Good StrategyBad Strategy

1. 1Restate the critical challenge from your diagnosis in one sentence — name the specific obstacle to forward progress, not a goal or aspiration. If you cannot name it, you have no strategy to overcome.
2. 2Identify the one or two pivot points: the factors where focused action multiplies effect. For Nelson it was enemy fleet coherence; for Petraeus it was civilian support of a legitimate government.
3. 3Design coherent actions that concentrate resources on those pivot points, and cut actions that dissipate them across conflicting demands. Coherence means the steps reinforce each other, not merely coexist.
4. 4Anticipate the real-world difficulties that resolution will trigger (foreclosures, insurgency, competitor imitation) and build mitigating actions into the plan rather than assuming the obstacle stays static.
5. 5Set specific guiding policies that structure daily action — like DARPA rotating program managers every four to six years and keeping overhead low — so the intent survives contact with the organization.
6. 6Track whether the gap between challenge and current state is actually closing; require evidence of movement on the pivot point, not activity or slogans.
7. 7Hold the position long enough for the strategy to work — Resnick waited years for processing scale to lock in — and refuse pressure to abandon it prematurely.

General Petraeus in 2007 inherited a diagnosed challenge U.S. units were not overcoming: a violent insurgency that search-and-destroy missions from secure bases were feeding rather than defeating. The prior approach had numerous high-sounding goals — freedom, democracy, reconstruction, security — but no coherent strategy tied to the actual mechanics of the insurgency. Petraeus's diagnosis named the pivot point: an insurgency can be beaten so long as the large preponderance of civilians support a legitimate government. His guiding policy shifted the military's focus toward protecting and winning that civilian population rather than raiding from isolated bases. Backed by the new Counterinsurgency Field Manual and five additional brigades, the coherent actions followed from that single insight. The challenge was overcome because the design attacked the factor that actually controlled the outcome, not because more effort or ambition was applied.

• If your plan lists broad goals (freedom, growth, being 'the best') but names no obstacle, then you have ambition, not a strategy that overcomes anything — rewrite around the challenge.
• If an action does not concentrate force on an identified pivot point, then cut it; a dog's dinner of objectives dilutes the punch needed to break the challenge.
• If resolving the challenge will provoke a reaction (competitor imitation, higher risk, insurgency), then pair the guiding policy with a mitigating mechanism — Lehman's 'more risk appetite' failed because it had none.
• If the challenge is entrenched and old (like Cold War-era national security structures), then assume the power required to overcome it is many times greater than for a fresh problem, and plan accordingly.
• If you feel pressure to quit before results appear, then check whether the pivot-point evidence is moving; if it is, hold — iron nerve and a five-to-ten-year horizon are often what completes the resolution.

• Mistaking a goal for the overcoming of a challenge. Chad Logan's 20/20 plan set 20% growth and 20% margin but never named an obstacle. The tell: the 'strategy' is a number or a vision statement with no problem in it.
• Substituting ambition or 'risk appetite' for a mechanism to surmount the challenge. Lehman grew faster into rejected deals with 3% equity and no risk mitigation. The tell: the plan scales up exposure to the very thing that was the threat.
• Falling back on a failed prior approach because no coherent one exists. U.S. units ran search-and-destroy — Vietnam's failed method — for lack of a strategy. The tell: default behavior resumes and everyone calls it 'staying the course.'
• Charisma and vision without attention to obstacles and action. The Children's Crusade moved thousands who died to no end. The tell: enthusiasm and followers are high, but no one can state which obstacle the marching removes.
• Declaring victory on activity rather than on the pivot point. The tell: reports full of effort and initiatives, but the diagnosed challenge is measurably unchanged.

• ▸Challenge (one sentence): the specific obstacle to forward progress, named as a problem not a goal.
• ▸Pivot point(s): the one or two factors where focused action multiplies effect.
• ▸Guiding policy: the overall approach chosen to cope with or overcome that obstacle.
• ▸Coherent actions: 3-5 steps that reinforce each other and concentrate resources on the pivot point.
• ▸Anticipated reaction: what the obstacle or rivals do when you act, and the mitigating step for each.
• ▸Evidence of movement: the observable measure showing the pivot point is shifting (not activity counts).
• ▸Holding horizon: how long resolution realistically takes and what pressure you will resist.
• ▸Cut list: actions removed because they dilute force or serve conflicting demands.
• ▸Verdict: overcome / partially mitigated / unresolved — with the specific evidence that justifies it.

1. 1Define the comparison baseline before claiming superiority: name the direct competitors, the industry average, or the past-performance benchmark you will measure against. Crown's 18.5% annual shareholder return meant nothing until placed beside the S&P 500's 8.6% for the same period.
2. 2Separate the outcome metric from the source of the outcome. Profitability, market share, and return on capital are results; identify the underlying strategic resource or coherent design that produced them, because current results reflect 'planting seasons long past,' not recent actions.
3. 3Trace superior performance back to a specific competitive advantage — a concentration and coordination of action and resources that lessens buyer bargaining power or applies strength against weakness. If you cannot name it, you have a lucky result, not a durable position.
4. 4Test whether the advantage is being maintained or eroded. Ask: what industry change (new technology, substitute product, shifting buyer demand) is undermining the basis of past superiority? Crown's advantage eroded when PET bottles cut product-changeover costs.
5. 5Build the next strategic resource off the current one before the current one decays, even if doing so reduces the immediate value of what you have. Xerox failed to build paper-handling capability because it protected the near-term value of its service system.
6. 6Reject goal-substitution: setting '20/20' targets or profit/market-share numbers is not a claim of superior performance. Attach every performance ambition to a specific process or accomplishment and a point of leverage — a key strength you build on or an industry change you exploit.
7. 7Guard against the success trajectory: superior performance breeds laxity and bloat, which breed decline. Institutionalize the disconnect between current profit and current action so leaders do not mistake plenty for proof of correct current behavior.

Under Connelly, Crown Holdings grew revenues at only 3.1% per year yet returned 18.5% annually to shareholders versus the S&P 500's 8.6% — demonstrably superior performance. The source was not effort or scale but a carefully designed strategy that focused the company on products and buyers where customer bargaining power was weakest, holding return on capital at a respectable 15.3%. When Avery took over, PET plastic bottles were cutting into soft-drink containers because changeover costs in plastics were far lower than in metal, eroding the basis of Crown's advantage. Rather than replace the advantage, Avery chased growth by acquiring into the PET business and buying CarnaudMetalBox. Revenue and market position climbed — Crown became the 'leading' container maker in the world — but return on capital fell below 10%, then below 5%, and shareholder returns collapsed to 2.4% per year against the S&P's 9%. The case shows superior performance is a harvest of a coherent advantage, not of size or growth; abandon the advantage and the results invert even as the top line grows.

• If a strong result is not tied to a nameable advantage that lessens rival or buyer power, then treat it as unexplained and unrepeatable, not as superior performance to build on.
• If growth in revenue or market share coincides with falling return on capital, then you are diving into a commodity and destroying value — stop and re-diagnose focus.
• If protecting a current profitable resource blocks building the next resource (as Xerox's service system blocked paper handling), then build the next resource anyway — near-term value is the harvest of a past you cannot re-plant.
• If leadership explains current profit by pointing to recent actions, dress codes, or policies, then suspect specious attribution and look for the strategic seeds planted long before.
• If the 'strategy' is a set of performance goals with no coherent actions or leverage point, then relabel it goal setting and refuse to call it strategy.

• Confusing growth with superior performance. The tell: sales and market position rise while return on capital drops (Crown post-Connelly, from 15.3% to under 5%).
• Attributing results to current actions. The tell: leaders 'point to their every action with pride' and books recommend copying the winner's vacation and parking policies.
• Resting on a powerful resource position. The tell: profit arrives 'without great effort,' laxity sets in, and rivals (Canon, Kodak, IBM) develop the capability you declined to build.
• Interpreting focus as merely a narrow product line rather than concentration and coordination of resources. The tell: a CFO saying 'we've always been focused — we serve a $150–200 billion slice' while ignoring the loss of competitive advantage.
• Substituting will-to-win rhetoric for a point of leverage. The tell: a leader like Chad Logan quoting 'never quitting until you win' or Jack Welch's 'reaching for the impossible' when asked what key strength the jump in performance rests on.

• ▸Performance metric: state the specific result (return on capital, shareholder return, market share).
• ▸Comparison baseline: name the competitor, industry average, or prior period and its number.
• ▸Margin of superiority: your figure minus the baseline figure.
• ▸Source advantage: name the coherent design or strategic resource producing the result.
• ▸Buyer/rival power lever: how the advantage weakens customer or competitor bargaining power.
• ▸Erosion threat: the technology, substitute, or demand shift undermining the advantage.
• ▸Next resource to build: the capability that extends the advantage even at near-term cost.
• ▸Attribution check: evidence the result comes from past design, not recent unrelated actions.
• ▸Goal-vs-strategy test: the specific coherent actions and leverage point, not just the target number.

1. 1Define the specific boundary of your advantage before defending it: name the exact product, application, and buyer segment where you produce at lower cost or deliver more perceived value. Whole Foods only holds advantage over Albertsons for certain products among higher-income shoppers who value organic goods—do not claim advantage beyond where cost and buyer characteristics actually support it.
2. 2Identify the isolating mechanism underneath the advantage—the reason rivals cannot duplicate the underlying resources. Sort candidates: patents, reputations, commercial and social relationships, network effects, dramatic economies of scale, tacit knowledge from experience. If you cannot name at least one, the advantage is not sustainable.
3. 3Test whether your advantage rests on chain-linked activities rather than a single policy. Ask: could a rival grab your business by copying just one activity and doing it well? If yes, you are exposed. IKEA's durability comes from policies (own stores, own designs, catalogs over salespeople, ready-to-assemble, large stores, logistics) that only work as a coherent set.
4. 4Verify all three IKEA conditions hold: each core activity is run with outstanding efficiency; the activities are chain-linked so single-policy imitation fails; the grouping is unusual enough that skill in one does not transfer to skill in the others. Missing any one collapses sustainability.
5. 5Convert existing resources into new resources instead of resting on them. Xerox used its plain-paper patent to build a world-class repair service—the correct next step was building paper-handling capability toward personal copiers and printers. Map what your current resource makes possible next.
6. 6Audit for the laxity that potent resources induce. When profits arrive regardless of business logic (a dominant patent, brand, or franchise), you stop doing strategy work. Watch for independent fiefdoms flourishing, conflicting initiatives multiplying, and integration becoming impossible.
7. 7Stay a moving target. Since advantages erode, deepen and extend the position over time rather than freezing it—broaden the brand into adjacent applications and refresh the resource base before it goes obsolete, so upstarts find no fixed opening.

Xerox owned the original patent on plain-paper copying—a potent isolating mechanism that made profits roll in almost regardless of business logic. It correctly used that patent to build a second strategic resource: a world-class fast-response repair service for its installed base of failure-prone leased machines, complemented by a profitable Xerox-brand paper business that jammed copiers less. The right next move was to build world-class paper-handling capability, which would have opened early positions in personal copiers, printers, and fax machines. But that would have reduced the immediate value of the service-system resource, so Xerox rested on its laurels. Canon, Kodak, and IBM developed superior paper-handling technology while Xerox chased a futile entry into computing. The lesson: a potent resource sustained profits for a time without genius, but the same resource became an impediment to innovation because it was embedded deeply in the enterprise's human fabric—the advantage decayed precisely because no ongoing strategy work was done.

• If an 'advantage' is available for purchase by all comers (an IT system, product, or training program on the market), then it is not an advantage—the term is a contradiction, so discard the claim.
• If a rival can grab your business by copying one policy and performing it well, then your position is not chain-linked and not sustainable—redesign toward a coherent set, not a single strength.
• If profits arrive regardless of how you arrange your business logic, then assume laxity is setting in—impose the discipline of tight integration deliberately, because the resource will mask decline until it is obsolete.
• If preserving a current resource blocks building the next capability the market rewards, then retire the old resource like obsolete machinery rather than protecting its immediate value.
• If you want to learn durable design-type strategy, then study an upstart's early conquests (young Microsoft besting IBM, young Wal-Mart besting Kmart, Nvidia stealing graphics dominance), not the mature giant's posturing.

• Confusing a potent resource with sustained strategy. The tell: profits are strong but no one is doing careful ongoing strategy work—success arrived and the discipline of tight integration quietly disappeared.
• Copying one policy of a chain-linked rival. The tell: a traditional furniture retailer adds a catalog but still lacks design, logistics, and large stores—expense rises with no real competition delivered.
• Overextending the advantage past its real boundary. The tell: you assume your lower cost or higher perceived value applies to all buyers and applications, ignoring that buyers differ in location, knowledge, and tastes.
• Protecting the current resource's immediate value at the cost of the next resource. The tell: the logical next capability would 'cannibalize' an existing profitable system, so it never gets built—and rivals build it instead.
• Faking youthful vigor with bolt-on acquisitions as growth naturally slows. The tell: independent fiefdoms flourish, conflicting products and standards multiply, and integration becomes impossible—the same posture IBM held in 1985.

• ▸Advantage boundary: name the exact product, application, and buyer segment where you hold cost or value advantage.
• ▸Source of advantage: state whether it is lower cost, more perceived value, or a mix—and by how much.
• ▸Isolating mechanism: name the specific resource rivals cannot cheaply duplicate (patent, reputation, relationships, network effects, scale, tacit knowledge).
• ▸Chain-link test: list the linked activities; mark whether copying any single one lets a rival grab your business.
• ▸Unusual-grouping check: state whether expertise in one activity transfers to the others (it should not).
• ▸Resource-to-resource path: name the next capability this resource makes possible and what building it would cost the current one.
• ▸Laxity indicators: flag fiefdoms, conflicting initiatives, bolt-on acquisitions, or profits arriving without strategy work.
• ▸Obsolescence horizon: estimate when the underlying resource base decays and what upstart move would exploit the opening.

1. 1Restate the presenting complaint separately from its cause. If leadership names 'underperformance' or 'low expectations,' write it down as a result, then force the question: why has this happened? The diagnosis lives in the answer, not the complaint.
2. 2Identify the one or two aspects of the situation that are pivotal and discard the rest. A good diagnosis simplifies overwhelming complexity by naming certain facts as critical — you are choosing what matters, not cataloguing everything.
3. 3Offer an explicit theory of the difficulty: state why the challenge is hard and why past efforts failed. Without this theory, no guiding policy or action can be evaluated as fitting.
4. 4Test any diagnosis against the actual mechanics of the situation, not against analogy or hope. Enumerate the physical, structural, and economic constraints and ask whether the proposed way of operating can actually work under them.
5. 5Strip fluff and Sunday words from the diagnosis. Rewrite each phrase in plain terms; if 'customer-centric intermediation' collapses to 'being a bank,' the diagnosis was empty and you must redo it.
6. 6Name the diagnosis as a clinician names a pathology — a specific label that defines a domain of action. The label should imply which guiding policies are on and off the table.
7. 7Confirm the diagnosis defines a bounded field of action: if it does not narrow what you will and won't do next, it isn't a diagnosis yet.

In summer 2000, Arthur Andersen presented Enron's bandwidth-trading move as a repeat of its gas and electricity success, arguing that 'commodity markets evolve in the same way, so the same business strategies apply to all.' That is a missing diagnosis dressed as fluff. A quality diagnosis would have named the pivotal fact: bandwidth is not analogous to gas or electricity. There was no spot price for basis trading, no standard of quality to define the deliverable, no way to ship capacity across geographies to balance supply and demand, and the incremental cost of a unit of bandwidth was zero — meaning spot price would collapse toward zero once capacity exceeded demand, which by 2000 it already did. Enron also traded the deliverable in gas and electricity but only capacity in bandwidth, and held no position in content. A leader who diagnosed the challenge as 'this market lacks the price and delivery structure our trading model requires' would have rejected the strategy; instead the analogy substituted for analysis, and no systematic bandwidth market ever developed.

• If the stated challenge is a result (underperformance, low share, low expectations), then it is not yet a diagnosis — require a theory of the cause before proceeding.
• If you cannot state why the challenge is difficult or why past attempts failed, then you cannot assess or improve any strategy built on it — stop and diagnose.
• If a phrase survives only because it sounds sophisticated, then translate it to plain language; if the plain version is trivial or empty, the diagnosis is fluff and must be redone.
• If the diagnosis rests on analogy to a past situation, then verify the mechanics match — analogy chooses the guiding policy, so a false analogy (Ahmadinejad as Hitler vs. Gadhafi) points action in opposite directions.
• If the diagnosis does not narrow the domain of action, then it has failed its job — a good diagnosis rules options in and out.

• Setting performance goals in place of diagnosing (Harvester's plan to raise market share and cut costs in every division). Tell: the 'strategy' is a list of ambitions with no account of why the company is stuck.
• Masking absence of substance with unnecessary complexity. Tell: fluffy, Sunday-word language ('customer-centric intermediation,' 'transformational leadership') that collapses to a triviality when translated.
• Diagnosing by analogy without checking the mechanics. Tell: 'markets evolve the same way, so the same strategy applies' — the differences between the cases go unexamined.
• Prescribing before diagnosing (asking for 'transformational leaders' with no theory of why leadership is weak). Tell: blue-sky objectives that assume away the very obstacles that make the challenge hard.
• Amalgamating separately-produced sub-diagnoses into one document with no unifying account of the core difficulty (Harvester's five divisional plans stapled together). Tell: no single critical issue is named; everything is a priority.

• ▸Presenting complaint: state the result or symptom in plain words
• ▸Named pathology: the single label defining the nature of the challenge
• ▸Theory of difficulty: why this challenge is hard right now
• ▸Why past efforts failed: the specific reasons prior attempts didn't work
• ▸Pivotal facts: the one or two critical aspects that matter most
• ▸Mechanics check: constraints (physical, structural, economic) the situation actually imposes
• ▸Fluff scan: rewrite each key phrase in plain language; flag any that collapse to nothing
• ▸Domain of action: what this diagnosis rules in and rules out

1. 1Start from your written diagnosis. A guiding policy has no meaning until you name the specific obstacle it addresses—so state the challenge in one sentence before drafting any approach.
2. 2Write the overall approach as a signpost, not a goal. It should mark a direction ('serve busy professionals rather than students') without prescribing every action. Test: does it read as an approach to an obstacle, or just as a performance target?
3. 3Force the approach to rule things out. If your policy is compatible with every action you might have taken anyway, it isn't guiding. A real guiding policy makes a vast array of possible actions off-limits.
4. 4Identify the source of advantage the policy creates or leverages. Kennedy leveraged America's much greater resources against a poorer Soviet effort; ask what asymmetry your approach exploits.
5. 5Push immediately toward action to test the policy. Ask 'what does this force me to do and stop doing?' The necessity to act is the sharpest tool for exposing a woolly policy.
6. 6Check the policy names one priority that takes precedence. If it tries to accommodate a multitude of conflicting demands, it is not focused—decide which priority wins.
7. 7Draft 2-4 coherent actions that coordinate and build on one another under the policy. If you can't specify feasible actions, the policy is still fluff.

Stephanie diagnosed her market's challenge as competing against larger stores by focusing on a distinct customer. Her guiding policy was to serve busy local professionals rather than students. Notice how this signpost directs and constrains without dictating detail: it rules out midnight hours (professionals don't shop at midnight) and rules out devoting space to student munchies. It simultaneously generates coordinated actions—a second checkout stand to handle the 5 p.m. rush, more parking in the alley, prepared high-quality take-home foods replacing munchies, and adequate staffing after work and at lunch. Each action follows from the same approach and reinforces the others. The policy created advantage by concentrating limited resources on one well-defined segment instead of spreading them across every possible shopper.

• If the statement could apply to any competitor in your industry, then it's fluff, not a guiding policy—rewrite it to exploit a specific asymmetry.
• If the policy names no obstacle from the diagnosis, then it's a goal in disguise—reconnect it to the challenge or discard it.
• If nothing gets ruled out when you adopt it, then it isn't guiding—add the actions and options it forbids.
• If you cannot derive at least two coordinated feasible actions from it, then it's too abstract to bring down to earth—sharpen it against the necessity to act.
• If the approach tries to satisfy all stated priorities at once, then choose which priority takes precedence before proceeding.

• Confusing the guiding policy with 'the strategy' and stopping there. Tell: there is a stated approach but no coherent action, leaving the chasm between strategy and implementation that produces the chronic 'execution problem' complaint.
• Substituting ambition for approach, as with Lehman's plan to 'gain market share by growing faster' and increasing risk appetite. Tell: the statement describes how much you want to win, not how you'll surmount the obstacle—and it carries no risk-mitigating actions.
• Marking pay grade instead of naming an approach. Tell: something is called 'strategic' only because a senior official decided it (a merger, a facility), with no cohesive response to a challenge behind it.
• Broad goals masquerading as policy—'Let's win,' vision, values, blue-sky objectives. Tell: it inspires but rules nothing out and dictates no direction; it's a verbal tic like 'growth strategy' or 'IT strategy.'
• Trying to accommodate a dog's dinner of conflicting demands. Tell: the policy embraces every priority as co-equal, so no resources get focused and no action can begin.

• ▸Diagnosed challenge (one sentence): name the critical obstacle this policy answers.
• ▸Overall approach (signpost): the direction forward, stated without prescribing every detail.
• ▸Source of advantage: the asymmetry or strength this approach creates or leverages.
• ▸Ruled out: list the major actions/options this policy now forbids.
• ▸Priority that wins: which competing demand takes precedence.
• ▸Derived actions (2-4): coordinated feasible steps that build on one another.
• ▸Fluff check: could a competitor state this same line? Yes/No—if yes, rewrite.
• ▸Action test: what does this force you to do now and stop doing?

1. 1Name the challenge first: write one sentence stating the specific high-stakes problem your actions must overcome (Jobs: 'the product lineup was too complicated and the company was bleeding cash'). Actions without a named challenge cannot be tested for coherence.
2. 2List every current policy, resource commitment, and action already in motion. Put them side by side—this exposes whether you have a strategy or a pile of separately-authorized decisions.
3. 3For each pair of actions, ask: does A make B more effective, less effective, or neither? Sort into three buckets—mutually reinforcing, conflicting, and merely non-conflicting-but-uncoordinated.
4. 4Cut the conflicting actions. Volvo-on-Jaguar-platform diluted both brands; when two actions undermine each other, one must go, not both survive as compromise.
5. 5Cut or defer the uncoordinated actions that only pass a 'requires top-management approval' test. Closing Akron, more advertising, and 360-feedback may each be sound operationally but they do not complement one another—so they are not strategy.
6. 6Identify the one or two pivot points where concentrated action multiplies effect, then redesign the remaining actions so each reinforces the others around that pivot (Jobs collapsed all desktops into the Power Mac G3 and dropped five of six retailers—a single simplifying logic ran through both moves).
7. 7Convert the design into a set of plausible, feasible immediate actions—these are the punch, not implementation details. If you cannot name concrete first moves, you have set goals, not a strategy.
8. 8Re-test the whole set: can you explain it in a few sentences without slides? Nelson's plan and Jobs's plan both compressed to a paragraph. If it needs a thick deck, coherence is probably missing.

In 1997 Apple was two months from bankruptcy, its product line sprawling and cash draining. Jobs named the challenge in plain terms: too many models a family friend couldn't choose between, no consumer machine under $2,000, and a company bleeding cash. Rather than announce revenue targets or a visionary future, he designed a coordinated set of actions that reinforced each other: replace all the desktop machines with one, the Power Mac G3; drop five of six national retailers because serving them forced too many models at too many price points; halt new operating-system development because he had already brought the industry's best OS from NeXT; and adopt Asian supply-chain manufacturing with short cycle times and negative working capital. Each move made the others work better—fewer retailers made a simplified line viable, a simplified line made supply-chain efficiency possible. This was not blind ax-wielding; it was a redesign of the whole business logic around a single simplifying idea. That coherence, not ambition, is what created the strength.

• If you cannot state the challenge the action set addresses in one sentence, then you have goals or a wish list, not a coherent strategy—stop and diagnose first.
• If two actions pull in opposite directions (safe Jaguar vs. distinctive Jaguar), then treat them as a conflict and eliminate one; a compromise satisfies no one.
• If actions are non-conflicting but none makes another more effective, then they are merely operational management; coordination is absent and so is advantage.
• If explaining the strategy requires a thick deck of slides, matrices, or triangles, then suspect there is no real pivot point—good strategy looks simple and obvious.
• If a decision is called 'strategic' only because a senior person authorized it, then reclassify it—pay grade is not coherence.

• Confusing goal-setting with strategy: setting market-share and profit targets and calling it strategy. Tell: a persistent 'execution problem' where you always fall short of the goals you set—the gap is because coherent actions were never defined.
• Pursuing multiple unconnected initiatives that symbolize progress with no approach beyond 'spend more and try harder.' Tell: a long list of parallel programs (alliances, 360-feedback, foreign markets, firmware updates) that nobody can tie to a single challenge.
• Conjoining actions that quietly conflict—buying Volvo and Jaguar then forcing a shared platform. Tell: your most passionate customers, dealers, and service shops are annoyed because the combined action erodes what each brand promised.
• Treating coherent actions as mere 'implementation' downstream of the real strategy. Tell: a wide chasm opens between 'strategy' and 'execution,' and strategy work becomes wheel spinning divorced from any specific action.
• Substituting exhortation—vision statements, balloon releases, 'global leadership, growth, high shareholder return'—for a cohesive response to a challenge. Tell: the 'strategy retreat' has everything but strategy.

• ▸Named challenge: state in one sentence the high-stakes problem all actions must overcome.
• ▸Pivot point(s): the one or two critical issues where concentrated effort multiplies results.
• ▸Action inventory: list every current policy, resource commitment, and action in play.
• ▸Reinforcement check: for each pair, mark whether A makes B stronger, weaker, or neither.
• ▸Conflicts to cut: actions that undermine one another—choose which survives.
• ▸Uncoordinated to defer: non-conflicting actions that reinforce nothing—remove or park them.
• ▸Reinforcing set: the surviving actions, each shown to strengthen the others around the pivot.
• ▸Immediate moves: 2-4 concrete, feasible first actions that carry the punch, not implementation notes.
• ▸One-paragraph test: write the whole strategy in a few sentences—if you can't, coherence is missing.

1. 1Write an explicit diagnosis first: name the one or two critical issues in the situation—the pivot points that multiply the effectiveness of effort. Without this, you cannot focus resources on a cure; you can only shift them around.
2. 2Estimate the threshold for each candidate objective—the critical level of effort below which you get little or no payoff. List only objectives that your available resources can push past that threshold.
3. 3Rank candidate objectives by return to concentration: which ones, if fully resourced, generate disproportionately larger payoffs because of constraints and threshold effects.
4. 4Cut the list down to the few your senior leadership group can actually hold in attention at once—treat leadership cognition and attention as a hard, limited resource, not an infinite one.
5. 5Reallocate financial capital, key talent, and leadership attention onto the surviving few objectives; explicitly name the actions and priorities you are stopping so resources are not spread to cover all bases.
6. 6Test for coherence: confirm your resource commitments and actions coordinate toward the guiding policy rather than fighting each other, then commit and execute now rather than perfecting later.
7. 7Concentrate application in time and space where useful—pulse effort into short periods or roll out region by region so you clear the threshold in one place before moving on.

Harold Williams used concentration to make the Getty a world presence in art rather than one modest player among many. The same logic appears in advertising: because a very small amount of advertising produces no result at all, spreading a budget evenly across the year or across every region wastes it below the threshold. The remedy is to pulse—concentrate spending into short periods—or to roll out a new product region by region, concentrating advertising where the product is new to spur adoption. A business strategist applies the identical rule when choosing to dominate a small market segment rather than hold an equal number of customers who are only a sliver of a larger market. Domination clears the threshold and produces returns; a thin, even spread does not. In each case the payoff comes not from more total resources but from refusing to divide them below the level at which they can affect the system.

• If there is no explicit diagnosis of the underlying problem, then stop—you have no basis for focusing resources and will only shift them around, as the 2008 crisis response shifted resources from the public to the banks.
• If an objective cannot be pushed past its threshold with available resources, then drop it; effort below the threshold has little payoff.
• If a strategy tries to cover all the bases rather than choose, then treat it as bad strategy driven by unwillingness to offend, and force a hard choice.
• If leadership attention is already committed to several priorities, then adding another dilutes all of them—an organization, like an individual, cannot solve five problems at once.
• If two schools fully turned around move public opinion more than marginal gains spread across many, then concentrate on the visible, threshold-clearing target.

• Treating strategy as goal setting rather than problem solving—the tell is high-sounding goals (freedom, democracy, most respected company) with no diagnosis and no concentration of action behind them.
• Confusing ambition, vision, or 'never quitting until you win' with strategy—the tell is slogans, motion pictures, and dramatic music at a 'strategy retreat' but no identified pivot point.
• Spreading resources thinly to avoid offending anyone—the tell is a plan that funds every priority a little and none enough to clear its threshold.
• Chasing all-out growth, which almost sank Crown—the tell is pursuit of size for its own sake displacing focus on the pivotal segment that made the business work.
• Letting past success breed laxity and decline—the tell is potent resources being spread loosely because prior wins made concentration feel unnecessary.

• ▸Diagnosis: the one or two critical issues / pivot points in this situation
• ▸Candidate objectives: every priority currently claiming resources
• ▸Threshold per objective: minimum effort needed before any real payoff
• ▸Resource fit: which objectives our capital/talent can push past threshold
• ▸Chosen few: the small set we will concentrate on (name them)
• ▸Explicit stops: what we will not fund or pursue, freeing resources
• ▸Attention check: can senior leadership hold this many at once? (yes/no)
• ▸Concentration method: how we pulse in time or roll out by region/segment
• ▸Coherence test: do these actions coordinate toward the guiding policy?

1. 1Name the specific challenge first — the single obstacle each action must serve. Coordination only exists relative to a target; without it you get a list of good ideas that pull apart.
2. 2Lay out every proposed action on one page and ask of each: which other action does this reinforce, and which does it undercut? Volvo-Jaguar shows conflict; the Akron/advertising/360-feedback list shows mere non-conflict. You want mutual reinforcement, not just absence of contradiction.
3. 3Trace the reinforcing chain explicitly. Jobs's cuts, the single Power Mac G3, dropping five of six retailers, and Dell-style supply chain each fed the same logic: simplify to stop the cash bleed. Write out the causal links so you can see the actions compounding.
4. 4Kill actions that require conflicting resource commitments or conflicting customer promises. Putting two distinct marques on one platform saves money but destroys the brand equity that justifies each — the coordination test fails because the actions demand incompatible outcomes.
5. 5Assign each action an owner and a shared reference point (the guiding policy), so that when trade-offs arise, silos resolve them toward the same signpost rather than optimizing locally.
6. 6Concentrate resources on the one or two pivot points, not spread across all fronts. Nelson broke enemy coherence by risking lead ships at the decisive point rather than distributing force evenly.
7. 7Re-check after any success: potent resources tend to induce laxity and drift, so re-verify that actions still work in concert rather than having reverted to independent silo agendas.

Jobs's 1997-98 Apple turnaround is coordinated effort made visible. He named one challenge — a product line too complicated while the company bled cash — and every action served it. He replaced the confusing desktop lineup with a single Power Mac G3, dropped five of six national retailers because meeting them meant too many models at too many price points, secured Microsoft Office so the machines stayed usable, and adopted Dell's Asian supply-chain model with short cycle times and negative working capital. None of these was novel — it was 'Business 101' — but they reinforced each other around one logic of simplification. Notably he announced no revenue or profit goals and no messianic vision; the power came from a focused, coordinated set of actions all pointed at the same problem. Contrast Ford's Volvo-Jaguar move, where buying both and forcing them onto a shared platform put actions in direct conflict, diluting each brand.

• If two actions demand incompatible customer promises or resource states, then they are not coordinated — cut or redesign one, however individually sound each is.
• If a list of actions merely doesn't contradict itself but none reinforces another, then you have operational management, not strategy — force each to serve the named challenge or drop it.
• If an action's only 'strategic' quality is that it needs top-management approval, then reclassify it as operations; pay grade does not make coordination.
• If you cannot name the single challenge every action serves, then stop — coordination has no reference point and will collapse into silos.
• If recent success has made resources abundant, then audit for laxity and drift before assuming actions still work in concert.

• Assembling a list of good-but-independent actions (close Akron, spend more on advertising, run 360-feedback). Tell: you can reorder or delete any item without the others noticing.
• Conjoining actions that conflict at the customer level — the safe-Jaguar / sporty-Volvo platform. Tell: your most passionate customers, dealers, and service shops are the ones complaining.
• Substituting broad goals for coordinated action (the 20/20 plan, 'the will to win'). Tell: when asked 'what has to happen?' the answer is drive and ambition, not a lever or mechanism.
• Template-style 'key strategies' (be the firm of choice, delight customers, grow 20%, culture of commitment). Tell: it took three weeks of buy-in but names no obstacle and coordinates nothing.
• Delegating goals while leaving each unit to figure out its own actions. Tell: parts optimize locally toward the same slogan while pulling in different directions — honest label is goal setting, not strategy.

• ▸Named challenge: the single obstacle every action must serve
• ▸Guiding policy / signpost: the direction all actions point toward
• ▸Action list: each proposed policy, resource commitment, or action
• ▸Reinforcement links: for each action, which other action it strengthens
• ▸Conflict flags: for each pair, any incompatible promise or resource state
• ▸Owner + shared reference: who executes and what they resolve trade-offs toward
• ▸Pivot-point concentration: which one or two actions get the bulk of resources
• ▸Cut list: actions that are sound operations but coordinate nothing here