The Kernel: What Actually Makes a Strategy Work

Most people confuse planning with strategy. They make a list of goals, attach some KPIs, and call it a strategy document. That is not strategy. It is a wishlist with numbers on it. Richard Rumelt nailed this in Good Strategy Bad Strategy, and the core idea is deceptively simple. Every real strategy has what he calls a kernel, and it consists of three parts. A diagnosis of the challenge at hand. A guiding policy for dealing with it. And coherent actions that actually carry out that policy. I spent years watching companies present 40-page decks that contained none of those three things. They had visions, missions, objectives, and initiatives. They did not have a diagnosis. They did not have a guiding policy. They had aspirations dressed up as plans. The difference matters because without a diagnosis, you do not know what problem you are solving, and without coherent actions, you have no way to actually solve it.

Good Strategy Bad Strategy: Why Most Strategies Fail Before They Start

Bad strategy shows up in a few familiar patterns. I see them constantly. Outcome goals dressed as strategy, like "we want to grow 30 percent." That is a target, not a strategy. Fluff words that sound impressive but mean nothing. Strategic goals that are just restatements of the vision. Failure to face the actual challenge, which is the most common one. Here is a specific case from my own work. We were brought in to fix a product launch for a mid-size SaaS company. Their "strategy" was essentially "become the category leader in vertical CRM within 18 months." The team had budget, headcount, and enthusiasm. What they did not have was a diagnosis of why they kept losing deals against two specific competitors who had half their features. I dug into the win and loss data for six months and found the pattern. They were losing on integrations, not on features. Their guiding policy needed to be about building an integration ecosystem, not about marketing louder. The coherent actions shifted from ad spend to partnership development and API investment. Revenue per quarter went from stagnating to growing 12 percent within the next cycle. The insight most people miss is that strategy is not about being aggressive or ambitious. It is about concentrating force on a specific obstacle. The best strategies are often the ones that say no to a lot of good opportunities because they do not serve the diagnosis.

Building a Real Strategy: The Three-Step Process

Start with the diagnosis. This is the hardest part because it requires honest assessment, not hope. You need to understand what is actually stopping you from where you want to be. Not what you think is stopping you. What is actually stopping you. Look at your data, talk to customers who left, examine your competitors' actual moves, and identify the critical barrier. If you cannot state the barrier in one clear sentence, you do not have a diagnosis yet. Once you have the diagnosis, develop a guiding policy. This is your approach to overcoming the barrier. It is not a goal. It is the logic of how you will defeat the obstacle. A guiding policy should be something you can explain in a meeting without pulling out a slide deck. If you need a deck to explain your strategy, you do not have a strategy yet. The third part is coherent action. These are the specific steps that follow logically from your guiding policy. They must be coordinated with each other. One department moving in a different direction than another is not strategy, it is confusion. Coherence means the actions reinforce each other and pull toward the same outcome. I usually see teams skip straight to actions. They start hiring, spending, and launching before they have done the diagnosis. This is why so many initiatives fail. They are solving the wrong problem with a lot of energy.

Where This Framework Breaks Down

Good Strategy Bad Strategy works well for focused, deliberate challenges. It is less useful in highly volatile environments where the terrain changes faster than you can diagnose it. If you are in a space where the competitive landscape shifts every quarter, a fixed diagnosis can become obsolete before you execute on it. In those cases, you need adaptive strategy, which relies more on rapid feedback loops and scenario planning than on deep initial analysis. The framework also assumes you have access to honest data. If your organization is politically driven, with stakeholders protecting their own budgets and agendas, the diagnosis will be contaminated. People will tell you the problem is marketing when the real problem is product quality, because the marketing budget is easier to defend. I have walked out of engagements because the internal politics made honest diagnosis impossible. For those situations, combining Rumelt's kernel with elements ofOODA loop thinking or scenario planning gives you a workaround. You keep the diagnosis but make it revisable. You set up regular checkpoints where you re-examine whether the barrier you identified is still accurate, rather than assuming the first diagnosis was final.

Practical Application in Different Contexts

In product development, the diagnosis might be that your user acquisition cost is unsustainable because your onboarding funnel has a 70 percent drop-off at step two. The guiding policy would be to redesign the activation experience around the core value proposition. The coherent actions would be a redesign sprint, analytics audit, and phased rollout plan. Everything else, including the shiny new feature requests from sales, becomes secondary until the funnel problem is addressed. In organizational change, the diagnosis is often that communication between teams is degraded by conflicting incentives. The guiding policy becomes aligning compensation and metrics around shared outcomes. The coherent actions involve restructuring reporting lines, revising bonus criteria, and establishing cross-functional review cadences. Simply announcing a new mission statement does not count as action. The key takeaway is that strategy is a coherent response to a challenge, not a collection of aspirations. If your strategy document does not contain a clear diagnosis, a guiding policy, and coherent actions, it is not a strategy. It is a wish list with extra steps.