The Innovator's Solution, with a New Foreword

1. 1Characterize the phenomenon first: describe exactly how the product performs against what customers in a given tier actually need — don't jump to prescriptions before you've observed the fit. This mirrors the theory-building sequence of observe, categorize, then explain.
2. 2Categorize the market circumstance into one of two mutually exclusive states: 'not good enough' (a performance gap remains, customers pay for improvements) or 'more than good enough' (a performance surplus exists, customers no longer value additional improvement).
3. 3Identify the salient boundary by testing whether crossing it forces a fundamentally different management technique — if the same approach succeeds in both conditions, the boundary is not meaningful for prediction.
4. 4Confirm the categories are collectively exhaustive and mutually exclusive so you know both what circumstance you're in and what circumstance you're not in; ambiguity here makes your innovation strategy random.
5. 5Match the innovation type to the circumstance: pursue sustaining improvements where performance is 'not good enough,' and shift to convenience, simplicity, or lower cost where performance is 'more than good enough.'
6. 6Re-test the circumstance over time, because tiers migrate: what was 'not good enough' becomes 'more than good enough' as technology overshoots, so watch for the conditions changing and change your methods accordingly.

Packet-switching illustrates a shift across the fit boundary. Early on, packet switching introduced a few seconds of latency delay, which made it 'not good enough' for voice telephony — a real performance gap against what voice customers needed. But that same performance was 'more than good enough' to serve an emerging market: data networks, where the latency didn't matter. Over time the technology improved until the latency delay of a packet-switched voice call became almost imperceptibly slower than a circuit-switched call. At that point the performance crossed the boundary and became good enough for voice, enabling VOIP telephony. The lesson: the identical technology was categorized differently in two market tiers, and the circumstance flipped as performance improved.

• If a performance gap exists in the target tier (customers still pay for improvements), then you are 'not good enough' and sustaining innovation is the right medicine.
• If customers no longer value or will not pay for additional performance, then you are 'more than good enough' and compete on convenience, simplicity, or price instead.
• If the same management technique succeeds regardless of which side of a boundary you're on, then that boundary is not a real category — discard it.
• If you cannot state both the circumstance you are in and the one you are not in, then stop; your categories aren't yet collectively exhaustive and mutually exclusive, and predictions will be random.
• If a technology overshoots what a tier needs over time, then reclassify the circumstance from 'not good enough' to 'more than good enough' and change strategy accordingly.

• Copying attributes instead of understanding circumstance — the 'wings-and-feathers' error: imitating what worked for an 'excellent' company without asking in which circumstance it applies. The tell: advice framed as universal best practice with no boundary conditions.
• Prescribing before diagnosing — handing out the same remedy to every case like the doctor who says 'take two and call me,' having cured his last two patients. The tell: strategy chosen before the performance-vs-need fit is characterized.
• Categorizing by attributes (industry, product vs. service) rather than by circumstance. The tell: managers asking 'does this apply to my industry?' — a sign they're using an attribute scheme that almost never predicts.
• Confusing correlation with causality — assuming performance surplus or gap is stable because it correlated with an outcome once. The tell: no explanation of the causal mechanism, only observed patterns.
• Missing the migration of the boundary over time — treating a 'not good enough' verdict as permanent while the technology quietly overshoots. The tell: no periodic re-test, so you keep investing in improvements customers stopped valuing.

• ▸Target market tier: name the specific tier of customers you're serving
• ▸Job/need in that tier: what performance dimension do these customers actually need improved
• ▸Current product performance: how the product performs on that dimension today
• ▸Gap or surplus: is there a remaining performance gap or an unused surplus
• ▸Willingness to pay for more: will customers pay for additional performance on this dimension (yes/no)
• ▸Circumstance category: mark 'not good enough' or 'more than good enough'
• ▸Boundary test: does crossing this line force a different management technique (yes = salient)
• ▸Matched innovation type: sustaining improvement, or convenience/simplicity/low-cost
• ▸Re-test date: when to recheck whether the technology has overshot and the circumstance flipped

1. 1Map the incumbent's profit model: identify their best customers, the margins those customers generate, and the fixed assets tuned to serve them (e.g., $8B integrated mills, roll film margins). This tells you where they are motivated to fight versus flee.
2. 2Position your product below their margin floor. Aim at new or less-demanding customers who value simpler, more convenient, or cheaper (Dell's cheapest portable, Kodak Funsaver for people with no camera) so that competing costs the incumbent more profit than they'd gain.
3. 3Test whether the asymmetry actually holds. Ask: if the incumbent responded, would they be defending high-margin customers or chasing low-margin ones? If defending their own profitable core (as with Internet banking's identical cost of money), you have a sustaining innovation, not a disruption — abandon or license it.
4. 4Confirm the incumbent's up-market motivation is stronger than their defensive one. Established firms are 'always motivated to go up-market, and almost never motivated to defend the new or low-end markets.' Your foothold should sit in the market they most want to exit.
5. 5Build in an improvement trajectory. Because technological progress outstrips customer needs, plan how the not-good-enough product intersects demanding customers over time (Japanese steelmakers adding continuous casting/basic oxygen furnaces as their auto customers grew).
6. 6If you are the incumbent facing this, house the disruptive effort in an independent, stand-alone profit center for which the innovation is pure opportunity — as newspapers did when they spun off online groups — so it stops replicating the mainstream business.
7. 7Re-test the asymmetry as you climb up-market; the incumbent's flee-or-fight calculation changes when you approach the customers and margins they actually care about.

Minimills disrupted integrated steel mills by exploiting asymmetric motivation. Minimills melt scrap in electric arc furnaces and make steel of any given quality for 20 percent lower cost than an $8B integrated mill. They entered at the bottom — rebar, the least profitable, lowest-quality tier — where integrated mills were happy to cede ground and retreat up-market to higher-margin products. Each time integrated players fled a tier, minimills improved and moved up into the vacated, still-more-profitable segment. The integrated mills did exactly what was in their best and most urgent interest at every step: satisfy their most important customers and invest where profits were most attractive. That rational, profit-seeking behavior is precisely what handed the market to the attacker.

• If the incumbent would have to abandon its most profitable customers to compete, then the asymmetry favors you — proceed as a disruption.
• If competing means defending an existing high-margin core (identical cost structure, same customers, sustaining impact), then expect the incumbent to fight and win — license it or leave it to established firms.
• If the entry point sits in a market the incumbent is eager to exit up-market, then the flee motivation is strong; anchor your foothold there.
• If you are the incumbent, then place the disruption in an independent unit that sees pure opportunity — never inside the mainstream organization whose processes only serve sustaining innovation.
• If the innovation only makes existing customers use it instead of the existing product (online newspaper cannibalizing print), then you're cramming disruption into the mainstream and the asymmetry works against you.

• Cramming the disruptive product into the mainstream market to serve existing customers — the tell: your target advertisers/readers are the same as the incumbent's, and cannibalism fears become self-fulfilling, as with early online newspapers.
• Assuming a lower-cost technology automatically disrupts — the tell: the cost of inputs is similar for all players (banks' cost of money), so the incumbent can adopt it as a sustaining move.
• Housing the disruption inside the core organization — the tell: the new group keeps replicating the mainstream business's services, suppliers, and revenue sources instead of diverging.
• Picking a foothold the incumbent is motivated to defend — the tell: you're offering better performance to demanding, high-margin customers, which triggers the established leader's strongest fighting response and near-certain victory.
• Believing you've disrupted when you merely caught the leader asleep at the switch with a sustaining strategy — the tell: your success depends on the incumbent not noticing rather than on their rational unwillingness to respond.

• ▸Incumbent's best customers: name the high-margin segment they must protect
• ▸Incumbent's profit model: gross margins and fixed assets tied to the core
• ▸Your entry tier: the new-market or low-end customers you target
• ▸Cost/benefit to incumbent of responding: profit they'd gain vs. profit they'd forgo
• ▸Up-market pull: the more profitable segment they are motivated to chase instead
• ▸Flee test (Y/N): would they rationally cede your tier to serve better customers?
• ▸Improvement trajectory: how your not-good-enough product intersects demanding needs over time
• ▸Organizational home (if incumbent): independent profit center or mainstream unit
• ▸Verdict: disruption / sustaining-so-license-it / caught-asleep bet

1. 1Map the venture's opportunity by order size and margin, then ask the concrete question: given our cost structure and revenue base, does winning this opportunity register as attractive to the people who allocate our resources? If it does not, alignment is absent regardless of the opportunity's absolute appeal.
2. 2Size the venture to the opportunity, not the parent to the opportunity. A $50 million business finds a $5 million order energizing; a $50 billion parent finds it a rounding error and starves it. Set the cost structure and headcount so that the small, low-margin orders the opportunity actually generates clear the internal hurdle for attention and funding.
3. 3Trace your resource-allocation process explicitly: budgeting, headcount approval, market research prioritization, and compensation. These processes and the values behind them decide which opportunities look attractive. Confirm that each step rewards, rather than filters out, the low-margin and small-volume work the venture depends on.
4. 4Test whether the opportunity fits the job to be done AND the venture's values simultaneously. VHS won because it aligned with the recording-time job; a venture wins durably only when the opportunity also aligns with what its own allocation process treats as worth pursuing.
5. 5Where the parent's values reject the opportunity, place it in a separate organizational unit whose cost structure and size make the same numbers attractive — do not try to force-fit it into processes tuned for large, high-margin sustaining business.
6. 6Avoid capital whose scale demands returns the opportunity cannot deliver early. Money that needs the venture to be big fast will push it up-market and out of alignment; match the source and size of investment to the small-order, low-margin reality of the target.

AT&T's serial growth failures show misalignment between opportunity and the values of an enormous cost structure. The wireline business kept landing back 'right where it had started,' and 'the company had to grow' — a phrase that repeats because a firm that size needs large, high-margin wins to move its numbers. So it bought McCaw Cellular for $11.6 billion (eventually $15 billion), then TCI and MediaOne for $112 billion, chasing opportunities big enough to register in its allocation process. When Wall Street couldn't value the higher-growth wireless business inside the lower-growth wireline company, AT&T created a separately traded stock in 2000 — an explicit attempt to give the growth opportunity a home whose size and expectations aligned with it. Over roughly ten years AT&T wasted about $50 billion, because the opportunities that could have grown were never sized or housed where the allocation process would treat them as attractive.

• If a target opportunity's order size or margin falls below what your budgeting and headcount processes reward, then house it in a smaller organizational unit rather than the parent.
• If 'the company had to grow' is driving you toward only large acquisitions, then your parent's values are misaligned with disruptive opportunities — expect to overpay and destroy value.
• If an opportunity aligns with the job to be done but not with your allocation process, then fix the organizational home before scaling; alignment with the job alone won't survive internal starvation.
• If Wall Street or internal review cannot value the venture inside the parent, then separate it structurally so its own cost structure sets the hurdle rate.
• If proposed capital requires the venture to be big fast, then decline it — that money will force the venture up-market away from its low-margin core.

• Judging an opportunity by its absolute size instead of by whether your allocation process finds it attractive. Tell: small but real orders keep dying in budget review while everyone agrees the market is 'interesting.'
• Attaching a growth venture to a large parent whose margins make its early revenue invisible. Tell: the venture is chronically under-resourced and repeatedly reframed as 'not material.'
• Confusing job-to-be-done fit with values fit. Tell: the product clearly serves customers, yet the organization keeps deprioritizing it internally.
• Chasing ever-larger acquisitions because the core keeps stalling. Tell: the phrase 'we had to grow' precedes each deal, and each deal returns you to your starting point at great cost.
• Taking capital sized to the parent's ambitions, not the opportunity's reality. Tell: pressure mounts to move up-market and abandon the small, low-margin customers that were the entry point.

• ▸Opportunity profile: typical order size and gross margin
• ▸Venture revenue base: current size of the unit that will pursue it
• ▸Internal hurdle: minimum deal size/margin your budgeting rewards
• ▸Alignment gap: does the opportunity clear the hurdle? (yes/no + gap)
• ▸Allocation processes touched: budgeting, headcount, research, comp
• ▸Organizational home: parent unit vs. separate/spun-out unit
• ▸Job-to-be-done fit: which specific job the opportunity serves
• ▸Capital source and size: does required return force up-market drift?
• ▸Decision: pursue in current home / re-house / decline

1. 1Define the 'job to be done' as a circumstance, not a customer demographic or product category. State the situation the customer is in and the progress they are trying to make, e.g., 'a commuter needs to fill a boring 20-minute drive and stave off hunger until lunch' rather than 'coffee drinkers aged 25-40.'
2. 2Inventory the customer's current alternatives honestly, including nonconsumption and makeshift solutions (a banana, a bagel, a candy bar, skipping the job entirely). These, not branded rivals, are your real competition for the job.
3. 3Rate how well each alternative gets the specific job done under real conditions of use — one-handed operation, mess, wait time, duration of satisfaction — so you can see where every existing solution fails the customer.
4. 4Design your offering to beat those alternatives on the exact dimensions the job demands, not on generic 'better/cheaper' attributes that matter to no one in the moment of hiring.
5. 5Test perceived superiority by observing whether customers actually 'hire' your offering over the makeshift solution in the moment, and whether they 'fire' the old alternative. Don't rely on stated preferences or generic satisfaction surveys.
6. 6Guard the concept against the internal shaping process: middle managers will recast the job-focused idea into a me-too improvement for existing customers backed by data from past successes. Force the value case to rest on the job, not on existing-customer market-size projections.
7. 7Iterate the offering against the tell — if customers keep reverting to the makeshift solution, the job is misdefined or your solution fails a dimension you haven't identified.

A fast-food chain wanted to sell more milkshakes and studied buyers by demographic, then improved the shakes on those dimensions — nothing changed. Reframing around the job to be done revealed that a large share of shakes were bought before 8 a.m. by lone commuters who 'hired' the shake to make a long, dull drive interesting and to hold off hunger until noon. Their alternatives — bananas (gone too fast, no lasting fullness), doughnuts (crumbly, greasy fingers on the wheel), bagels (dry, need two hands) — all lost to a thick shake that lasts twenty minutes and is drunk one-handed through a straw. Understanding that job, the fix was to make the shake even thicker and more viscous and to move it to a self-serve machine so hurried commuters could grab it fast. That is a different product change than any demographic-based improvement would have suggested, because the competition being beaten was nonconsumption and makeshift breakfasts, not other milkshakes.

• If you can't name a nonconsumption or makeshift alternative you're beating, then you haven't found the real job — keep observing.
• If the value case depends on data about existing customers and similar past products, then you're likely shaping a me-too innovation, not a job-based one.
• If customers say they prefer your offering but keep hiring the old makeshift solution in the moment, then trust the behavior — the perceived value isn't there yet.
• If your improvement raises an attribute the job doesn't actually require (a demographic-driven feature), then it will not move purchase behavior — stop investing in it.
• If the offering targets a job served poorly by all current alternatives and today's market looks small, then that small market may be the disruptive opening, not a reason to kill it.

• Categorizing by product or demographic instead of by job — the tell: your improvements delight focus groups but sales don't move.
• Benchmarking only against branded competitors while ignoring the banana, the bagel, and doing-without — the tell: you 'win' comparisons yet customers still don't switch.
• Letting middle managers reshape the concept into 'make existing customers happier' to secure funding — the tell: the business plan now rests on assured, sizable existing markets.
• Optimizing average satisfaction rather than performance under the actual conditions of use (one-handed, twenty-minute drive) — the tell: praise in the abstract, abandonment in context.
• Killing a job-based offering because today's market is small — the tell: rejection language cites insufficient market-size data, exactly the weeding-out the shaping process is built to do.

• ▸Job statement: the circumstance + progress the customer seeks, in one sentence
• ▸Conditions of use: when/where/how the job happens (time, hands, setting)
• ▸Current alternatives: list all, including nonconsumption and makeshift fixes
• ▸Where each alternative fails: the specific dimension it falls short on
• ▸Your offering's edge: which failing dimensions you beat and how
• ▸Hire/fire evidence: observed behavior showing switching, not stated preference
• ▸Wrong-attribute check: features you're tempted to add that the job doesn't need
• ▸Shaping-risk flag: is the value case leaning on existing-customer/market-size data?

1. 1Separate the parts of your strategy that must be fixed from the parts that must stay flexible. In a new venture the viable business model is unknown at launch, so treat your initial plan as a hypothesis, not a mandate to execute rigidly.
2. 2Fund the venture in stages tied to what you learn, not to milestones of the original plan. AT&T poured $7.4 billion into NCR, then $2 billion more 'trying to make the acquisition work' before abandoning the vision in 1996 — that is the pattern of doubling down on a flawed plan instead of pivoting.
3. 3Read market feedback as data about which business model works, not as an implementation problem to overcome. When AT&T's computer division lost $200 million annually 'unassailable from the outside,' the signal was to change strategy, not bet bigger.
4. 4Convene the team explicitly around the question of Chapter 8: 'When is flexibility important, and when will flexibility cause us to fail?' Decide, before committing capital, whether you are still searching for the model (flexibility) or executing a proven one (commitment).
5. 5Protect the venture from the corporate growth imperative that forces premature scaling. The market discounts expected growth into today's stock price, so the pressure to hit a number fast can push you to lock in a plan before feedback has revealed a viable one.
6. 6Keep each pivot small and cheap enough that a wrong turn is survivable. Losing $50 billion over a decade, as AT&T did, comes from betting big on unvalidated strategies; converging on a model requires many low-cost course corrections.

AT&T's decade of failed growth is the anti-pattern for adaptability. After the 1984 divestiture it bet that computing and telephony would converge, built its own computer division losing $200 million a year, then in 1991 bought NCR for $7.4 billion and lost another $2 billion 'trying to make the acquisition work' — rigidly executing a flawed thesis rather than pivoting off the feedback. It finally sold NCR in 1996 for $3.4 billion, roughly a third of what it invested. But 'the company had to grow,' so it repeated the pattern with wireless ($15 billion in, valued at $10.6 billion by 2000) and then broadband (TCI and MediaOne for $112 billion combined), selling the cable assets to Comcast for $72 billion in 2000. In just over ten years AT&T 'wasted about $50 billion,' each time treating a failing plan as an implementation problem to be solved with more money rather than a signal to converge on a different model.

• If early market feedback contradicts your revenue or adoption assumptions, then treat the plan as the thing to change, not the market — pivot before you invest more.
• If you cannot yet name which initial customers form a viable foundation, then you are still searching and flexibility must dominate; do not commit large capital.
• If losses are growing because you are 'trying to make it work,' then stop funding the original thesis — that phrase is the tell of rigid execution.
• If a proven model is finally identified and repeatable, then switch to commitment and execute; flexibility at this stage causes failure.
• If the corporate growth imperative is pushing you to scale before the model is validated, then resist — pursuing growth the wrong way is worse than no growth at all.

• Escalating commitment: betting bigger on a failing thesis. Tell — spending framed as 'trying to make the acquisition work' rather than testing whether the model is right (AT&T + NCR).
• Confusing risk appetite with adaptability. Executives 'evidenced a strong stomach for risk' yet still failed; a bold bet on the wrong fixed plan is not a pivot.
• Serial repetition of the same pattern across ventures. Tell — 'the company had to grow' drives another big, unvalidated bet immediately after the last one failed (AT&T's computing, then wireless, then broadband).
• Letting the growth imperative dictate timing so you lock in a strategy before feedback exists. Tell — the plan is justified by the stock price needing a number, not by validated customer demand.
• Treating outcomes as pure luck ('creating growth is simply unpredictable') and therefore not studying the process, so you never learn from feedback well enough to pivot deliberately.

• ▸Initial hypothesis: state the business model you assumed at launch, in one sentence.
• ▸Market feedback observed: the specific customer/revenue/adoption data since last review.
• ▸Contradiction check: which assumption does the feedback confirm or break?
• ▸Search vs. execute: are we still finding the model (flexibility) or running a proven one (commitment)?
• ▸Capital-at-risk this stage: the amount committed and what learning it must buy.
• ▸'Making it work' test: are new dollars testing the model or forcing a flawed plan? (yes = stop).
• ▸Pivot options: two concrete alternative models the feedback points toward.
• ▸Decision + trigger: pivot or persevere, and the metric that will reverse this call.

1. 1Separate the two things investors demand: current profitability and future growth. Track them independently so you don't paper over a growth vacuum with a temporary earnings rebound like Cabot did.
2. 2Diagnose whether your growth initiative fits a disruptive circumstance or a sustaining one. Disruptors align every process and decision to fit the disruptive circumstance; sustaining bets into a rival's core get killed.
3. 3Choose the initial foothold market by profitability asymmetry, not size. Enter where the incumbent is relieved to flee (low-margin-to-them, attractive-to-you), as minimills did in rebar, then move up-market where margins are richer.
4. 4Model what happens after you win. If victory means low-cost producer versus low-cost producer in a commodity, price collapses and no one earns attractive returns—plan the up-market migration before you win the current tier.
5. 5Set the growth target against the rate investors have already discounted into the stock price. Only growth that exceeds the priced-in expectation delivers above-market returns, so size initiatives to beat that hurdle, not just to grow.
6. 6Stage the capital: take money that demands early profitability while impatient for growth, and refuse capital whose timeline forces you to scale before the model is proven.
7. 7Sequence the nine decisions—product, initial customer, scope, commoditization defense, organization, strategy process, funding, CEO role—as aligned choices, not independent bets, because misalignment on any one lets competitors kill the venture.

The minimills illustrate profitability-through-disruptive-foothold done right, then its trap. They entered rebar, the least attractive tier, where a 20 percent cost advantage let them earn attractive profits against higher-cost integrated mills, who happily reconfigured lines toward richer products. By 1979 the minimills drove the last integrated mill out—and rebar's price collapsed 20 percent, so low-cost minimill versus low-cost minimill earned no attractive profit. The lesson: their profitability depended on competing against a higher-cost incumbent, not on the segment itself. Relief came from looking up-market to angle iron and thicker bars where integrated mills earned ~12 percent gross margins in a market twice as large. Sustained venture profitability came from repeatedly migrating up-tier, not from winning any single commodity tier.

• If your growth plan bets into a competitor's higher-margin core on sustaining terms, then expect to be killed—reshape the venture to be disruptive first.
• If entering a tier means eventually low-cost-you versus low-cost-you in a commodity, then plan the up-market move before you win, because winning alone destroys margin.
• If a temporary earnings rebound comes from supply/demand shifts rather than new growth (as Cabot's carbon black did 1991–1995), then don't mistake it for a solved growth problem.
• If the growth you can deliver merely matches what investors already discounted into the price, then you have not created shareholder value—only exceeding the priced-in rate counts.
• If capital is impatient for scale but patient for profit, then it is bad money for an unproven model; take money patient for growth and impatient for profit.

• Chasing growth by ever-larger acquisitions near the core and calling it strategy—AT&T wasted ~$50B across NCR, McCaw, and TCI/MediaOne. Tell: each deal justified by 'the company had to grow' rather than by a fit-to-circumstance thesis.
• Building growth platforms after the core matures, then sacking management when losses drag earnings—Cabot's specialty metals/ceramics. Tell: Wall Street first triples then halves the stock, board mandates 'refocus on core,' leaving successors right back where predecessors started.
• Treating a profitability rebound as sustained success. Tell: earnings recover from external supply/demand luck, not from any new growing business.
• Assuming growth is irreducibly unpredictable and managing odds (let a thousand flowers bloom, fail fast). Tell: you rely on portfolio luck—'two of ten hit home runs'—instead of understanding the process inside the black box.
• Choosing the foothold market by size or prestige instead of profitability asymmetry. Tell: you enter where the incumbent will fight hard rather than flee relieved.

• ▸Growth thesis: why this venture, in one sentence (not 'we had to grow')
• ▸Circumstance: disruptive or sustaining, and the evidence
• ▸Foothold market: where does the incumbent flee relieved, and its margin-to-you vs margin-to-them
• ▸Post-win outcome: what margins remain once you defeat current rivals
• ▸Up-market path: the richer, larger tier you migrate to next and its gross margin
• ▸Investor hurdle: growth rate already priced into the stock you must exceed
• ▸Capital fit: is the money patient for growth / impatient for profit?
• ▸Alignment check: which of the nine decisions is currently misaligned

1. 1Map the full value chain of your product, from raw inputs through the finished good to the end customer's job. Mark at each stage whether current performance is already good enough for customers or still not good enough.
2. 2Locate where commoditization is actively occurring — the stage where products have become interchangeable and margins are collapsing (disk drives fell from 60% to 15% gross margins even while getting sixty times better). This is the signal that differentiation has migrated elsewhere.
3. 3Find the reciprocal de-commoditization stage: whenever commoditization destroys differentiability in one link, a proprietary, not-good-enough link opens up adjacent to it. That adjacent link is where profit is moving.
4. 4Position your growth business at the spot where performance is not yet good enough, because that is where proprietary, interdependent product architectures still command premium pricing.
5. 5Continuously re-scan, because the locus of differentiation shifts as each new wave of disruption washes through. Skate to where the money is going in the value chain, not where it presently sits.
6. 6Test each candidate improvement against the customer's willingness to pay: only pursue improvements customers 'enthusiastically reward with premium prices' and drop those greeted with indifference.
7. 7Reposition established businesses proactively toward the de-commoditizing link rather than defending the commoditized one, before margins fully erode.

The one-gigabyte 3.5-inch disk drive is the cautionary case. When introduced in 1992, these drives let manufacturers earn 60 percent gross margins because the ability to mechanically position a head across tracks only 0.00008 inch apart without misreading was genuinely proprietary and not-good-enough. Yet within years, drives sixty times better were earning only 15 percent margins — the product became an undifferentiable commodity despite being a mechanical and microelectronic marvel. The lesson is not that the engineering stopped being hard; it is that 'good enough' performance turned the differentiable attribute into a commodity attribute. Meanwhile, de-commoditization was occurring in an adjacent link of the chain, where performance was still not good enough. A venture avoids the drive-maker's fate by building at that adjacent link rather than at the one where over-served performance has erased the customer's willingness to pay a premium.

• If performance at your stage has become more than good enough for the customer's job, then expect commoditization and margin collapse — move to where performance is still not good enough.
• If a customer greets an improvement with indifference rather than premium pricing, then that dimension is already commoditized — stop investing there.
• If commoditization is visibly destroying margins in one link of the value chain, then look for the reciprocal de-commoditizing link nearby and reposition toward it.
• If your product architecture can be assembled from interchangeable modules, then differentiation has left your link — seek the interdependent, proprietary link instead.
• If a segment is not yet good enough, then an interdependent, differentiated design will command attractive returns — build there.

• Believing commoditization is inevitable fate and resigning to it — the tell is executives defending falling margins as 'unfair' rather than scanning the chain for where differentiation moved (as with the disk-drive lament).
• Continuing to invest in engineering marvels after performance has overshot what customers value — the tell is spectacular technical improvement accompanied by shrinking gross margins (sixty-times-better drives at 15% margins).
• Chasing growth by acquiring adjacent platforms without checking whether the target link is de-commoditizing — the tell is investor enthusiasm at purchase followed by margin drag, as when Cabot's specialty-metals and ceramics bets triggered losses and a halved stock, forcing a retreat to the core.
• Fighting to keep differentiation at the commoditized link rather than repositioning to the de-commoditizing one — the tell is defending the historic core while profits visibly migrate down the chain.
• Confusing product improvement with premium-worthy improvement — the tell is launching upgrades customers meet with indifference instead of premium prices.

• ▸Value-chain stages: list every link from inputs to the customer's job
• ▸Performance vs. need at each stage: mark 'good enough' or 'not good enough'
• ▸Margin trend per stage: rising, flat, or collapsing
• ▸Commoditizing link: the stage where products are interchangeable and margins fall
• ▸Reciprocal de-commoditizing link: the adjacent not-good-enough stage capturing profit
• ▸Customer premium test: does the target improvement earn premium pricing or indifference?
• ▸Architecture check: is the link modular (commodity) or interdependent (proprietary)?
• ▸Reposition action: where will you build or move the venture to catch the profit
• ▸Re-scan date: when you will re-run this because the locus shifts with each disruption

1. 1Accept the base rate before you plan: only about 5% of Fortune 50 firms sustain growth and another 4% reignite it after a stall, so treat a 10% probability of sustained growth as a generous ceiling, not a stretch goal.
2. 2Stop explaining failure with the three false causes — below-average managers, risk aversion, or irreducible unpredictability — because the evidence refutes all three (executives bet billions on innovations like IBM's System 360 and Corning's optical fiber). Instead diagnose the process by which ideas get shaped inside your firm.
3. 3Audit your middle-manager idea-shaping filter. Identify what evidence your midlevel managers demand before backing a proposal — typically market-size data and endorsements from existing large customers — and recognize that this filter systematically kills disruptive, new-market ventures that cannot show that evidence.
4. 4Screen each candidate venture against the disruption test questions: Is there a large population of nonconsumers blocked by cost or skill (new-market disruption), or overserved low-end customers willing to accept fewer features for a lower price (low-end disruption)? If neither, expect a sustaining fight the incumbent wins.
5. 5Confirm a viable low-cost business model exists at the required price point before committing — a discount price alone is not disruption if your cost structure (e.g., the cost of money for a bank) matches incumbents'.
6. 6Set the growth bar against market expectations, not against last year. Because the market prices in an expected 'fade,' merely meeting expectations generates no meaningful shareholder reward; you must exceed the discounted growth rate to move the stock.
7. 7Repeat deliberately: build the capability to launch a sequence of disruptive ventures rather than betting the firm on a single home run, since the parent's growth rate depends on a stream, not one win.

A large toy company and BIG face the same market, yet BIG repeatedly finds, develops, and licenses a sequence of exciting growth products while the toy company suffers a chronic 'dearth of opportunities.' Managers first blamed insufficient creativity or executive risk aversion, so they cycled a parade of new people through the toy company — none cracked it. The real cause is the shaping process: midlevel managers there refuse to back a concept whose market is not assured, because a failed launch wastes millions, so they demand credible market-size data and feedback from significant existing customers. That evidence can only come from current customers and existing markets, which screens out precisely the disruptive, new-market ideas that produce sustained growth. BIG's structure — developing to a working prototype and licensing out — bypasses that filter. The lesson: the parent firm's growth stalls not from bad people but from a good, self-protecting process doing exactly what it was designed to do.

• If a proposed venture cannot document market size using existing-customer data, then do not route it through the standard funding filter — house it where it can survive without that evidence, or you will kill it.
• If the target market is dominated by nonconsumers blocked by cost or skill (e.g., hundreds of millions without residential air conditioning in China), then pursue a new-market disruption and design the cheap, simplified business model first.
• If you cannot construct a low-cost business model that earns attractive profit at the discount price required to win low-end business, then abandon the low-end disruption play (as with Internet banking, where the cost of money is similar for all banks).
• If the venture only offers a better product to customers already unwilling to pay premium prices, or relies on an advantage competitors can copy (like low-cost labor), then expect no durable growth.
• If a business is merely meeting the growth rate already discounted into the share price, then treat it as failing to create shareholder value, regardless of positive absolute growth.

• Blaming stalled growth on 'below-average' or uncreative managers — the tell is a rotating parade of new hires who each fail to fix the problem, proving the fault is in the process, not the people.
• Assuming risk aversion is the barrier — the tell is executives who have already bet billions on big innovations (System 360, Kevlar, optical fiber) yet still cannot sustain growth.
• Treating new-business creation as inherently unpredictable and defaulting to 'let a thousand flowers bloom' or 'fail fast' — the tell is a venture-capital-style portfolio logic (2 fail, 6 walking wounded, 2 home runs) substituting for a diagnosis of which strategy will win.
• Demanding hard market data on ventures aimed at nonconsumers — the tell is proposals dying in middle management for lack of 'credible' numbers about markets that do not yet exist.
• Averaging feedback across a demographic segment and building to the mean — the tell is a 'one-size-fits-none' product, like a milkshake reformulated to please an averaged segment that serves neither the commuter's job nor the parent's job.

• ▸Venture name: the candidate new business in one line.
• ▸Disruption type: new-market, low-end, or neither (sustaining) — state which and why.
• ▸Nonconsumption / overserved evidence: who is blocked by cost or skill, or overserved at the low end.
• ▸Low-cost business model check: can you profit at the required discount price? yes/no + the cost driver.
• ▸Shaping-filter risk: what standard funding evidence this idea cannot supply, and where you'll house it to protect it.
• ▸Market-expectation hurdle: the growth rate already discounted into the parent's share price that this must beat.
• ▸Sequence fit: how this venture extends a repeatable stream of disruptive launches, not a one-off bet.
• ▸Go / No-Go: decision plus the single strongest reason.

1. 1Run the new-market disruption litmus first: ask whether a large population historically lacked the money, equipment, or skill to do this thing and so went without or paid an expert, and whether current use forces customers to an inconvenient, centralized location. If either (ideally both) is yes, shape the idea toward new-market disruption competing against nonconsumption.
2. 2If the new-market tests fail, test for low-end disruption: identify customers at the bottom of the existing value network who are overserved — they don't need the full performance package and would accept a simpler, cheaper offering. Confirm you can serve them profitably through a different business-model formula.
3. 3Define the value network explicitly: fix the performance metric that matters in that context, the target cost structure, and the inventory/asset-turn model. New-market disruptions require a new vertical axis (simplicity, portability, cost), not the incumbent's performance yardstick.
4. 4Manage the shaping process consciously rather than leaving it to whoever writes business plans — most sustaining ideas could be reshaped into disruptive plans with greater growth potential, but only if executives force the litmus questions before funding.
5. 5Check the asymmetry of motivation: verify that when you attack, the incumbent is motivated to flee upmarket toward higher margins rather than defend. If the numbers make retreat rational for them, you have a defensible entry.
6. 6For an incumbent choosing to respond, sequence framing: frame the disruption as a threat during resource allocation to win top-level commitment, then move commercialization to an autonomous unit that frames it as an opportunity and can target nonconsumers instead of cramming it into the mainstream market.
7. 7Build the business model to earn acceptable profit through a different formula (e.g., faster asset turns at lower margins), not by matching incumbent margins. Then improve performance up the trajectory to move into the mainstream over time.

Intuit's QuickBooks is a new-market disruption. Established accounting-software leaders let small-business managers run sophisticated analytical reports; QuickBooks, derived from the personal-finance product Quicken, simply helped very small owners (mostly fewer than five employees) keep track of their cash. This targeted a large population that historically did not keep books on a computer at all — classic nonconsumption. Within two years QuickBooks took 85 percent of the small-business accounting market, mainly by creating new growth rather than stealing existing customers. The stealing came later, as functionality improved and the product climbed the performance trajectory. The strategic choice was to compete on a different value proposition (simplicity, tracking cash) against nonconsumers, not to out-feature the incumbents on their own metric.

• If a large population lacks money/skill/equipment and goes without, or must travel to a centralized location to consume, then shape the idea as a new-market disruption competing against nonconsumption.
• If the target customers are the overserved low end of an existing market who would accept less performance for lower price, then pursue low-end disruption with a different business-model profit formula.
• If entering forces the incumbent to choose between defending low margins and retreating to higher-margin tiers, then attack — they are motivated to flee, not fight.
• If you are an incumbent, then frame the project as a threat to secure commitment, and only afterward hand it to an autonomous organization that can frame it as an opportunity.
• If the idea only improves performance for your existing best customers on the existing metric, then it is sustaining — expect to fight incumbents head-on with no asymmetry in your favor.

• Cramming a disruptive technology into the mainstream market to protect current customers — the tell is threat framing that makes cannibalism a self-fulfilling prophecy, as newspapers did by putting the paper online for the same readers and advertisers.
• Leaving the shaping process to dispersed, instinctive business-plan writers, so disruptive potential gets defaulted into a sustaining plan — the tell is a plan measured only on the incumbent's performance axis with no new value network.
• Concluding disruption failed because the incumbent wasn't instantly killed — the tell is treating disruption as an event rather than a process whose forces operate continuously, sometimes over decades.
• Concluding the theory is false because an incumbent caught the wave — the tell is ignoring that management harnessed the forces (a flying plane doesn't disprove gravity); the leader succeeded through autonomous framing, not by the forces disappearing.
• Trying to match incumbent margins instead of building a different profit formula — the tell is margins plummeting when you price below the incumbent, because you kept a cost structure built for the old value network rather than one earning acceptable ROCII through faster turns.

• ▸Nonconsumption population: who historically lacked money/skill/equipment and went without?
• ▸Convenience gap: does current use require an inconvenient, centralized location?
• ▸New performance axis: what metric (simplicity, portability, cost) will you win on instead of the incumbent's?
• ▸Low-end target: which overserved customers at the bottom would accept less for cheaper?
• ▸Profit formula: what margins x asset turns produce acceptable ROCII in your model?
• ▸Incumbent motivation: will they flee upmarket or defend? (state the margin math)
• ▸Value-network fit: cost structure, suppliers, and channels this business requires
• ▸Framing plan (incumbents only): threat frame for commitment, then autonomous unit for opportunity
• ▸Classification: new-market / low-end / hybrid — and the evidence for it

1. 1Define the two architectural poles for your product: interdependent/proprietary (components co-designed so performance is wrung from tight fit) versus modular/open (standardized interfaces so pieces can be mixed, customized, and swapped by users).
2. 2Locate where your target customer sits on the performance trajectory: if the mainstream is not yet good enough on the metric they care about, weight toward interdependent architecture; if performance has overshot what they can use, weight toward modular/open.
3. 3Identify the value network you intend to serve—the customers, cost structure, and channel—and specify what 'performance' means in that context, since it differs from the original network's metric.
4. 4For a new-market disruption, design the architecture around simplicity, portability, and low cost so nonconsumers with less money or skill can self-serve; open, modular design lets those users create their own outcomes without expertise.
5. 5For a low-end disruption, adopt a modular/open architecture that plugs into a cheaper business model (higher inventory turns, lower gross margin) so you profit at prices the incumbent must flee.
6. 6Match the architecture to the enabling technology: if a shift (e.g., microprocessors, relational databases, electronic ticketing) lets users do their own configuration or reporting, choose modular; if you are still chasing raw performance in a demanding application, keep it interdependent.
7. 7House the disruptive architecture in an autonomous unit with its own cost structure and processes so the mainstream business does not force it back toward the interdependent standard.
8. 8Re-evaluate over time: as an initially modular new-market product improves, it pulls applications out of the old value network—plan the architecture's evolution rather than freezing it.

Oracle's relational database beat Cullinet and IBM by choosing a modular, open architecture over an interdependent one. IBM's and Cullinet's hierarchical/transactional databases ran on mainframes and were built to generate standard financial reports—performance-optimized for a demanding, expert-served application. Oracle's relational, modular architecture ran on cheaper minicomputers and later microprocessor-based machines, and, decisively, let users without deep programming expertise create their own custom reports and analyses. That architectural choice traded some raw mainframe performance for flexibility and customization the incumbents' interdependent design could not offer. The result was disruption: Oracle served users the incumbents ignored, then improved. The lesson is that the architecture choice was the strategy—modularity opened a new population of self-serving users.

• If the mainstream customer is not yet satisfied on the core performance metric, then choose interdependent/proprietary architecture to extract maximum performance.
• If performance has overshot what customers can absorb, then choose modular/open architecture to compete on flexibility, speed, cost, and customization.
• If your target is a nonconsumer who lacks money or skill, then make it modular and simple so they can configure and use it themselves—as QuickBooks and Oracle let non-experts do their own work.
• If you are attacking overserved low-end customers, then pair modular architecture with a leaner business model (faster inventory turns, acceptable lower margins) so the incumbent flees rather than fights.
• If you must beat an incumbent in the exact application it already serves, then only an interdependent architecture that is genuinely better on that metric will do—and the odds are very, very low (the EMC-over-IBM exception).

• Choosing modular/open when the market is still not-good-enough on performance—the tell: customers reject you as 'not powerful/accurate enough' and stay with the interdependent incumbent.
• Choosing interdependent/proprietary when customers are overshot—the tell: you keep adding performance nobody pays for while a cheaper modular entrant takes your low end.
• Trying to win in the incumbent's own application with a merely-different architecture—the tell: you're competing head-on where a large, resourced rival will fight and win (only EMC-style genuine superiority survives).
• Bolting the disruptive modular architecture onto the mainstream business model—the tell: the parent's cost structure and processes pull the design back toward the proprietary standard; the cure is an autonomous unit like HP's separate ink-jet business.
• Treating the architecture choice as permanent—the tell: you don't plan for the modular product improving and pulling applications out of the old value network (as plastics did to steel and wood), so you're surprised when it moves upmarket.

• ▸Target value network: name the customers, cost structure, and channel you will serve.
• ▸Performance metric in this network: state what 'good' means here (differs from the incumbent's metric).
• ▸Not-good-enough vs. overshot: mark whether mainstream customers still want more performance or have too much.
• ▸Disruption type: new-market (compete with nonconsumption) or low-end (attack overserved customers).
• ▸Architecture pole: interdependent/proprietary (performance) or modular/open (flexibility, speed, customization).
• ▸User self-service test: can a non-expert configure and get outcomes themselves? (yes favors modular).
• ▸Business-model fit: state gross margin and inventory-turn/cost model the architecture must profit within.
• ▸Organizational home: mainstream unit or autonomous unit with its own processes.
• ▸Evolution path: how the architecture improves and which applications it will pull from the old network next.

1. 1Classify the venture on the disruption spectrum first: determine whether it is a low-end disruption (attacking overserved customers at the bottom of the existing value network) or a new-market disruption (competing against nonconsumption in a new value network). This classification drives the organizational choice.
2. 2Test whether the venture's required cost structure conflicts with the mainstream. Ask: at what gross margin and inventory-turn (or equivalent) does this new business make acceptable profit? If those numbers are lower than what the core organization needs to justify its resource allocation, the venture cannot survive inside the mainstream.
3. 3Map the resource allocation battle explicitly. Identify the scarce resource the venture competes for (e.g., retail floor space, engineering attention, sales force time) and forecast how the mainstream's profit-maximizing logic will starve it — the same logic that makes department stores flee low-end goods for 50% cosmetics margins.
4. 4If the venture requires a distinct business model, cost structure, or set of values, house it in a structurally and managerially autonomous unit with its own charter, separated physically and organizationally from the core.
5. 5Give the autonomous unit an unfettered charter to build a genuinely new business model, not a mandate to replicate the parent's processes. IBM's PC unit in Florida and HP's ink-jet division separate from Boise are the template.
6. 6Match the unit's profit formula to its own value network — let it define performance on the axis its new customers value, not the axis the mainstream values, and let it set prices and margins accordingly.
7. 7Decide CEO involvement: keep hands off the operating details but protect the unit's independence from the core's resource-allocation gravity, since only senior authority can override that pull.

Hewlett-Packard faced the ink-jet printer, a new-market disruption relative to its own laser jet business — compact, light, low-cost enough that a whole new population of computer owners, primarily students, could individually own a printer, even though early ink-jets were slow and produced fuzzy images. Rather than house ink-jet development inside its Boise laser jet division, HP created a completely independent ink-jet business unit. That separation let the ink-jet unit compete against nonconsumption and improve on its own trajectory without being judged against laser jet margins or starved by the core's resource-allocation priorities. Ink-jets eventually became the mainstream printer of choice, pushing laser jets to the high end. Because HP owned the disrupting unit, it stayed atop the industry rather than being displaced by an entrant — the same pattern by which IBM survived minicomputers and PCs through separate business units.

• If the venture's viable profit formula (margins, asset turns, cost structure) is lower than what the mainstream needs, then house it in an autonomous unit — the core will rationally flee or starve it.
• If the venture is a new-market disruption competing against nonconsumption, then give it an independent charter to define its own value network and performance metrics rather than the parent's.
• If the venture competes for the same scarce resource as high-margin core activities (floor space, engineering, sales attention), then separate it, because the mainstream's resource logic will always reallocate away from the lower-margin newcomer.
• If the disruption is sustaining relative to your existing business and fits your current cost structure and customers, then it can be pursued inside the mainstream organization.
• If you cannot give the new unit an unfettered charter and structural independence, then do not expect it to survive the parent's values and processes.

• Housing a lower-margin disruptive venture inside the mainstream — the tell: the core keeps reallocating resources toward higher-margin opportunities (like department stores choosing 50%-margin cosmetics over defending 20%-margin hard goods), and the venture is quietly starved.
• Judging the new unit by the parent's performance metrics — the tell: early products look inferior (slow, fuzzy ink-jets) on the axis the core values, so leadership kills it before its own value network matures.
• Giving the unit a constrained charter that forces it to replicate the parent's business model — the tell: the new business ends up with the same cost structure it was supposed to escape and can't reach profitable nonconsumers.
• Framing the venture's market using the parent's IT data (products, existing customers, organizational units) — the tell: you aim at phantom targets and size the opportunity as too small because nonconsumers don't appear in your reports.
• Assuming size and resources will win the fight, so keeping the venture in-house for 'synergy' — the tell: the well-resourced incumbent still flees the low end while the separate entrant model takes root uncontested.

• ▸Disruption type: low-end, new-market, or hybrid — state which and why
• ▸Target customers: existing overserved customers vs. current nonconsumers — name them
• ▸Required profit formula: target gross margin and asset/inventory turns for the venture
• ▸Core's required profit formula: margins and turns the mainstream needs to justify allocation
• ▸Cost-structure conflict test: is the venture's formula below the core's threshold? (yes/no)
• ▸Scarce resource at stake: the shared resource the core will reallocate away (e.g., floor space, engineering)
• ▸Placement decision: mainstream vs. autonomous unit — and the trigger that justifies it
• ▸Charter scope: what the unit is free to do differently (model, metrics, pricing, customers)
• ▸CEO/governance role: what leadership protects (independence) vs. leaves alone (operations)

1. 1Classify the venture's near-term reality: does it have a proven, profitable business model yet, or is it still searching for one? This determines which funding profile fits.
2. 2For an early, unproven new-growth business, seek capital that is 'patient for growth but impatient for profit' — money that lets you stay small while you find a viable, profitable formula, and that pressures you to become profitable early rather than scale losses.
3. 3For an established core business with a known profit engine, accept capital that is 'impatient for growth but patient for profit' — money that funds aggressive scaling because the profit formula is already proven.
4. 4Map the investor's stated expectations to their behavior: read what they will demand in the first 18-24 months (scale or profitability?) and confirm it matches the venture's stage.
5. 5Screen out the mismatch: refuse large slugs of growth-hungry capital for a business that has not yet discovered how it makes money, because that money forces premature scaling before the model works.
6. 6Sequence sources by stage — take patient-for-growth/impatient-for-profit money to discover the model, then switch to growth capital once the profit formula is proven and repeatable.
7. 7Guard the venture against parent-company or investor demands to grow big fast when what it needs is time to find the right customers and cost structure.

AT&T illustrates the wrong funding logic applied under pressure. The company 'had to grow' to build shareholder value, so it repeatedly deployed enormous, impatient-for-growth capital: $15 billion into wireless via McCaw Cellular, then $112 billion for TCI and MediaOne to become the largest U.S. cable operator. Each move demanded rapid scale rather than tolerating small, profitable experimentation to find a workable model. When implementation and integration failed, AT&T sold the cable assets to Comcast for $72 billion, and over roughly a decade wasted about $50 billion. The capital was impatient for growth in ventures that had not yet proven they could profit — the opposite of what an unproven business needs. Contrast Intuit's QuickBooks, which grew by first serving a new market of tiny businesses and earning its position (85 percent share in two years) before scaling into established firms' customers.

• If the business model is still unproven, then take capital patient for growth but impatient for profit — force early profitability, stay small.
• If the profit formula is proven and repeatable, then accept capital impatient for growth but patient for profit — scale hard.
• If investors demand rapid scale before you know how the venture makes money, then that capital is the kiss of death — decline it.
• If you feel pressure to grow big fast (the 'we had to grow' reflex), then treat it as a signal to check whether the model is proven before deploying growth capital.
• If you can only satisfy investors by burning larger losses at bigger scale, then you have the wrong funding profile for this stage.

• Deploying massive growth-hungry capital into an unproven venture; the tell is escalating acquisition spend chased by the phrase 'the company had to grow' (AT&T's $50 billion).
• Matching capital to the parent's growth needs instead of the venture's stage; the tell is investors valuing scale over a path to profit.
• Assuming outcomes are pure luck (the VC maxim of two failures, six walking wounded, two home runs) and therefore not matching money to circumstance; the tell is a 'let a thousand flowers bloom' spray-and-pray posture.
• Bringing in impatient-for-profit money too late, after you already have a working model, starving a scalable business of growth capital; the tell is a proven formula left underfunded.
• Reading only the investor's reputation or endorsements rather than their actual first-24-month demands; the tell is picking a 'camp' based on the dust jacket, not the circumstance.

• ▸Venture stage — mark: profit model PROVEN or UNPROVEN.
• ▸Evidence of proven model — one line of repeatable-profit evidence, or 'none yet'.
• ▸Needed profile — patient-for-growth/impatient-for-profit OR impatient-for-growth/patient-for-profit.
• ▸Investor's first 24-month demand — scale or profitability? (one line).
• ▸Match verdict — FIT or MISMATCH (kiss of death).
• ▸Amount appropriate to stage — small-to-find-model vs. large-to-scale.
• ▸Next-stage switch trigger — the specific proof that lets you take growth capital.
• ▸Rejection note — why this money is declined if mismatched.

1. 1Separate your intended strategy (the deliberate, top-down plan you announced) from your emergent strategy (the actual pattern that forms from the day-to-day resource-allocation decisions you make as opportunities appear). Name both explicitly before you commit capital.
2. 2Diagnose which process fits your situation: use a deliberate, analytical process when the winning strategy is knowable in advance; use an emergent, discovery-driven process when customer needs, the right market, and the business model are still unknown.
3. 3For emergent conditions, structure investment so you learn cheaply. Fund the first pass to test assumptions, not to scale, and set the expectation that the business model will change as new-market or low-end signals arrive.
4. 4Watch where resources actually go, because resource allocation is where strategy is really made. In retailing this is floor and shelf space; in your firm it is engineering hours and capital. The formula (gross margin times inventory turns) decides what gets funded, regardless of the stated plan.
5. 5When early results reveal the real market, deliberately shift from emergent to deliberate. Once QuickBooks-style signals confirm who the customer is and how you make money, lock the strategy and execute against it.
6. 6Insulate the new-growth effort from the core's economics by giving it an autonomous unit with its own cost structure and profit formula, so the parent's higher-margin comparisons don't strangle it.
7. 7Re-examine your process choice against circumstance, not reputation. Decide whether focusing on the core or Fosteresque creative destruction fits your specific conditions rather than picking the camp with the better dust-jacket endorsements.

Hewlett-Packard faced the choice between defending its established laser jet business through a deliberate plan and pursuing an uncertain new-market opportunity in ink-jet printers. Ink-jets started slow and produced fuzzy images, and it was not obvious who would buy them, so a purely analytical, top-down forecast would have killed them against laser jet margins. Instead HP set up an autonomous ink-jet business unit with its own cost structure to compete against its own laser jet business, letting the new unit discover its market — primarily students who could now individually own a compact, light, low-cost printer. That emergent, new-market path let the product improve until ink-jets became the mainstream choice and pushed laser jets to the high end. Contrast this with the full-service department stores facing discounters: their business model made fleeing the rational deliberate choice (cosmetics at 50% margins and three turns yielded 150% ROCII versus 60% by defending hard goods). HP survived because it chose an emergent process protected from the core's economics; the department stores' fixed process guaranteed retreat.

• If the winning strategy is knowable and the market, customer, and business model are clear, then run a deliberate top-down process; if any of those three is unknown, then run an emergent, discovery-driven process.
• If you are entering nonconsumption (new-market) territory where customers previously lacked money or skills, then do not fund for scale — fund to learn, because forecasts will be wrong.
• If early data reveals the real customer and revenue formula (as QuickBooks did within two years), then switch from emergent to deliberate and execute hard.
• If the new business must be judged against the core's margins and turns to get funded, then it will be starved — house it in an autonomous unit with its own economics.
• If two credible prescriptions conflict (focus on core vs. creative destruction), then decide by matching to your circumstance, not by the authors' reputations.

• Treating the announced plan as the real strategy while resource allocation quietly follows a different emergent path. Tell: the capital and engineering hours flow somewhere other than the slides say.
• Applying a deliberate analytical process to a genuinely unknown market and killing the idea on early performance. Tell: you reject a slow, fuzzy first product (like early ink-jets) because it loses to core metrics.
• Letting the parent's profit formula veto the new business. Tell: the growth unit is evaluated on gross margin and inventory turns identical to the core and always looks unattractive.
• Personal-strategy drift: emergent decisions default to short-term wins with no guiding purpose. Tell: day-to-day choices accumulate into a direction nobody intended and no one chose.
• Picking a strategy camp by author reputation and endorsements rather than fit. Tell: your justification is 'these consultants say so' instead of 'these are our circumstances.'

• ▸Intended strategy: the plan you have announced, in one sentence
• ▸Emergent pattern: where resources actually flow today, in one sentence
• ▸Knowability check: is the winning strategy knowable in advance? (yes = deliberate / no = emergent)
• ▸Unknowns list: customer, market, and business model — mark each known or unknown
• ▸Resource-allocation reality: what gets funded (shelf space / engineering hours / capital) and by what formula
• ▸Learning milestone: the specific signal that will confirm the real customer and revenue model
• ▸Switch trigger: the result at which you move from emergent to deliberate execution
• ▸Autonomy check: does the new business have its own cost structure separate from the core? (yes/no)
• ▸Circumstance fit: focus-on-core or creative-destruction — and the condition that makes it correct here