What happens when you can design the organization around AI from day one instead of retrofitting legacy. The structural advantage of building without inherited debt—and the risks of over-investing in technology before organizational capacity is established.
Between 2020 and 2025, fewer than 95 new banks opened in the United States—down from more than 1,200 in the preceding decade. Yet these de novo institutions, alongside digital subsidiaries, spin-offs, and new corporate entities, represent the clearest test case for a question that dominates enterprise technology strategy: can you build a better organization by starting from zero?
The evidence from 15 new-entity technology builds across banking, insurance, financial services, and corporate venturing reveals a paradox. Greenfield organizations enjoy operating costs up to 70% lower than incumbents, break even in three to four years, and can generate steady-state returns on equity exceeding 15% (McKinsey, 2018). Yet approximately 80% of neobanks remain unprofitable, 50% of corporate spin-offs valued above $1 billion fail to create shareholder value within two years (HBR, 2022), and multiple high-profile digital subsidiaries—including JPMorgan Chase's Finn and RBS's Bó—have been shut down and reabsorbed.
The greenfield advantage is real, but it is organizational, not technological. New entities that sequence capability building before technology acquisition outperform those that replicate the incumbent pattern of over-investing in infrastructure before establishing the organizational capacity to use it. This brief extends the "Greenfield Designer" archetype introduced in Enterprise AI Investment 2026 Outlook (RB-Q2), which identified this pattern as one of five investment archetypes shaping enterprise AI returns.
The formation of entirely new financial and technology entities in the United States effectively collapsed after 2008. The FDIC approved an average of 153 new bank charters per year from 1990 through 2008. From 2009 through 2024, the total was 95—over sixteen years. The drought was not limited to banking. ILC (industrial loan company) charter approvals, a route favored by fintech and corporate entrants, saw a complete moratorium from 2006 through 2013, followed by a de facto freeze that produced only three approvals in the next decade: Nelnet Bank (2020), Square Financial Services (2021), and Thrivent Bank (2024).
The landscape is now shifting. In 2025, the OCC received 14 de novo charter applications for limited purpose national trust banks—nearly matching the total from the prior four years combined. Many of these applications involve fintech and digital-asset firms seeking to move core activities inside a regulated banking perimeter. Acting FDIC Chairman Travis Hill stated that the agency intends to "find ways to support more de novo bank formation, including those that involve novel or innovative business models." In July 2025, the FDIC conditionally approved Ford Motor Company and General Motors for ILC deposit insurance—signaling renewed openness to commercial parent entities operating banks.
Simultaneously, incumbent organizations have launched digital subsidiaries and spin-offs at an accelerating pace. Goldman Sachs launched Marcus in 2016, reaching $100 billion in deposits by 2020. JPMorgan Chase launched a digital retail bank in the United Kingdom, "designed from scratch to specifically meet the needs of customers." DBS launched Digibank. Santander launched Openbank. MassMutual launched Haven Life as a fully digital insurance subsidiary. The pattern extends beyond financial services: BCG Digital Ventures has launched nearly 200 new businesses since 2014 in partnership with Fortune 500 organizations.
McKinsey's analysis of greenfield digital banks establishes the structural economic case. Well-executed greenfield digital banks can achieve operating costs up to 70% lower than traditional banks while generating steady-state returns on equity exceeding 15%. Without the burden of large fixed physical infrastructure, these organizations can pass efficiencies to customers in the form of better pricing—attracting deposits and loan volumes that incumbents lose gradually.
The cost advantage extends beyond banking. McKinsey's research on corporate venture building shows that the weighted average investment required before a new venture breaks even has decreased from $125 million to $77 million, with digital ventures breaking even faster than physical-product ventures. BCG reports a 66% success rate in its corporate venturing engagements, compared to an industry average closer to 47% for ventures that meet or exceed growth expectations.
The technology investment required has also decreased. Core banking platform investments for greenfield institutions now range from $10 million to $40 million, compared to $100 million or more for incumbents attempting progressive modernization of legacy core systems. Cloud-native platforms—Thought Machine, Mambu, Temenos Infinity—have reduced the time from charter approval to operational launch, enabling new entities to deploy modern technology stacks in months rather than years.
| Dimension | Greenfield Entity | Incumbent Retrofit | Source |
|---|---|---|---|
| Operating cost ratio | Up to 70% lower | Baseline | McKinsey, 2018 |
| Core banking platform cost | $10M–$40M | $100M+ (legacy modernization) | McKinsey, 2025 |
| Break-even timeline | 3–4 years | Ongoing (no baseline reset) | McKinsey, 2018 |
| Venture break-even investment | $77M (weighted avg) | $125M (prior year avg) | McKinsey, 2024 |
| Steady-state ROE potential | 15%+ | Varies (legacy drag) | McKinsey, 2018 |
| Profitability rate (neobanks) | ~20% profitable | N/A | Fortune BI, 2024 |
Eighty percent of neobanks remain unprofitable despite structural cost advantages (Fortune Business Insights, 2024). The cost advantage of building from zero is real but insufficient. The organizations that convert the advantage into returns share a common characteristic: they invested in organizational capability before technology scale.
The greenfield advantage should, in theory, produce consistently superior outcomes. The absence of legacy technology debt, entrenched processes, and organizational inertia means new entities can design from first principles. Yet the evidence shows that most greenfield organizations default to the same structural mistakes as their incumbent counterparts.
Deloitte's Tech Trends 2026 report identifies this directly: "Winners are not layering onto broken processes; they are rebuilding operations from the ground up with focused, measurable results." Yet only 1% of IT leaders report that their organizations have no significant operating model changes underway—indicating that even new entities are immediately confronting the gap between technology ambition and organizational readiness.
The pattern repeats across sectors. JPMorgan Chase launched Finn as a mobile-first digital bank in 2018, targeting digitally native customers. It was shut down within a year and customers were migrated to existing Chase accounts. RBS launched Bó as a standalone digital bank in the UK; it was closed in 2020 after attracting only 11,000 customers. Goldman Sachs's Marcus, while reaching $100 billion in deposits, accumulated billions in consumer lending losses as the organization struggled to build the operational and risk management capabilities required for a consumer business.
Gartner predicts that 30% of generative AI projects will be abandoned after proof of concept by the end of 2025, due to poor data quality, inadequate risk controls, escalating costs, or unclear business value. The prediction applies with equal force to new entities, where the absence of legacy infrastructure creates an illusion of readiness that masks the absence of organizational capability.
This is the paradox: the greenfield advantage exists, but most new entities squander it by over-investing in technology before establishing the organizational capacity to absorb it. They replicate the 93/7 spending ratio—93% technology, 7% people and process—that Enterprise AI Investment 2026 Outlook identified as the dominant failure pattern across all organizations.
If starting from zero confers a structural advantage, why do most new entities still default to the legacy investment pattern—and what do the exceptions do differently?
No single research program identifies what distinguishes the greenfield entities that generate returns from those that do not. But when the evidence from 15 documented new-entity technology organizations is placed side by side—de novo banks, ILC charter holders, digital subsidiaries, corporate spin-offs, and AI-native ventures—a convergence pattern emerges that none of the individual studies articulate.
The FDIC data on de novo bank formation shows that 95 banks were chartered between 2009 and 2024, compared to 1,243 in the prior nine years. The survivors share a pattern: focused business models with constrained product scope. Craft Bank (Atlanta, 2020) broke even before the end of its second year by concentrating on commercial lending in a single market. Thrivent Bank (2024)—only the third ILC charter approved since 2008—limited its scope to serving an existing membership base of 2.3 million rather than competing broadly. Constraint, not ambition, predicted survival.
Among 15 documented digital subsidiary launches by incumbent institutions since 2016, the pattern bifurcates sharply. Successes—Marcus (Goldman Sachs), Chase UK, Digibank (DBS), Openbank (Santander)—invested in organizational design, talent architecture, and operational processes before scaling technology. Failures—Finn (JPMorgan Chase), Bó (RBS), and multiple unnamed ventures—launched on modern technology stacks without adequate organizational infrastructure. McKinsey's analysis confirms: the technology itself was not the differentiator. What failed was the organizational integration.
McKinsey's 2024 survey of corporate venture builders reveals that expert builders—those that have built six or more ventures in five years—achieve double the success rate of novice builders and generate 12 times more revenue by a venture's fifth year, despite investing only twice the capital. BCG reports a 66% success rate in its corporate venturing model. The differentiator is organizational capability—the accumulated knowledge of how to sequence people, process, and governance decisions before the technology is deployed.
McKinsey's analysis of 400+ public companies shows that organizations with advanced product operating models have 60% greater total returns to shareholders and 16% higher operating margins. Among new entities, those that adopted product-centric operating models from inception—organizing around outcomes rather than projects—outperformed those that adopted traditional IT delivery structures. The operating model decision made at founding persists. Organizations rarely reorganize from project-based to product-based after initial scaling.
The convergence insight: The greenfield advantage resides in the opportunity to sequence organizational capability before technology investment—the absence of legacy technology is secondary. The 15 new-entity builds examined in this brief reveal a consistent pattern: entities that invested first in talent architecture, governance design, and operating model definition—before selecting and scaling technology—achieved profitability at 2–3 times the rate of those that led with technology. The advantage lies in building in the right order, regardless of starting position.
Across the 15 new-entity builds studied, five distinct archetypes emerge—each reflecting a different structural relationship between organizational capability investment and technology deployment.
The following diagnostic helps leadership teams assess whether their new-entity initiative is structured for capability-first success. It is organized around the Four Capability Bands from The Intelligence Organization, applied specifically to the greenfield context where legacy constraints are absent but organizational capability must still be built deliberately.
| Capability Band | Assessment Question | Score 1–5 | Greenfield Implication |
|---|---|---|---|
| Band 1: Right-Fit Technology | Is the technology scope constrained to 3–4 core capabilities, or does the roadmap attempt to replicate an incumbent’s full stack? | ___ | Score <3: narrow the technology scope before proceeding. The greenfield advantage is lost when the entity tries to build everything. |
| Has the organization assessed whether it has the people, processes, and data infrastructure to actually operate the technology it plans to build—or is the build plan based on what the technology can do? | ___ | Score <3: the technology plan exceeds the organization’s absorption capacity. Scope each platform choice against what the team can sustain, not what the vendor can deliver. | |
| Band 2: People & Purpose | Has the talent architecture been defined—roles, capabilities, and organizational structure—before the technology stack was selected? | ___ | Score <3: pause technology procurement and complete organizational design first. This is the binding constraint for 80% of unprofitable greenfield entities. |
| Have you identified who in the organization is already experimenting and willing to lead adoption, versus who will need demonstrated proof before engaging—and are those groups being activated differently? | ___ | Score <3: a single adoption strategy applied uniformly will fail. Early adopters need freedom to experiment; skeptics need evidence of results, not training. Design for both. | |
| Band 3: Operational Integration | Do internal data assets have assigned owners, defined quality standards, and service commitments—or is data treated as a shared resource with no individual accountability? | ___ | Score <3: assign data ownership before the first product launches. Greenfield entities can build this discipline from day one—incumbents spend years retrofitting it. |
| Are cross-functional terms defined consistently from day one—does “customer,” “risk,” or “product” mean the same thing to every team—or is each function developing its own definitions? | ___ | Score <3: unify definitions now. This is the single largest greenfield advantage over incumbents, where reconciling legacy definitions across functions takes years and costs millions. | |
| Band 4: Adaptive Governance | Are low-risk activities—internal experimentation, non-customer-facing tools, standard operational decisions—pre-authorized to proceed without individual approval? | ___ | Score <3: separate governance speed by risk level. Low-risk work should move in days. High-risk decisions (customer-facing, regulatory) get deliberate review. One speed for everything is the pattern that slows incumbents. |
| Is decision authority assigned to whoever has the most relevant expertise for each domain, or does it default to seniority and hierarchy? | ___ | Score <3: greenfield entities have a narrow window to establish expertise-based decision rights before hierarchy calcifies. Once seniority governs decisions, it is structurally difficult to reverse. |
Total 32–40: Capability-first entity. Organizational readiness precedes or matches technology investment. Focus on sustaining the sequence as the entity scales—the gravitational pull toward technology spending increases with growth.
Total 20–31: Mixed sequence. Some organizational capability is established, but gaps exist in one or more Bands. Identify the lowest-scoring Band and address it before scaling the technology investment. The most common gap is Band 2 (People & Purpose)—the talent architecture was deferred in favor of platform selection.
Total 8–19: Technology-first entity. The greenfield advantage is being consumed by the same investment pattern that constrains incumbents. Halt technology scaling and invest the next 90 days in organizational design across all four Bands. The evidence shows this pause generates better long-term returns than continued technology investment without organizational infrastructure.
Focused Charter (Archetype 01): Bands 2 and 4 are the priorities. The charter constrains scope (Band 1), but talent and governance must be established before the entity accepts its first customer. Craft Bank's sub-two-year breakeven demonstrates what is possible when these Bands are addressed at founding.
Incubated Subsidiary (Archetype 02): Band 3 is the critical test. If the subsidiary shares the parent's project-based operating model, it is not truly independent—regardless of whether the technology stack is separate. Genuine operational independence requires a distinct operating model, not just a distinct brand.
Corporate Venture Build (Archetype 03): All four Bands apply, but Band 4 (Adaptive Governance) determines whether the parent organization will allow the venture to operate with the speed required for market entry. The 66% success rate in BCG's model reflects ventures where governance was designed to enable, not constrain.
Regulated Entrant (Archetype 04): The regulatory perimeter imposes Bands 1 and 4 externally. The diagnostic should focus on Bands 2 and 3, where the entity has discretion. The ILC capital requirements ($667M for GM, $1.5B for Ford) force patient capitalization that, combined with capability-first sequencing, may produce the strongest long-term foundations.
AI-Native Rebuild (Archetype 05): Band 2 is the binding constraint. Deloitte's emerging roles—AI collaboration designers, edge AI engineers—represent capabilities that do not yet exist at scale. The entity must either build these capabilities internally or design organizational structures that make them acquirable. Technology is the least constrained Band for this archetype.
Decision support aligned with The Intelligence Organization · Four Capability Bands (Starkey, 2026) · Greenfield sequencing as primary determinant of new-entity outcomesThis research is the foundation for our new-entity capability architecture executive workshop series. If your organization is evaluating a de novo charter, digital subsidiary, or AI-native entity build, we should talk.
Schedule a ConversationIndustry Research: McKinsey & Company, "Ten Lessons for Launching a Greenfield Digital Banking Business," McKinsey Financial Services, Oct 2018. McKinsey & Company, "How CEOs Are Turning Corporate Venture Building Into Outsize Growth," Oct 2024, survey of 757 senior executives. McKinsey & Company, "The Way to Win in Corporate Venturing: Serial Building and AI," 2024. McKinsey & Company, "The Bottom-Line Benefit of the Product Operating Model," Dec 2023, analysis of 400+ public companies. BCG, "A Proven Model for Corporate Venturing," 2022. BCG, "Accelerating Business Building for Companies," Nov 2023, survey of 1,000+ senior leaders across 17 countries. Deloitte, "The Great Rebuild: Architecting an AI-Native Tech Organization," Tech Trends 2026, Dec 2025. Deloitte, "AI Comes of Age: 17th Annual Tech Trends Report," Dec 2025.
Regulatory & Institutional: FDIC, Bank Applications Data and Deposit Insurance Decisions, 2009–2024. FDIC, "Deposit Insurance Application Approved for Erebor Bank, N.A.," 2025. FDIC, Acting Chairman Travis Hill, "Update on Key Policy Issues," 2025. FDIC, Conditional Approvals for Ford Credit Bank and GM Financial Bank, 2025. Congressional Research Service, "De Novo Banks: Policy Issues for the 118th Congress," IF12697, 2024. Congressional Research Service, "Industrial Loan Companies (ILCs): Background and Policy Issues," R46489, 2024. OCC, Comptroller Gould, "De Novo Charters," 2025. Gartner, "30% of Generative AI Projects Will Be Abandoned After Proof of Concept by End of 2025," Jul 2024.
Market Data & Case Studies: Fortune Business Insights, "Neobanking Market Size, Share, Growth," 2024, market valued at $143.29 billion. Goldman Sachs, "Marcus by Goldman Sachs: Consumer Finance Platform," company disclosures, 2016–2020. JPMorgan Chase, "Digital Consumer Banking in the U.K.," company disclosures, 2021. Banking Dive, "Thrivent Receives ILC Charter," Jun 2024. American Banker, "Will the Slow Pace of New Bank Formation Pick Up in 2026?," Dec 2025. Goodwin Law, "FDIC Approves ILC With Traditional Bank Business Model," Jul 2024.
RBD. Research: Starkey, M.C., The Intelligence Organization, 2026. RBD., "Enterprise AI Investment 2026 Outlook: From Technology-First Budgets to Capability-First Returns," RB-Q2 2026. RBD. cross-industry new-entity build analysis and greenfield sequencing synthesis, 2026.