Good Strategy Does Not Fail. It Gets Killed Before It Can Work.

I used to give that answer too. Then I spent 20 years watching sound strategy get approved in a conference room and then bleed out slowly on the way to the market, dying not from a single fatal flaw but from two structural forces that almost nobody names out loud. The first is degradation. Strategy loses coherence as it passes through dozens of execution touchpoints, each one a small place where the plan gets compromised, watered down, or quietly ignored. The second is impatience. Boards and stakeholders pull the plug before the program has run long enough to produce the data that would tell them whether it was working.

Neither of those is a strategy failure. Both of them get blamed on the strategy anyway. And the place you land on who actually owns the result says a lot about how you think marketing really works.

The Phenomenon Nobody Warned You About

The research on this is clearer and more damning than most marketers realize. Across decades of studies from HBR, McKinsey, and Bain, the strategy execution failure rate sits somewhere between 60 and 90 percent. That is an enormous number, and the instinct is to assume it means most strategies are bad. It does not mean that at all.

Strategy failure rarely happens at the point of formulation. The thinking is usually sound. The ambition is often right. What breaks down comes later, in the translation from intent to behavior, from slide deck to daily decision. The plan is approved. Then operational reality quietly takes over, and the strategy that actually reaches the market bears progressively less resemblance to the one that was designed. By the time anyone measures the result, they are measuring the degraded version, not the strategy itself. Then they conclude the strategy did not work, when what actually happened is that the strategy was never fully executed in the first place.

This is the gap between intended strategy and realized strategy, and in marketing it is especially wide, because strategy loses coherence as it cascades across channels, teams, agencies, and time horizons. Every one of those handoffs is a place where a little bit of the plan goes missing. None of them feels fatal in the moment. The cumulative effect is fatal almost every time.

Force One: Death by a Thousand Touchpoints

A marketing strategy is not a single action. It is a connected system of dozens of decisions that depend on each other, executed by different people across different functions over weeks or months. And at every single one of those touchpoints, the strategy can be degraded by a choice that feels locally reasonable and is globally destructive.

A leader overrides one recommendation because it threatens their preferred approach. A department declines a process change because the change is inconvenient. A team executes a tactic at a mediocre level because they are stretched thin. A handoff between two functions drops a piece of context that the next step depended on. A data signal gets ignored because acting on it would mean admitting an uncomfortable truth. A timeline gets compressed, so the testing phase gets cut. A budget gets reallocated mid program, so a channel that was supposed to reinforce the others goes dark.

Not one of these decisions kills the program on its own. Each one just removes a small piece of the system. But a connected strategy depends on its pieces connecting. Pull enough of them, even the small ones, and the system that reaches the market is no longer the system that was designed. It is a fraction of it, missing the specific connections that made the whole thing work. And a fraction of a connected system does not produce a fraction of the results. It often produces almost none, because the pieces were supposed to compound, and you cannot compound what is not there.

This is why “execution” is such a misleading word. It sounds like a single phase that either happens well or does not. In reality it is hundreds of micro decisions distributed across an organization, and the strategy survives only if enough of them go the right way. The most mature teams execute with 92 percent confidence versus 57 percent for everyone else, and that gap does not come from better strategy. It comes from aligning the strategy, the processes, and the measurement so that fewer pieces get dropped at the handoffs. The mature teams protect the integrity of the plan across the touchpoints. Everyone else lets it bleed out one reasonable compromise at a time.

I have written before about how you cannot scale nurture when the rest of the buyer journey is leaking. This is the same principle applied to strategy itself. A plan that depends on connected execution cannot survive a journey full of leaks, and most organizations have more leaks than they realize, because each individual leak looks too small to matter.

Force Two: Impatience Pulls the Plug Before the Data Can Speak

The first force degrades the strategy. The second force kills it before anyone can tell whether the degraded version was even the problem.

Marketing produces signal over time. A positioning shift, a content engine, a demand program, a journey redesign. These things compound. The early data is noisy and incomplete because the system has not run long enough to stabilize, the audience has not been exposed enough times to respond, and the optimization loop has not had enough cycles to find what works. Every experienced operator knows that the first weeks of a program tell you almost nothing about its ceiling. They tell you that it has started.

Boards and stakeholders do not operate on that timeline. They operate on the quarter, the board meeting, the update cycle. And so a familiar pressure builds. The program has been running for a few weeks. The numbers are not dramatic yet, because they cannot be yet. A stakeholder asks why it is not working. The pressure to show something forces a premature judgment. And the program gets cut, paused, or quietly defunded before it ever reached the point where the data would have been meaningful.

The cost of this is enormous and almost completely invisible, because nobody ever finds out what the program would have produced if it had been allowed to run. Among successful implementations, 58 percent achieved ROI within six months, versus only 33 percent of failed or neutral cases. Read the timeline in that. The winners measured success over months, not weeks. The programs that get killed in week three are not being judged against a fair standard. They are being judged against an expectation of immediate results that no real marketing program can meet, and then cited as evidence that the strategy failed.

This is the part that should bother every operator and every honest stakeholder. A program cut before it can be optimized was never actually tested. The optimization is where most of the value lives. The first version of any program is a hypothesis. The value comes from running it, reading the real data, finding the two or three things that are actually broken, fixing them, and running it again. Cut the program before that loop completes and you have not learned that the program does not work. You have learned that you were not willing to run it long enough to make it work. Those are completely different conclusions, and organizations confuse them constantly.

And there is a quieter failure hiding inside the impatience, one that is arguably worse than cutting too early. Plenty of organizations leave a program running but refuse to actually work it. They will not tinker. They will not run the small experiments. They will not change one variable, watch what it does, and change the next. They set it live, watch the numbers underwhelm, and treat that as a verdict instead of an invitation to start optimizing. Optimization is not something that happens to a program while you wait. It is something you do to a program, actively, week after week, by testing the subject line, the audience, the offer, the timing, the channel mix, the message, one deliberate micro experiment at a time until the results start moving. The organizations that win at this are restless with their own programs. They are always tweaking something, always running a small test, always hunting for the specific lever that unlocks the outcome. The organizations that lose set the program live and then sit back, arms folded, waiting to be impressed, and when they are not impressed quickly enough, they declare it broken. A program that was never actively optimized did not fail. It was never actually run the way real programs are run.

The standard for declaring something dead is not “the first numbers were underwhelming.” The standard is “we ran enough deliberate experiments, for long enough, to understand both that it will not work and specifically why.” Almost nobody clears that bar before they pull the plug. They cut on the first weak read, having changed nothing, tested nothing, and learned nothing about why the numbers looked the way they did.

Most strategies fail not because they are bad, but because they never get executed well, and a strategy that gets killed before the optimization loop runs is the purest example of that. The thinking never got a fair test. The execution never got to mature. And the verdict gets recorded as “the strategy did not work,” which is the one thing we actually do not know, because nobody let it run long enough to find out.

Why the Two Forces Compound Each Other

These forces are bad on their own. Together they are lethal, because they create a self reinforcing trap.

The touchpoint degradation makes the early results worse than they should be. The strategy reaches the market in a compromised, fractional form, so the initial numbers are weaker than a fully executed version would have produced. Then the impatience kicks in, and a stakeholder looks at those artificially weak early numbers and concludes the program is failing. They cut it. And because they cut it, the optimization loop never runs, which means nobody ever discovers that the real problem was the dropped touchpoints, not the strategy.

So the degraded execution produces weak early signal, the weak early signal triggers premature cancellation, and the premature cancellation guarantees that the actual cause, the degradation, never gets diagnosed. The organization concludes the strategy was wrong. They go looking for a new strategy. They run it through the same degrading touchpoints and the same impatient timeline. And they get the same result, for the same reasons, which they will once again misattribute to the strategy.

This is why the strategy execution failure rate has stayed stuck between 60 and 90 percent for decades despite every company having access to better strategic thinking than ever before. The bottleneck was never the quality of the thinking. It is the integrity of the execution and the patience of the timeline, and most organizations have a structural deficit in both. Better strategy cannot fix a system that degrades strategy and then refuses to let it run.

The Part Where I Implicate the Strategy Side

I am not going to pretend the people who build strategy are blameless here, because that would be too easy and you would be right not to trust it.

Whoever owns a strategy owns more than the thinking. They own the adoption. If a plan is so fragile that it cannot survive a normal organization’s touchpoints, that is partly a strategy problem, because real strategy has to be built for the organization that will actually execute it, not the idealized one that exists on the slide. A strong strategist anticipates where the plan will get degraded and designs around it. They build buy in so departments do not override the pieces that matter. They sequence the work so early wins create the patience that protects the program later. They set expectations with stakeholders up front about what the timeline actually requires, so the impatience does not get a foothold. The best operators do not just hand over a sound plan and hope. They protect the plan through the execution and manage the stakeholders who might kill it.

That is real, and anyone who does this work for a living should hold it. The question nobody can fully resolve is where the line sits. Because there is a point past which no amount of strategic skill survives an organization that is structurally committed to degrading the plan and impatient enough to cut it early. You can anticipate the touchpoints, build the buy in, and set the expectations, and still watch a stakeholder pull the plug in week three because the quarter looks soft. You can do everything right on the strategy side and still lose to the two forces, because owning the strategy is not the same as controlling the execution or the boardroom. Different people control those, and they answer to different pressures.

This is exactly why the strategy versus execution distinction is so slippery. The two are not cleanly separable. A strategy that cannot survive a normal organization was arguably not complete. But an organization that degrades and then prematurely kills a sound strategy was arguably never going to succeed regardless of what it was handed. Both are true at once, which is precisely why this is a real disagreement and not a settled question.

What This Means for How You Run Marketing

If you sit on the stakeholder side of this, there are a few things worth sitting with, because the goal is not to win an argument. It is to stop killing strategy that might have worked.

Protect the integrity of the plan across the touchpoints. When a department wants to override a recommendation or skip a process change, understand that you may be removing a load bearing piece of a connected system. The override feels local and reasonable. The damage is systemic and invisible until the results come in weak. Before you let a piece get dropped, ask whether the rest of the plan depended on it, because connected strategies fail quietly when their connections get cut.

Set the timeline before you start, not after the numbers come in. The single most destructive thing a board can do to a marketing program is judge it against a timeline nobody agreed to up front. Decide before launch how long the program needs to run before the data is meaningful, and hold that line when the pressure builds in week three. The winners measure ROI over months. If your governance cannot tolerate that, your governance will kill good programs on a schedule, and you will keep concluding that marketing does not work when what actually does not work is your patience.

Let the optimization loop run, and actually work it, before you render a verdict. The first version of any program is a hypothesis, not a finished product. The value is in running it, reading the real data, fixing the two or three things that are actually broken, and running it again. But running it is not passive. Optimization is something you do, not something you wait for. Change a variable, watch what it does, change the next. Test the audience, the offer, the message, the timing, the channel mix, one deliberate experiment at a time, until the results start moving. A program left running but never actively tinkered with is not being optimized. It is just being watched. And a program cut before that loop completes, or before anyone bothered to run a single real experiment, was never tested, only started. The bar for declaring something dead is not “the first numbers underwhelmed.” It is “we ran enough deliberate experiments, for long enough, to understand that it will not work and specifically why.” Almost nobody clears that bar before they pull the plug.

And audit the execution before you blame the strategy. When a program underperforms, ask the honest question first: did we actually run the strategy, or a degraded fraction of it? Did the pieces connect, or did half of them get dropped at the handoffs? Did we let it run long enough to optimize, or did we cut it early and call that a result? Most “the strategy failed” conclusions cannot survive that audit. This connects to the measurement problem I covered in The KPI Mirage. If you are reading weak early numbers from a degraded, prematurely judged program as proof the strategy was wrong, you are not measuring reality. You are measuring a verdict you reached before the evidence existed.

So, Whose Fault Is It?

Back to the question, because I promised I would not dodge it.

My honest answer is that it is split, and the split is not 50/50. The strategy side owns the part where a plan has to be built for the real organization and protected through the people who might degrade or kill it. Adoption and stakeholder management are skills, and a plan that dies in execution was, in part, not defended well enough. That is real, and I am not going to wave it away.

But the larger share sits with organizations that degrade sound strategy across a hundred small compromises, then refuse to actively work the program once it is live, then cut it before the optimization loop can run, and then record the result as a strategy failure, which is the one thing it almost never is. You cannot diagnose a strategy that was never fully executed. You cannot judge a program you never bothered to tinker with. You cannot conclude something will not work when you changed nothing, tested nothing, and quit before the data could tell you why. At some point the organization owns the integrity of its own execution, the restlessness of its own optimization, and the patience of its own timeline, and no quality of strategic thinking can compensate for a system that bleeds plans out, refuses to work them, and pulls the plug too early.

That is where I land. The strategy is usually sound, the execution is usually a degraded fraction of it, the program is usually never actively optimized, the timeline is usually too short, and the verdict is usually rendered on evidence that was never allowed to exist. The strategy side could always defend the plan better and manage stakeholders harder. Both are true, which is the whole point.

Now I want your answer, because this is where reasonable people disagree and I want to know where you actually stand. When sound strategy degrades across dozens of touchpoints and then gets cut before it can be optimized, whose fault is the result? Is it on the strategy for not surviving the organization, or is there a point where the organization owns what it did to a plan that might have worked?

Tell me where you land. I have given you mine.

For more on building marketing strategy that survives execution and earns the patience to work, visit the Transmyt blog.