiation between intention and system. The moment you see it this way, an obvious conclusion follows. Strategic change and organizational change are not two workstreams. They are the same workstream, and pretending otherwise is the root cause of the eighteen-month failure cycle.
Strategy is a bid for attention
Let me make the same argument from a different angle, because this is the part I understand better now than when I first wrote about it.
Strip away the frameworks and a strategy is one thing: a bid for the organization’s attention. It’s a request that thousands of hours of human focus be redirected from their current objects to new ones. That’s all “priorities” means.
But attention in an organization is not lying around waiting to be claimed. Every hour of it is already spoken for — by the standing meetings on the calendar, the metrics in the dashboards, the incentives in the comp plan, the habits in the hallways, the thousand small mechanisms of the existing OS. These are standing claims on attention, renewed automatically, defended by inertia, invisible precisely because they’re everywhere.
A strategy deck, by contrast, makes its claim exactly once — at the kickoff, in the all-hands, in the email nobody finishes. Then it’s a PDF, and the PDF goes to war with the calendar. The calendar wins. It always wins, because the calendar makes its claim fifty times a week and the deck made its claim in Q1.
If you want to know an organization’s real strategy, don’t read its strategy document. Read its calendar, its dashboards, and its promotion decisions. Those are where attention actually goes, and where attention goes is what the organization actually does. A new strategy succeeds only when it stops being a document and starts being embedded in those attention-allocating mechanisms — when the standing claims themselves are rewritten. Which is, again, operating system work.
Three ways the old system eats the new strategy
When a strategy dies, it rarely dies of opposition. Nobody stands up and fights it. It dies of digestion — absorbed and neutralized by an OS doing exactly what it was built to do. Three mechanisms do most of the eating.
The asymmetry of risk. Legacy organizations are exquisitely tuned to the risks of doing — every proposed action gets scrutinized, reviewed, and de-risked into mush. The risks of not doing are nearly invisible; there’s no review board for opportunity cost, no post-mortem for the move never made. A new strategy is, almost by definition, a commitment to step into the unknown. Run that commitment through a system that punishes commission and ignores omission, and watch each bold stroke get sanded down until the “new direction” is the old direction with fresh vocabulary.
The gravity of silos. In most large organizations, people’s identities attach to their function long before they attach to the enterprise. So when a strategy demands genuinely collective action, the OS translates “cross-functional” into its native tongue: a working group with one representative from every conceivable faction, growing until it can no longer schedule itself, producing alignment theater instead of aligned action. The strategy needed a coalition. The system supplied a committee.
The frozen structure. Organizations ossify as their work becomes known — and to be fair, that ossification buys real efficiency for known work. But a strategic shift presupposes mobility: re-teaming around new opportunities, redrawing decision rights, tolerating ambiguity while the new terrain gets mapped. Handing a change mandate to a structure optimized for efficiency is like asking a sprinter to run a marathon. The motivation may be sincere. The physiology has been built for something else.
Notice the pattern: none of these are villains. They’re features of an OS doing its job. That’s precisely why strategy work that doesn’t touch the OS is doomed — you’re not fighting resistance, you’re fighting design.
What AI changes (it’s not what you think)
This argument has been true for decades, but something happened recently that raises the stakes considerably.
The strategy artifact — the analysis, the market scan, the options, the beautiful deck — used to be expensive. Months of senior attention, maybe a seven-figure consulting engagement. That expense created a kind of differentiation: not everyone could afford world-class strategic thinking, so having it was an edge.
That edge is evaporating. When every organization in your industry can consult the same frontier models, trained on substantially the same corpus of business knowledge, the documents converge. Everyone’s market analysis cites the same dynamics. Everyone’s strategic options land in the same neighborhood. The deck — the part of strategy that organizations have always treated as the hard part — is becoming a commodity, and commodities confer no advantage.
So where does advantage migrate? To the part that was always the actual hard part and can’t be generated: the organization’s capacity to metabolize a strategy. The trust that lets a team move before everything is certain. The decision rights that let opportunity be seized at the speed it appears. The culture in which people give their genuine energy to a direction rather than performing compliance with it. Two competitors can now hold functionally identical strategy documents — and one of them will eat the other, because strategy-as-artifact has been flattened while strategy-as-organizational-capability has not. This isn’t a sentimental argument about keeping humans in the loop. It’s a cold strategic one: the OS is becoming the last place sustainable differentiation lives.
From strategy to meta-strategy
There’s a tell that distinguishes organizations that have understood all this from those that haven’t: the size of their strategic moves.
Organizations running legacy operating systems lurch. The dramatic pivot, the disruptive reorg, the transformation program with its own logo — these aren’t signs of boldness. They’re signs that an organization has been unable to adapt continuously and must now adapt convulsively, the corporate equivalent of a New Year’s resolution: a fit of aspirational planning that promises we will magically become someone we’ve never been, followed by the familiar collapse when the first quarter punches back. The proponents of big moves always underestimate the churn they generate and overestimate how much trajectory actually changes.
Organizations that have done the deeper work don’t need the lurch. They’ve shifted their effort from strategy to meta-strategy: the ongoing care and fitness of the operating system itself — the sensing, sensemaking, and experimenting machinery that lets them pursue opportunities as they appear and abandon dead-ends before they metastasize into existential crises. Their strategy still changes, sometimes substantially. It just changes the way a healthy organism changes: continuously, at the edges, without requiring a near-death experience to trigger it.
If your organization is contemplating a strategic shift right now, this is actually good news, because a strategy moment is the single best opening for OS work you will ever get. Inertia is briefly interrupted; the status quo has confessed its inadequacy; attention is, for once, available. Use it. Let the teams carrying the new strategic work operate outside the usual bureaucratic envelope, and treat their speed as evidence in the argument against the old way. Treat the strategy’s assumptions as hypotheses to be tested in weeks, not articles of faith to be defended for years. And start the way you intend to finish: a strategy meant to make you adaptive cannot be executed through command-and-control, plan-and-predict methods, any more than you can waterfall your way to agility.
But hold the order of operations clearly in mind. The deck is not the strategy. The deck is a proposal. The strategy is what your operating system does with it — and that, not the thinking, is where the next version of your organization will be won or lost.
A note on lineage: an earlier version of this argument appeared in 2021 as “Why Strategic and Organizational Change Must Happen Together," written during my decade at The Ready. The OS framing and the meta-strategy conclusion began there; the attention argument and the AI convergence argument are where the thinking has gone since.