Business Model Innovation Fails: 6 Systemic Causes and How to Spot Them

An automotive corporation invests 14 billion euros in a new software unit. Four years and 8 billion euros in losses later, it pivots to partnerships with Chinese and American startups. An insurer launches a car-sharing joint venture and sells it three years later because its own distribution network sabotaged the new model. A machinery manufacturer needed 15 years until the business model innovation worked — not because of the technology, but because the field sales force couldn’t sell the new model.

If you’re responsible for business model innovation at a DACH company, you probably recognize at least one of these patterns. The question that’s rarely asked: Why do the same failure patterns repeat — across different industries, with different budgets, among competent leadership teams?

The answer: Business model innovation (BMI) fails for different reasons than product innovation or business design projects. Product innovation fails due to technology or market demand. Business design projects fail due to methodology — treating the canvas as a result instead of a hypothesis, lacking validation, using wrong metrics. Business model innovation fails at something more fundamental: the organization itself. The numbers confirm the pattern: 83 percent of companies worldwide name innovation as a top-3 priority, but only 3 percent have the organizational infrastructure to execute it.¹ In Germany, a leading position in knowledge generation (rank 1 in the BDI Innovation Indicator) is offset by a commercialization efficiency of just 61 percent.²

This article documents six systemic causes — each with academic grounding, a recognizable corporate scenario, and a diagnostic question to help you assess your own situation. The causes are not independent: Dominant logic (cause 1) creates the exploitation trap (cause 2), which feeds the immune system (cause 3), which intensifies the logic conflict (cause 5). Addressing only one cause treats a symptom.

Cause 1: Dominant Logic — The Prison of Your Own Success

C. K. Prahalad and Richard Bettis coined the term “dominant logic” in 1986: the mental models through which a leadership team understands and steers its company.³ This logic is not a conscious decision — it is a filter through which all information is processed. What doesn’t fit the existing business model is not recognized as an opportunity but filtered out as noise.

Clayton Christensen describes the mechanism in the RPV framework: A company’s resources can be reallocated. Its processes and values (the criteria by which priorities are set), however, are structurally rigid.⁴ A machinery manufacturer that has sold equipment for 30 years has processes optimized for unit-cost accounting and values that reward revenue per unit sold. In this system, a pay-per-use model is not an innovation — it is a threat to the calculation logic.

The DACH example: Volkswagen invested over 14 billion euros in CARIAD, its in-house software unit for the software-defined vehicle.⁵ Cumulative operating losses exceeded 8 billion euros over four years. Of 6,000 employees, only about 10 percent actually wrote code — the rest managed processes inherited from the hardware world.⁵ VW had the resources (budget, personnel), but the processes and values of the hardware corporation could not be converted into a software organization by executive decree. In 2024, VW pivoted to partnerships with Xpeng and Rivian — a de facto admission that internal BMI had failed.

Mary Tripsas and Giovanni Gavetti documented the same mechanism using the example of Polaroid: The company developed excellent digital camera technology but couldn’t change its business model because the entire organization was built around the razor-and-blades model (cheap device, expensive film).⁶ Cognitive inertia was stronger than technological competence.

What this feels like in practice: You’re sitting in a strategy meeting. A colleague presents a pay-per-use model for your industrial equipment. The first question from the room: “What does the unit cost calculation look like?” The second: “What happens to the sales commissions that disappear?” Neither question is wrong — but both presuppose the existing business model as the evaluation framework. The new model is measured against the logic of the old one and found lacking. Not because it’s bad, but because the evaluation criteria structurally disadvantage it.

Diagnostic question: Take your company’s last innovation evaluation. Open the scoring matrix or evaluation sheet. Do the criteria (payback period, contribution margin per unit, revenue in year 1) come from the existing business model? If so: Dominant logic is acting as a filter — and it systematically screens out everything that works differently.

Cause 2: The Exploitation Trap — Optimization as Addiction

James March described the fundamental problem of organizational learning in 1991: Exploitation (leveraging the known) systematically displaces exploration (investigating the unknown).⁷ The mechanism is self-reinforcing: Every successful optimization delivers measurable results and confirms the conviction that optimization is the right strategy. Exploration, on the other hand, produces uncertainty, failures, and no measurable returns in the short term.

Daniel Levinthal and March call this the “myopia of learning”: Organizations learn too slowly to recognize threats, too superstitiously (they draw wrong conclusions from small samples), and too narrow-mindedly (they optimize near their existing competencies instead of exploring new territory).⁸

Why this hits especially hard in DACH: The BDI Innovation Indicator 2025 documents the German paradox: Knowledge generation at the reference value of 100 (best among all countries compared), commercialization efficiency at just 61.² Germany is world-class at optimizing existing knowledge and translating it into incremental improvements — circular economy (rank 1), new materials (rank 2), production technologies (rank 5). But in translating knowledge into new business models, the country falls behind: digital connectivity (rank 10), biotechnology (rank 15).

The Bitkom study 2025 confirms the picture at the company level: 53 percent of German companies have problems with digitalization. Only 3 percent find developing digital business models “easy.”⁹ 31 percent develop no digital business models at all. This is the exploitation trap in its purest form: Companies know they need to change their business model, but their entire system is geared toward optimizing the existing one.

In budget negotiations, exploitation projects can present concrete profitability calculations. Exploration projects have vague future prospects. Promotion systems reward efficiency gains. Quarterly optimization successes are easier to communicate than long-term investments in uncertain future markets.⁸

Diagnostic question: Open your innovation portfolio or this year’s budget plan. Categorize each project into one of two buckets: “improves the existing business model” or “tests a new business model.” Calculate the ratio. If it’s above 90:10, you’re in the exploitation trap. Second test: How many of your innovation projects over the last three years had validated customer contact before budget approval? If the answer is “none” or “we don’t know,” the exploitation trap is structurally embedded.

Cause 3: The Organizational Immune System

Henry Chesbrough identified the central barriers to business model innovation in established companies in 2010: Conflicts with existing assets and business models, the need to “cannibalize” the established business, and the cognitive inability to recognize the value of new business models.¹⁰

Donald Sull describes the mechanism as “active inertia”: Organizations respond to disruptive change not with passivity but by intensifying the very routines that caused the problem.¹¹ A new business model is a foreign body in the organism of the company. The existing structures — budget processes, approval cascades, career incentives, KPIs — are the antibodies that identify and neutralize this foreign body. This doesn’t happen out of malice. It happens because the system was built precisely for this purpose: to protect stability.

What this feels like in practice: You’ve developed a new subscription model and now need a pilot contract. Legal says: “Our terms and conditions don’t cover this, and the revision will take six months.” Procurement says: “This vendor isn’t on our approved list.” Finance says: “The monthly revenue recognition doesn’t fit our reporting.” None of these objections is unreasonable. But collectively, they neutralize the new model before it reaches a single customer. Every department is doing its job — and that is precisely the problem.

The DACH example: BMW and Daimler founded ShareNow in 2019 as a merger of DriveNow and car2go — the European market leader in car-sharing. In 2022, they sold the company to Stellantis’ Free2Move.¹² The Free2Move CEO commented: “They didn’t find the recipe. We already found it.” The core of the problem wasn’t operational. It was structural: BMW and Daimler sell cars. Car-sharing reduces the need to buy cars. The immune system of the parent corporations — sales incentives, dealer networks, volume targets — worked against the new business model, even though the boards strategically supported it.

Steve Blank distinguishes three forms of innovation theater that the immune system produces:¹³ Organizational theater (restructurings that keep everyone busy for a year without changing anything), innovation theater (hackathons and design thinking workshops that rarely produce deployable products), and process theater (attempts to reform procurement or legal without an overarching innovation doctrine). The immune system produces activity as a substitute for results.

Diagnostic question: Is there a BMI initiative in your company that has existed for more than 12 months without a single paying customer using the new business model? If so, the immune system may be working faster than the innovation.

Cause 4: Business Model Innovation as a One-Off Project

You recognize this pattern from a typical sentence in a status report: “The innovation initiative has been successfully completed.” A business model that still needs to be validated cannot be “completed.” It can be tested, adapted, scaled, or abandoned — but “completing” presupposes a plannable outcome, which BMI by definition does not have. Rita McGrath argues in The End of Competitive Advantage that sustainable competitive advantage is a myth in most industries.¹⁴ Advantages are transient — they emerge, are exploited, and erode. In this world, BMI is not a one-time transformation but a permanent capability.

Oliver Gassmann demonstrates with the St. Gallen Business Model Navigator that 90 percent of all new business models are recombinations of 55 identified patterns — not radical reinventions.¹⁵ But this recombination is not copy-and-paste: It requires iterative testing, adapting, and discarding. Over two-thirds of new products and solutions miss customer expectations. Business model innovation is an iterative recombination process with a high failure rate — not a linear project with a plannable outcome.

Alexander Osterwalder quantifies the effort for large enterprises: Approximately 250 tested projects over three years are needed to find one outlier success. 70 percent of projects should be shut down within three months.¹⁶ Bosch, for example, filtered 214 ideas through a discovery phase; 19 made it to execution — a graduation rate of 8.9 percent.¹⁶

The time factor: Hilti needed over 15 years to fully realize the shift from tool sales to fleet management.¹⁷ The biggest obstacle was not technology but sales: Retraining field sales representatives who had spent decades demonstrating tools to site managers to conduct conceptual sales conversations at C-level took years, not months. The cash-flow effect compounded the problem: “It’s easier to sell a tool and receive your money after 30 days than to deliver tools and receive monthly payments over four years.”¹⁷ In return, customer retention in fleet management was five times higher than in the sales model — and the model helped Hilti survive the 2008 financial crisis.

Most DACH companies expect BMI results within 18 to 24 months. Hilti shows: Even a successful servitization BMI requires over a decade of continuous commitment.

Diagnostic question: Does your BMI initiative have a planned “end date”? If so, it will most likely be terminated before reaching its goal — because business model innovation is not a project but a capability.

Cause 5: The Logic Conflict — Causation vs. Effectuation

Brenk et al. documented in a 2019 longitudinal study of a German machinery manufacturer what may be the most precise mechanism for BMI failure in the DACH region.¹⁸ The company — a classic mid-market firm with a strong engineering culture — attempted to shift from a product business model to a smart service model. The study documents in detail how the process failed: Every decision regarding the new business model had to pass through the same governance bodies that managed the core business. The governance was optimized for causation: predicting the future, minimizing contingency, calculating risks. Business model innovation, however, requires effectuation: acting under uncertainty, generating new information through experiments, learning from failures rather than avoiding them.

What makes the Brenk study particularly revealing: The actors within the company recognized the conflict — yet couldn’t resolve it. The engineers wanted the smart service model. The sales leadership saw the necessity. But the decision-making processes demanded the same evidence (ROI calculation, five-year plan, risk analysis) for every investment in the new model as for core business investments. For a model that by definition cannot yet deliver reliable figures, this is a structural death sentence.

The crux: The decision-making logic that successfully governs an existing business model is structurally incompatible with the decision-making logic that a new business model needs. Brenk et al. conclude: Alternative business models must be organizationally separated when their value creation logic contradicts the existing one.¹⁸

Vijay Govindarajan describes the same conflict with his Three-Box Solution: Box 1 (manage the existing business), Box 2 (forget what made the company successful), and Box 3 (create new models for the future).¹⁹ The central challenge is Box 2 — active unlearning. The mindsets and methods that make Box 1 excellent are fundamentally incompatible with Box 2 and Box 3.

Why this is especially relevant in DACH: The KfW/ZEW study shows that only 7 percent of mid-market companies pursue comprehensive business model innovation.²⁰ At the same time, 46 percent say digitalization has changed their business model.⁹ The gap between passive change and active shaping arises precisely at this logic conflict: The planning systems of the Mittelstand are built for a world of predictable outcomes. BMI is by definition unpredictable.

Two DACH-specific structural features intensify the logic conflict: First, the engineering culture that embeds technical perfection and controllability as core values. Effectuation — acting on incomplete information, learning from mistakes — fundamentally contradicts this cultural DNA. Second, codetermination (Mitbestimmung): Supervisory boards and works councils are structurally oriented toward protecting the status quo. Personnel changes necessary for a BMI unit require social compatibility and interest balancing — legitimate concerns that nevertheless slow down the pace of business model innovation. In US corporations, a CEO can build a separate unit with its own staff in weeks; in DACH corporations, the same process takes months.

Diagnostic question: Does your company require the same business case for BMI initiatives as for core business investments? If so, the causation logic is active — and it will block every initiative that cannot deliver a predictable ROI.

Cause 6: Wrong Separation — or Missing Separation

Charles O’Reilly and Michael Tushman describe organizational ambidexterity as the ability to conduct exploitation and exploration simultaneously.²¹ Nancy Bocken and Thijs Geradts identify the specific barriers in their empirical study: Incentives geared toward short-term efficiency, institutional logics dominated by shareholder value, and resource allocation driven by political influence rather than innovation potential.²²

The practical question is not whether to separate — but how. Two DACH examples illustrate the difference:

Klöckner/XOM Materials — correct separation: Klöckner launched its B2B trading platform XOM Materials as an independent unit, deliberately separated from the core business.²³ The rationale: “The platform and the core business had potentially conflicting business goals and KPIs.” The culture change — convincing employees and customers — was harder than the technology. But the structural separation allowed different metrics, different decision-making processes, and different time horizons. Digital channels reached 32 percent of total revenue by Q4 2019.

VW/CARIAD — wrong separation: CARIAD was founded as a separate unit but had no independent budget control. The brands — Audi, Porsche, VW — retained decision-making power while CARIAD was expected to deliver.⁵ The result: A CARIAD insider described the situation as “We were just a hose.” Three brand factions built parallel systems instead of a unified platform. Some brands actively blocked decisions. Separation without autonomy is not separation — it is an alibi.

A study by DHBW Karlsruhe confirms: Placing innovation decisions outside the core business at top management level carries the highest probability of failure.²⁴ And: Innovations without strategic connection to the company generate strong resistance. Separation must therefore meet two conditions: Autonomy in execution and strategic connection to the overall company.

Diagnostic question: Does your BMI unit have its own budget, its own KPIs, and its own decision-making authority? Or does it report to the same governance bodies that manage the core business? If the latter: You have a unit without autonomy — and the immune system will neutralize it.

Cross-Section: What Connects These Six Causes

All six causes share a common denominator: The existing business model shapes the organization so that it reproduces that model — and expels alternative models.

Cause	Mechanism	Typical Symptom
Dominant logic	Mental models filter new business models as “noise"	"That doesn’t fit our core business”
Exploitation trap	Optimization successes confirm the existing strategy	Innovation budget >90% for core business improvement
Immune system	Structures neutralize foreign bodies (new models)	Innovation labs without deployable results
One-off project	BMI treated as a project with an end date	”The innovation initiative is completed”
Logic conflict	Causation logic blocks exploration	Business case required for every BMI initiative
Wrong separation	Autonomy without connection or connection without autonomy	Separated unit reporting on core business KPIs

Where to start? When all six causes are at work simultaneously, trying to address them all at once is itself a failure pattern. The evidence suggests a sequence: First, resolve the logic conflict (cause 5) through structural separation — without a protected unit with its own decision-making authority, there is no room for experimentation. Then address the dominant logic (cause 1) by having the separate unit develop its own evaluation criteria. The remaining causes (exploitation trap, immune system, one-off project, separation errors) can only be addressed once this foundation is in place. Klöckner/XOM Materials shows this path; VW/CARIAD shows what happens when separation is attempted without autonomy.

Industry variance: The weighting of the six causes varies. Machinery and automotive suffer disproportionately from dominant logic (unit-cost thinking, hardware processes). Insurance and financial services fail more often at the immune system (compliance cascades, regulatory approval processes). Technology companies are more likely to fall into the exploitation trap (optimizing existing platforms instead of new business models). The diagnostic questions at the end of each cause help determine your own industry weighting.

The organizational architecture dimension: All six causes share a common root: hierarchical decision-making structures optimized for exploitation. In traditional line organizations, every BMI decision must pass through governance bodies that judge by core business logic. An alternative approach — role-based governance with distributed authority — addresses the logic conflict structurally: Someone authorized in a BMI role to make decisions using effectuation logic does not need approval from a governance body that judges by causation logic. This is not theoretical speculation: Bosch deliberately uses different governance structures for its “Grow” units than for the core business.¹⁶ Klöckner gave XOM Materials independent decision-making authority.²³ The separation described in cause 6 is fundamentally a question of organizational architecture — not org chart cosmetics.

The seven structural innovation killers describe organizational barriers from the perspective of hierarchy research. The seven business design mistakes address methodological errors within a BMI project. This article adds the third level: the systemic causes that explain why the project never gets the chance to correct its methodological errors.

A Counterpoint: When BMI Is the Wrong Strategy

Not every company needs business model innovation. Rita McGrath emphasizes: If your existing business model still generates healthy returns and no disruptive competitor is visible, aggressive BMI can destroy more value than it creates — through organizational disruption, talent attrition, and distraction from the core business.¹⁴

March warns of the opposite side of the exploitation trap: the failure trap. Organizations that obsessively explore without ever scaling a validated model burn resources and talent without building a sustainable business.⁷ The right answer is almost always a portfolio approach — but the balance depends on industry lifecycle and competitive position.

The decisive strategic question is not “Are we innovating enough?” but “Are our exploration and exploitation investments in the right ratio for our competitive situation?”

Frequently Asked Questions

Why do business model innovations fail more often than product innovations?

Business model innovation changes the value creation architecture — not a single product. This activates all six systemic causes simultaneously: Dominant logic filters the new model, the exploitation trap starves it of budget, the immune system mobilizes resistance, the project logic sets an end date, the logic conflict blocks decisions, and wrong separation prevents autonomy. Product innovation works within the existing system. BMI works against it.

What percentage of business model innovations fail?

Chesbrough (2010) estimates that over 90 percent of BMI attempts do not achieve their intended outcomes.¹⁰ Osterwalder quantifies it from the portfolio perspective: Of 250 tested projects, typically one produces an outlier success.¹⁶ The critical point: Individual project failure is expected and healthy. Portfolio failure — no outlier despite sufficient experiments — indicates a systemic problem.

How does this article differ from the business design mistakes?

The business design mistakes cover methodological errors within a BMI project: treating the canvas as a result, lacking validation, using wrong metrics, premature scaling. This article addresses why the project never gets the conditions to correct its methodology errors in the first place — the systemic causes at the organizational level.

What is the exploitation trap?

The exploitation trap describes the self-reinforcing mechanism by which optimization successes confirm the conviction that optimization is the right strategy — systematically displacing exploration.⁷⁸ In DACH companies, this is especially pronounced: Engineering culture, process competence, and quality management are exploitation strengths that simultaneously undermine the ability to innovate business models.

Which German companies have failed due to a lack of business model innovation?

VW/CARIAD (over 8 billion euros in cumulative losses attempting to build a software organization), BMW/Daimler ShareNow (sold to Stellantis in 2022 because the car-sharing model collided with car sales), and VW/MOIA (Hannover location closed in 2025, pivot to technology provider). Hilti is the positive counterexample — but the 15-year transformation duration illustrates that even success is not a sprint.

What can companies concretely do?

Four operational starting points, in this order:

1. Conduct the diagnosis: Answer the six diagnostic questions at the end of each cause honestly. Three or more “yes” answers indicate a systemic problem. Prioritize the causes with the clearest manifestation.

2. Create a separated unit: Give the BMI initiative its own budget (not as a line item in the core business budget but as an independent cost center), its own KPIs (learning velocity, hypotheses validated per quarter, customer contacts — not revenue in year 1), and its own decision-making authority (who decides on pivots and kill decisions without a board submission). The Klöckner model, not the CARIAD model.

3. Make the time horizon explicit: Communicate to the board and supervisory board: BMI portfolios typically deliver outlier results after 3-5 years, not after 18 months. Hilti’s fleet management took 15 years. Set milestones for learning progress, not for revenue.

4. Protect the exploration budget: Define a fixed percentage (Osterwalder recommends at least 10 percent of the innovation budget for explore projects) and protect it from quarterly reallocation. If the exploration budget is up for debate in the next budget round, the exploitation trap has won.

Methodology & Sources

This article is based on 24 academic and practitioner sources on business model innovation, including four empirical DACH studies (KfW/ZEW 2024, Bitkom 2025, BDI 2025, DHBW Karlsruhe 2023), five foundational works (Prahalad/Bettis, March, Chesbrough, Christensen, O’Reilly/Tushman), a longitudinal study on BMI failure in German machinery manufacturing (Brenk et al. 2019), and three documented DACH company examples (VW/CARIAD, BMW/Daimler ShareNow, Klöckner/XOM Materials).

SERP finding: No German-language SERP result treats BMI-specific failure as a standalone topic with academic grounding and enterprise context. The existing results are either generic innovation failure lists (startup-focused), BMI process guides with a failure sidebar, or canvas usage errors at the tactical level.

Limitations: The DACH company examples illustrate typical patterns but are not complete case studies. The quantitative data on failure rates (Chesbrough >90%, Osterwalder 250:1) come from different contexts and methodologies. Survivorship bias in innovation research is a general problem: We study Hilti (success) more intensively than the dozens of mid-market companies that attempted servitization and quietly failed.

Disclosure: SI Labs supports companies in business model development. We have endeavored to base recommendations on published sources and to transparently name the limitations of the models.

References

Footnotes

1. BCG. “83% of Companies Rank Innovation as Top-Three Priority.” BCG Publications, June 2024. Based on a survey of 1,000+ C-level executives worldwide. ↩

2. BDI, Roland Berger, Fraunhofer ISI, and ZEW. Innovationsindikator 2025. Berlin: Bundesverband der Deutschen Industrie, 2025. Germany rank 12 of 35 countries. ↩ ↩²

3. Prahalad, C. K. and Richard A. Bettis. “The Dominant Logic: A New Linkage Between Diversity and Performance.” Strategic Management Journal 7, no. 6 (1986): 485—501. ↩

4. Christensen, Clayton M. The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail. Harvard Business School Press, 1997. RPV framework: Chapter 8. ↩

5. Based on reports from Der Autopreneur (2025), All-Electronics (2025), and Heise Online (2025). EUR 14 billion investment, EUR 8+ billion cumulative operating losses 2021—2024. Xpeng and Rivian partnerships announced 2024. ↩ ↩² ↩³

6. Tripsas, Mary and Giovanni Gavetti. “Capabilities, Cognition, and Inertia: Evidence from Digital Imaging.” Strategic Management Journal 21, no. 10—11 (2000): 1147—1161. ↩

7. March, James G. “Exploration and Exploitation in Organizational Learning.” Organization Science 2, no. 1 (1991): 71—87. ↩ ↩² ↩³

8. Levinthal, Daniel A. and James G. March. “The Myopia of Learning.” Strategic Management Journal 14, no. S2 (1993): 95—112. Three forms of myopia: temporal, spatial, and regarding false inferences. ↩ ↩² ↩³

9. Bitkom. Digitalisierung der Wirtschaft 2025. Berlin: Bitkom Research, 2025. Sample: n=603 companies with 20+ employees. ↩ ↩²

10. Chesbrough, Henry. “Business Model Innovation: Opportunities and Barriers.” Long Range Planning 43, no. 2—3 (2010): 354—363. ↩ ↩²

11. Sull, Donald N. “Why Good Companies Go Bad.” Harvard Business Review 77, no. 4 (July-August 1999): 42—52. Sull describes “active inertia” as the mechanism by which successful organizations respond to disruption by intensifying their existing strategies, processes, and values — and thereby fail. Additionally: Rumelt, Richard P. Good Strategy Bad Strategy: The Difference and Why It Matters. Crown Business, 2011. Ch. 3: “Bad Strategy” through organizational inertia. ↩

12. BMW/Daimler ShareNow: Sold to Stellantis/Free2Move in July 2022. Sources: BMW press release, Electrive, Wikipedia. ↩

13. Blank, Steve. “Why Companies Do ‘Innovation Theater’ Instead of Actual Innovation.” Harvard Business Review, October 2019. Three theater types: organizational, innovation, process. ↩

14. McGrath, Rita Gunther. The End of Competitive Advantage: How to Keep Your Strategy Moving as Fast as Your Business. Harvard Business Review Press, 2013. Transient advantages replace sustainable competitive advantages. ↩ ↩²

15. Gassmann, Oliver, Karolin Frankenberger, and Michaela Csik. Geschäftsmodelle entwickeln: 55 innovative Konzepte mit dem St. Galler Business Model Navigator. Carl Hanser Verlag, 2014. 90% recombination, 55 patterns. ↩

16. Osterwalder, Alexander, Yves Pigneur, Fred Etiemble, and Alan Smith. The Invincible Company: How to Constantly Reinvent Your Organization with Inspiration From the World’s Best Business Models. Wiley, 2020. 250-project heuristic, Bosch example. ↩ ↩² ↩³ ↩⁴

17. Hilti fleet management: Based on Harvard Business School Cases 717427 and 717465 and Strategyzer Library. Over 15 years transformation duration, 5x higher customer retention. ↩ ↩²

18. Brenk, Sebastian et al. “Learning from Failures in Business Model Innovation: Solving Decision-Making Logic Conflicts through Intrapreneurial Effectuation.” Journal of Business Economics 89 (2019): 1097—1124. Longitudinal study of a German machinery manufacturer. ↩ ↩²

19. Govindarajan, Vijay. The Three-Box Solution: A Strategy for Leading Innovation. Harvard Business Review Press, 2016. Box 1 (manage), Box 2 (forget), Box 3 (create). ↩

20. KfW Research, ZEW, and Fraunhofer ISI. Studie zu Geschäftsmodellinnovationen im Mittelstand. Bonn: KfW Bankengruppe, 2024. 7% comprehensive BMI, 14% core BMI, 28% broad BMI. ↩

21. O’Reilly, Charles A. and Michael L. Tushman. “Organizational Ambidexterity: Past, Present, and Future.” Academy of Management Perspectives 27, no. 4 (2013): 324—338. ↩

22. Bocken, Nancy M. P. and Thijs H. J. Geradts. “Barriers and Drivers to Sustainable Business Model Innovation: Organization Design and Dynamic Capabilities.” Long Range Planning 53, no. 4 (2020): Article 101950. ↩

23. Klöckner/XOM Materials: Based on Harvard Business School Case 820035, Etventure Case Study, and Applico analysis. 32% digital revenue share by Q4 2019. ↩ ↩²

24. DHBW Karlsruhe. Organisationale Erfolgsmuster für digitale Geschäftsmodellinnovation. Karlsruhe: Duale Hochschule Baden-Württemberg, 2023. ↩