SOIL Research Methodology White Paper

1. Introduction and Rationale

1.1 The Knowledge Gap

Organizational mortality represents one of the most consequential yet understudied phenomena in social science (Jenkins, 2016; Ucbasaran, 2013). While estimates suggest that the majority of new ventures fail and organizational death affects millions of individuals annually, our systematic understanding of why organizations die remains remarkably primitive.

Current knowledge suffers from several critical deficiencies:

Survivor Bias Dominance. The vast majority of organizational research examines successful organizations, extrapolating “success factors” from survivors. This approach fundamentally cannot distinguish between factors that contribute to success and factors that are merely common among both survivors and casualties.

Anecdotal Evidence Base. Existing failure knowledge consists primarily of post-hoc narratives, case studies selected for dramatic value, and self-serving founder accounts. These sources lack systematic structure, standardized measurement, and independent verification.

Temporal Compression. When failures are studied, analysis typically focuses on proximate causes — the final crisis that precipitated closure. The developmental trajectory leading to vulnerability remains largely unexplored.

Single-Perspective Limitation. Most failure accounts represent a single stakeholder view (typically the founder), missing the multi-dimensional reality of organizational dysfunction.

1.2 The Medical Analogy

Modern medicine developed through systematic autopsy — the careful examination of deceased bodies to understand disease processes. Before autopsy became standard practice, medicine relied on theory, speculation, and case reports. The shift to systematic post-mortem examination created the empirical foundation for pathology, which in turn enabled diagnosis, prognosis, and treatment.

Organizational science currently resembles pre-autopsy medicine. We possess abundant theory about organizational health but lack the systematic empirical foundation that would allow us to:

• Classify organizational pathologies taxonomically
• Identify early warning indicators with predictive validity
• Distinguish survivable crises from terminal conditions
• Develop evidence-based intervention protocols

SOIL proposes to address this gap by establishing organizational autopsy as a rigorous research practice.

1.3 Research Objectives

Primary Objective: Build a comprehensive, verified database of organizational mortality cases sufficient to enable pattern identification, hypothesis testing, predictive model development, and cross-disciplinary analysis.

Secondary Objectives:

• Develop and validate instruments for organizational autopsy data collection
• Establish reliability standards for retrospective organizational assessment
• Create taxonomies of organizational pathology
• Identify early warning indicators with predictive validity
• Generate testable hypotheses about organizational mortality mechanisms
• Provide therapeutic value to participants while maintaining research rigor

2. Theoretical Foundations

2.1 Methodological Separation: Collection vs. Analysis

SOIL's approach rests on a fundamental methodological distinction between data collection and data analysis.

Data Collection: Theory-Neutral, Comprehensive
Like a medical autopsy, organizational autopsy must collect comprehensive data without pre-committing to any particular diagnosis. A pathologist does not examine only the heart because they suspect cardiac failure — they examine the entire body systematically. Similarly, SOIL's interview protocols capture the complete organizational trajectory across all functions, allowing the data to reveal patterns rather than confirming pre-existing theories.

• Prevents confirmation bias (collecting only what supports a favored explanation)
• Preserves information that any single theory might overlook
• Enables retrospective analysis using frameworks not yet developed
• Allows comparison of explanatory power across different theoretical approaches

Data Analysis: Multi-Disciplinary Interpretation
Once data is collected, analysis proceeds through multiple disciplinary lenses. Organizational mortality has been studied in fragments — economists examine market failures, psychologists study founder grief, sociologists analyze institutional pressures, ecologists model population dynamics (Amankwah-Amoah, 2016). Each discipline illuminates aspects invisible to others.

• Reveals patterns invisible to any single discipline
• Enables comparison of explanatory power across frameworks
• Produces richer understanding through disciplinary triangulation
• Builds cumulative knowledge rather than fragmented disciplinary silos

2.2 The Five Disciplinary Lenses

SOIL's Phase 1 methodology integrates five scientific disciplines. Medicine occupies a special role — providing both the data collection methodology (systematic autopsy protocols) and an analytical lens (diagnostic frameworks). The other four disciplines — Ecology (Hannan, 1977; Hannan, 1989; Carroll, 2000), Psychology (Shepherd, 2003; Shepherd, 2009; Ucbasaran, 2013), Systems Theory (Perrow, 1984; Weick, 1993), and Sociology (Baum, 1991) — primarily contribute analytical frameworks for interpreting collected data.

2.3 The Twelve Lenses: Extended Analytical Framework

The five lenses above represent Phase 1 priorities. The complete framework encompasses twelve perspectives, symbolized by SOIL's navigational symbol — the Roman Dodecahedron, an ancient artifact whose purpose remains unknown, much as organizations often die without understanding why.

A note on terminology: This list includes both established scientific disciplines (Biology, Psychology, Economics) and theoretical approaches that draw on multiple disciplines (Systems Theory, Cybernetics). We use “lenses” rather than “disciplines” to acknowledge this diversity — what unites them is their capacity to reveal different aspects of organizational mortality.

2.4 Temporal Dynamics Model

Organizational mortality is not an event but a process (Weitzel, 1989; Hambrick, 1988; Amankwah-Amoah, 2016). Our theoretical model posits three distinct phases requiring different data collection approaches:

• Phase 1: Genesis and Early Development. The period from founding through initial operations. Data collection focuses on founder characteristics, founding conditions, initial resource endowments, and early strategic choices (Stinchcombe, 1965).
• Phase 2: Peak Operations. The period when the organization achieved its highest functionality. This becomes our primary temporal anchor for comparative analysis — capturing organizational structure and health at maximum capability.
• Phase 3: Decline and Termination. The period from peak through closure. Data collection focuses on degradation dynamics, crisis events, response patterns, and terminal processes (Hambrick, 1988; Weitzel, 1989).

3. Research Design Philosophy

3.1 Mixed Methods Integration

SOIL employs a convergent mixed-methods design, collecting both quantitative and qualitative data through integrated instruments rather than parallel streams.

Quantitative Components: Structured organizational metrics (headcount, revenue, funding, etc.), standardized function assessment scales, timeline markers and durations, Likert-scale health indicators.

Qualitative Components: Open-ended narrative responses, critical incident descriptions, meaning-making and attribution accounts, contextual explanations.

Integration Strategy: Quantitative data provides the skeleton for systematic comparison; qualitative data provides the flesh for understanding mechanisms. Neither alone is sufficient.

3.2 Retrospective Design Considerations

Organizational autopsy is inherently retrospective — we study organizations after they have died. This creates well-known methodological challenges that we address through specific design features.

Challenge: Memory Decay and Reconstruction. Humans reconstruct memories rather than replay recordings. Over time, memories become simplified, schematized, and influenced by subsequent events and current beliefs.

Mitigations: Temporal anchoring at Peak Operations (a salient, typically positive memory), multiple external memory cues (documents, timelines, contemporaneous records), focus on structural/factual data before interpretive accounts, collection of supporting documentation where available, and multi-stakeholder verification for key facts.

3.3 Unit of Analysis

The primary unit of analysis is the organizational mortality episode — the complete arc from an organization's founding through its termination, as documented through autopsy.

Secondary units include:

• Functional subsystems within organizations
• Critical events within organizational timelines
• Stakeholder perspectives on the same organization
• Environmental contexts surrounding multiple organizations

3.4 Comparison Strategy: Survivors and Near-Death Cases

Understanding why organizations die requires comparison with organizations that survived similar conditions. SOIL addresses this through a multi-pronged comparison strategy including near-death survivor analysis and matched comparison design.

4. Data Collection Methodology

4.1 Sampling Strategy

Target Population: Organizations that have permanently ceased operations after a period of active functioning.

Inclusion Criteria:

• At minimum: 2+ people involved at some point OR external stakeholders (paying customers, investors, suppliers)
• Clearly definable founding and termination dates
• Identifiable founder(s) or senior leader(s) willing to participate
• Sufficient documentation or verification sources

Exclusion Criteria:

• Solo side projects without external stakeholders
• Organizations still operating (even if transformed)
• Mergers and acquisitions (unless the original entity genuinely ceased to exist)
• Temporary projects with predetermined endpoints

4.2 Data Collection Instruments

Data collection occurs through a modular wizard system, with each module capturing distinct aspects of organizational life and death:

4.3 Verification System

Verification addresses a fundamental challenge in self-report research. Each cenotaph requires 3 confirmations from verifiers before publication.

Verifier Categories: Former employees, former co-founders, former customers, former suppliers, former investors, former partners.

4.4 Data Quality Tiers

5. Validity and Reliability Framework

5.1 Construct Validity

Challenge: Do our instruments measure what they purport to measure?

• Content Validity: Instrument development informed by multiple organizational theory traditions, expert review, and iterative refinement.
• Convergent Validity: Multiple indicators for key constructs, correlation analysis, multi-stakeholder agreement.
• Discriminant Validity: Distinct constructs measured by distinct instruments, factor analysis, examination of unexpected correlations.

5.2 Internal Validity

Challenge: Can we validly draw causal inferences from retrospective data?

Approaches: Explicit temporal markers throughout data collection, event sequencing within and across modules, documentary evidence of timing where available, systematic collection of environmental factors, and pattern matching across cases.

Limitations Acknowledged: Retrospective design fundamentally limits causal inference. We frame findings as patterns and associations requiring prospective validation. Strong causal claims require prospective follow-up studies.

5.3 External Validity

Challenge: Do findings generalize beyond our sample?

Approaches: Stratified sampling with deliberate inclusion across organization types, regions, sizes, and sectors. Subgroup analysis to examine pattern consistency. Transparent reporting of sample composition.

5.4 Reliability

For qualitative data coding, we employ rigorous Inter-Rater Reliability (IRR) protocols: Krippendorff's alpha (α ≥ 0.70 for exploratory, α ≥ 0.80 for confirmatory), standardized coder training, calibration checks, disagreement resolution processes, and drift monitoring.

6. Analytical Approaches

6.1 Descriptive Analytics

Building on organizational demography methods (Carroll, 2000), we employ:

• Mortality demographics: Distribution of deaths by organization type, region, stage, sector
• Functional profiles: Function presence rates, health indicator distributions, formalization patterns
• Financial patterns: Typical trajectories, crisis event frequency, response pattern effectiveness

6.2 Pattern Recognition

Failure Archetype Identification: Using cluster analysis and latent class methods to identify recurring mortality patterns (Weitzel, 1989). What combinations of dysfunctions commonly co-occur? What temporal sequences characterize different mortality pathways?

Early Warning Signal Detection: Using sequence analysis and survival modeling (Levinthal, 1991; Barron, 1994). What observable indicators precede mortality? How much lead time do different indicators provide?

6.3 Cross-Disciplinary Comparison

The theory-neutral data collection enables systematic comparison of explanatory power through deriving predictions from each discipline, operationalizing predictions using collected data, assessing empirical support, and developing integrative synthesis where warranted.

6.4 Predictive Modeling

• Survival Analysis: Cox proportional hazards and related methods for time-to-failure prediction
• Machine Learning: Classification and regression methods for mortality probability prediction
• Natural Language Processing: Sentiment trajectories, topic modeling, linguistic marker analysis

6.5 Qualitative Analysis

Thematic analysis for emergent themes, process tracing for mechanism identification, and case comparison for theory development through contrast.

7. Ethical Framework

7.1 Core Principles

• Dignity: Every founder and organization deserves respectful treatment. Failure is human; our methodology honors rather than exploits this reality.
• Autonomy: Participants control their data. Consent is informed, specific, and revocable.
• Beneficence: Research should benefit participants (through therapeutic value) and society (through knowledge generation).
• Non-maleficence: Research must not harm participants, named individuals, or the broader community.
• Justice: Benefits and burdens of research should be fairly distributed.

7.2 Privacy and Confidentiality

• Encryption at rest and in transit
• Access controls based on role
• Audit logging of all data access
• Compliance with applicable data protection regulations (GDPR, CCPA, etc.)
• Default anonymization in all research outputs

7.3 Therapeutic Benefit

SOIL explicitly designs for therapeutic benefit, not merely harm minimization (Shepherd, 2003; Shepherd, 2009):

• Closure: The structured reflection process facilitates psychological closure on failure experience (Cope, 2011).
• Meaning-Making: Narrative modules support constructive interpretation of experience (Ucbasaran, 2013).
• Contribution: Framing participation as valuable contribution to knowledge provides redemption narrative.
• Community: Connection with others who share similar experiences reduces isolation.
• Recognition: Beautiful memorialization validates the effort invested (Harris, 1986).

8. Limitations and Mitigation Strategies

We acknowledge several methodological challenges inherent in studying organizational mortality (Mantere, 2013; Jenkins, 2016):

9. Research Agenda

10. Invitation to Collaboration

10.1 Why Collaboration is Essential

SOIL does not claim to have solved the methodological challenges of studying organizational mortality. This white paper presents our current best thinking — a foundation for critique, refinement, and collaborative development.

We recognize that:

• No single team possesses all relevant expertise
• Methodological blind spots are inevitable without external review
• Scientific credibility requires community validation
• The field will develop faster through collaboration than competition

10.2 Disciplinary Expertise We Seek

SOIL's cross-disciplinary approach requires expertise we do not possess. We explicitly invite collaboration from researchers in our priority disciplines:

10.3 Open Questions for Collaborative Development

We explicitly invite collaboration on the following methodological challenges: Comparison strategy design, causal inference from retrospective data, cross-cultural validity, longitudinal design integration, AI/ML methodological integration, and therapeutic-research balance.

10.4 Forms of Collaboration

• Academic Partnerships: Joint research projects, methodological critique, instrument validation studies
• Data Access: Anonymized dataset access for hypothesis testing, collaborative analysis projects, dissertation research support
• Practitioner Input: Instrument validation from professional experience, pattern validation from advisory/VC practice
• Methodological Review: Expert critique of instruments and protocols, peer review of analytical approaches

10.5 How to Engage

Researchers, practitioners, and methodologists interested in collaboration are invited to contact SOIL:

• Academic partnership inquiries: research@soilplatform.org
• Methodological feedback: methodology@soilplatform.org
• General inquiries: hello@soilplatform.org

11. Conclusion

11.1 Contribution to Knowledge

SOIL addresses a fundamental gap in organizational science: the systematic, empirical study of organizational mortality. By treating organizational death with the same rigor that medicine brings to human death, we aim to build the foundation for a new discipline — Organizational Medicine.

Methodological Contributions: Theory-neutral data collection enabling multi-disciplinary analysis through five scientific lenses, multi-stakeholder verification addressing self-report validity, integration of therapeutic and research objectives, temporal modeling distinguishing genesis, peak, and decline phases, and comprehensive coverage across organizational functions and dynamics.

Substantive Contributions: First large-scale, verified database of organizational mortality, empirical basis for failure archetype taxonomy, identification and validation of early warning indicators, comparison of disciplinary and theoretical explanatory power, and foundation for predictive modeling and prevention.

11.2 Practical Implications

• For Founders: Therapeutic closure process, learning from others' experiences, reduction of failure stigma, community and support
• For Practitioners: Evidence-based portfolio risk assessment, early warning monitoring frameworks, intervention targeting, due diligence enhancement
• For Educators: Empirically grounded case material, failure literacy curriculum, organizational health assessment training
• For Policymakers: Understanding of organizational mortality in different contexts, evidence for support program design, ecosystem health assessment frameworks

11.3 Closing Reflection

“Just as the systematic study of human death gave rise to modern medicine, so the systematic study of organizational death may give rise to organizational medicine — the scientific understanding, prediction, and prevention of organizational mortality.”

Every organizational death represents an investment of human time, energy, creativity, and hope. Most of this investment is currently wasted — the lessons disappear with the organization. SOIL aims to transform this waste into wisdom, creating a systematic foundation for understanding why organizations die and how we might help more of them live.

The methodology presented in this white paper is not final. Science advances through iteration, critique, and improvement. We offer this framework as a serious beginning, not a definitive answer. We welcome the engagement of the scholarly community in refining these methods and building this new field together.

References

The following works are cited throughout this white paper. Click on any reference to access the original publication.