📅 2025-10-11 11:00
🕒 Reading time: 19 min
🏷️ PDCA
The week following the resolution of AgriTech Solutions Africa's JTBD utilization case, a consultation regarding continuous improvement in drug development arrived from Europe. The eighth case in Volume 19 "New Frontiers of Analysis" concerned challenges of organizational improvement that learns from failures and increases success probability.
"Detective, we are a European pharmaceutical company engaged in innovative drug development, but clinical trial failures continue, and we feel there are fundamental problems in our development process. While we can learn from failures, we lack systematic mechanisms to connect that to future success."
Helmut Schmidt, from Germany and Head of R&D at BioPharmaTech Europe, visited 221B Baker Street with a serious expression. In his hands, he held clinical trial data from the past five years and accompanying voluminous failure analysis reports.
"We are a company conducting research and development of next-generation biopharmaceuticals across Europe. While scientifically at a high level, process improvement remains ad-hoc, and we can't learn and grow as an organization."
BioPharmaTech Europe's R&D Capabilities and Challenges: - Founded: 2016 (emerging European biopharma company) - Research Areas: Innovative therapeutics for cancer, immunology, neurodegenerative diseases - Development Pipeline: 15 preclinical, 12 in clinical trials - Research Investment: 20 billion yen annually (85% of revenue to R&D) - Researcher Count: 450 (85% with PhDs, highly skilled personnel)
The numbers certainly indicated strong commitment to R&D. However, Helmut's expression bore deep concern.
"The problem is that while individual researchers are excellent and conduct detailed analysis of failed trials, those learnings aren't accumulated organization-wide and don't lead to improved success rates in future projects."
Contradiction of High R&D Investment and Low Success Rate: - Clinical Trial Success Rates: Phase I 65%, Phase II 28%, Phase III 45% (below industry average) - Development Period: Average 12 years (longer than 10-year industry average) - Development Cost: Average 18 billion yen per drug (exceeds 15 billion yen industry average) - Approval Acquisition: Only 2 cases in past 5 years (investment effectiveness challenges) - Learning Utilization: 80% of failure analyses not utilized in next projects
"We conduct 'high-quality failure analyses' but can't achieve 'continuous improvement.' We keep repeating similar failures."
"Mr. Helmut, in your current R&D process, how are improvements and learning conducted?"
Holmes quietly inquired.
Helmut began explaining the current situation with a confused expression.
"We conduct detailed reviews at project completion and create reports. However, mechanisms don't exist to systematically reflect those insights into next project planning."
Current R&D Process (Absence of Improvement System):
Plan Stage: - Research Hypothesis: Hypothesis setting based on literature and prior research - Trial Design: Trial protocol design emphasizing scientific validity - Resource Planning: Budget, personnel, timeline planning - Issue: Past failure experiences not reflected in planning
Do Stage: - Preclinical Trials: Animal testing, safety evaluation implementation - Clinical Trials: Staged implementation of Phase I-III - Data Collection: Trial results, adverse effects, efficacy data collection - Issue: Problem discovery and response during execution always reactive
Check Stage: - Result Analysis: Statistical analysis, scientific evaluation implementation - Failure Factor Analysis: Detailed causal analysis when failed - Report Creation: Detailed analysis report creation and storage - Issue: Analysis results end with individual projects
Action Stage: - Individual Response: Localized improvement in relevant project - Lesson Organization: Documentation and storage of failure factors - Next Preparation: Individual preparation for next project - Issue: Insufficient organizational improvement and standardization
I focused on the lack of learning accumulation at each stage.
"PDCA elements are being implemented, but continuity as cycles and organizational learning are insufficient."
Helmut answered with a serious expression.
"That's exactly right. We execute the 'four letters' but don't have the 'cycle.' It's like starting from zero every time."
Specific Problem Cases from PDCA Dysfunction:
Cancer Immunotherapy Drug Development Project (2019-2023):
Phase I (2019): Safety Confirmation Trial - Plan: Safe dosage setting based on literature - Do: Safety trial with 24 patients - Check: Unexpected hepatotoxicity in 10 patients - Action: Reduced dosage by 50% to proceed to Phase II
Phase II (2021): Efficacy Confirmation Trial - Plan: Efficacy evaluation at reduced dosage - Do: Efficacy trial with 150 patients - Check: Insufficient efficacy (18% response rate, target 30%) - Action: Considered dosing schedule modification
Phase III (2023): Final Confirmation Trial - Plan: Large-scale trial with modified dosing schedule - Do: Started comparison trial with 600 patients - Check: Interim analysis showed efficacy criteria not met - Action: Trial discontinued, development abandoned
Fundamental Problem: PDCA Cycle Dysfunction - Plan Stage: Past hepatotoxicity data from similar drugs not reflected in planning - Do Stage: No early warning detection system - Check Stage: No comparative analysis with other projects - Action Stage: Learnings not utilized in next projects
"A project with 4 years and 18 billion yen investment, but we lack mechanisms to utilize learnings in next projects."
"Failure is learning opportunity. PDCA transforms individual failures into organizational wisdom."
"True improvement lies in repetition. Cycles are the source of growth."
"PDCA is not mere procedure. It's management philosophy for creating learning organizations."
The three members began their analysis. Gemini displayed a "Pharmaceutical Research-Specific PDCA Design" framework on the whiteboard.
Pharmaceutical R&D-Specific PDCA Design Principles: - Plan: Hypothesis and design integrating past learnings - Do: Real-time monitoring and early detection system - Check: Multi-faceted analysis, comparative evaluation, learning extraction - Action: Organizational learning accumulation, standardization, next reflection
"Mr. Helmut, let's redesign BioPharmaTech Europe's R&D process with learning-accumulation PDCA cycles."
BioPharmaTech Europe's Learning-Accumulation PDCA Redesign:
Enhanced Plan Stage: "Learning-Integrated Planning"
Traditional Plan (Individual Planning): - Literature Survey: Public paper and competitor information collection - Hypothesis Setting: Hypotheses based on scientific theory - Trial Design: Statistically valid design emphasis - Issue: Internal past experiences not reflected
Improved Plan (Learning-Integrated Planning): - Internal Knowledge Base Utilization: Past project failure and success pattern analysis - Risk Prediction Model: Problem occurrence probability prediction for similar compounds and trials - Design Optimization: Trial design improvement based on past learnings - Early Discontinuation Criteria: Clear judgment criteria to avoid wasteful investment
Specific Improvement Content: - Compound Database: Internal data accumulation of molecular structure, toxicity, efficacy - Failure Pattern Analysis: Systematization of failure factors over past 20 years - Success Element Extraction: Common success element analysis of approved drugs - Risk-Benefit Evaluation: Quantitative GO/NO-GO judgment criteria
Enhanced Do Stage: "Real-time Learning Execution"
Traditional Do (Plan Execution): - Protocol Compliance: Trial execution as planned - Data Collection: Regular data acquisition - Problem Response: Individual response when problems occur - Issue: Early problem discovery and response difficult
Improved Do (Learning Execution): - Real-time Monitoring: Anomaly early detection through AI and data analysis - Adaptive Modification: Swift plan modifications upon problem discovery - Parallel Learning: Continuous learning and improvement during execution - Preventive Response: Advance countermeasures through problem prediction
Specific Improvement Content: - Biomarker Monitoring: Real-time tracking of early indicators for toxicity and efficacy - AI Prediction System: Predicting adverse effects and results from patient data - Adaptive Trials: Real-time trial design modification based on interim results - Early Warning System: Monitoring similarity to past failure patterns
Enhanced Check Stage: "Multi-dimensional Learning Evaluation"
Traditional Check (Result Evaluation): - Statistical Analysis: Statistical evaluation of primary endpoints - Safety Evaluation: Adverse events and safety assessment - Efficacy Evaluation: Scientific evaluation of therapeutic effects - Issue: One-time evaluation in individual projects
Improved Check (Learning Evaluation): - Comparative Analysis: Comparison with similar projects and competitor drugs - Meta-analysis: Cross-project pattern extraction - Prediction Accuracy Evaluation: Gap analysis between advance predictions and actual results - Learning Value Evaluation: Quantifying learning value from failures and successes
Specific Improvement Content: - Cross-project Data Analysis: Integrated analysis of all project data - Failure Factor Classification: Systematic failure factor categorization and pattern analysis - Success Probability Model: Success probability prediction model improvement at each stage - Investment Effect Analysis: Learning effects and future value of research investments
Enhanced Action Stage: "Organizational Learning Accumulation"
Traditional Action (Individual Improvement): - Report Creation: Documentation of detailed analysis results - Lesson Organization: Organizing and storing failure factors - Next Planning: Individual planning for next project - Issue: Insufficient organizational learning accumulation and utilization
Improved Action (Organizational Learning): - Knowledge Base Update: Sharing and accumulating learning content organization-wide - Standardization Promotion: Standardizing successful patterns into processes - Education and Training: Educational deployment of learning content organization-wide - Next Plan Integration: Mandatory reflection of learning outcomes in next projects
Specific Improvement Content: - Learning Management System: Integrated management of all project learnings - Best Practice Standardization: Standardizing successful patterns into processes - Researcher Education Program: Regular education on failure and success cases - Plan Review System: Reviewing degree of past learning reflection
Claude reported an important discovery.
"This is clear. BioPharmaTech Europe implements individual PDCA elements but lacks continuity as cycles and organizational learning accumulation. Learning-accumulation PDCA transforms failures into organizational wisdom."
Most Important Discovery: "Individual PDCA vs. Learning-Accumulation PDCA"
Previously executed independent PDCA for each project, but cross-project learning accumulation and utilization weren't happening. Organization-wide continuous learning system needed.
Following detailed learning-accumulation PDCA analysis and R&D process redesign, BioPharmaTech Europe's continuous improvement strategy became clear.
Transition from "Individual Project PDCA" to "Organizational Learning PDCA Cycles":
Core Problem: Learning Individualization and Dissipation
BioPharmaTech Europe executed high-quality PDCA for each project, but mechanisms didn't exist for that learning to be organizationally accumulated and utilized, leading to repeated similar failures.
Learning-Accumulation PDCA Implementation Strategy:
Phase 1: PDCA Foundation System Construction (6 months)
Organizational Learning Infrastructure Construction: - Integrated Knowledge Base: Centralized management of all project learnings - AI Learning System: Continuous learning of pattern recognition and prediction models - Real-time Monitoring: Early anomaly and risk detection in project progress - Comparative Analysis System: Multi-faceted comparison with similar projects and competitors
Learning Accumulation Process Standardization: - Plan Stage Standard: Mandatory reference and reflection processes for past learnings - Execution Stage Standard: Real-time learning and adaptive modification processes - Evaluation Stage Standard: Multi-faceted analysis and learning extraction processes - Improvement Stage Standard: Organizational learning accumulation and next reflection processes
Phase 2: Learning Cycle Implementation (12 months)
Cross-Project Learning System:
Plan Integration: Past Learning Integration Planning - Internal Failure Database: Past 15-year failure pattern and factor analysis - Success Element Model: Common success element and condition analysis of approved drugs - Risk Prediction AI: Risk prediction from compound characteristics and trial design - GO/NO-GO Criteria: Early decision-making through quantitative judgment criteria
Do Integration: Adaptive Execution Learning - Biomarker Integrated Monitoring: Multi-dimensional real-time tracking of toxicity and efficacy - Adaptive Trial Design: Real-time trial design modification based on interim results - Preventive Intervention: Early countermeasures through similarity to past patterns - Parallel Project Learning: Learning sharing across simultaneous projects
Check Integration: Multi-dimensional Learning Evaluation - Meta-analysis System: Cross-project cross-sectional pattern extraction - Competitor Comparative Analysis: Position and differentiation factor analysis across industry - Prediction Accuracy Evaluation: Advance prediction model accuracy verification and improvement - Learning ROI Evaluation: Future value quantification of learnings from failures and successes
Action Integration: Organizational Learning Accumulation - Knowledge Base Auto-update: Automatic classification, accumulation, and search system for learning content - Best Practice Standardization: Organizational standard process creation from success patterns - Researcher Continuous Education: Regular education and skill updates on latest learnings - Next Project Forced Integration: Mandatory reflection of past learnings in next planning
Phase 3: Continuous Improvement Culture Establishment (Ongoing)
Learning Organization Culture Construction: - Failure-Welcoming Culture: Positive evaluation of failures as learning opportunities - Knowledge Sharing Promotion: Sharing individual and team learnings as organizational assets - Continuous Improvement Mindset: Pursuing continuous improvement rather than status quo satisfaction - Long-term Perspective Emphasis: Prioritizing long-term learning and growth over short-term results
External Learning Integration: - Industry Learning Integration: Organizational learning integration of industry-wide failure and success cases - Academic Collaboration: Latest knowledge sharing with universities and research institutions - Regulatory Authority Collaboration: Organizational accumulation of learnings from approval reviews - Patient Feedback: Long-term learning of actual clinical drug effects
Comparison with Successful Companies:
PDCA-Utilizing Successful Company (Swiss Company A): - Same scale and field pharmaceutical company - Before PDCA Enhancement: Stagnant success rates, prolonged development periods - After PDCA Enhancement: 30% success rate improvement, 25% development period reduction - Success Factors: Organizational learning accumulation, continuous improvement culture
BioPharmaTech Europe's Improvement Potential: Significant success probability and efficiency improvements expected through similar approach
Holmes summarized the comprehensive analysis.
"Mr. Helmut, PDCA cycle's essence is 'learning organization' construction. Not individual excellent analyses, but organization-wide continuous learning and improvement transforms failures into wisdom and increases success probability. PDCA is not mere procedure but management philosophy for creating organizations that keep learning."
Learning-Accumulation PDCA Strategy: Transition from "Individual Improvement" to "Organizational Continuous Learning"
Basic Strategic Policy: Organizational Learning Excellence
Phase 1: Learning Foundation System Construction (6 months)
Integrated Learning Platform: - Knowledge Integration System: Integrated management of all project learnings - AI Learning Engine: Continuous improvement of pattern recognition and prediction accuracy - Real-time Monitoring: Early anomaly and risk detection and response - Comparative Analysis Engine: Multi-faceted comparison of internal and external data
Learning Process Standardization: - Plan Standard: Mandatory integration process for past learnings - Do Standard: Adaptive execution and real-time learning process - Check Standard: Multi-dimensional evaluation and learning extraction process - Action Standard: Organizational learning accumulation and utilization process
Phase 2: Continuous Learning Cycle Implementation (12 months)
Cross-organizational Learning System: - Failure Learning Integration: Pattern extraction and countermeasure development from all failure cases - Success Learning Integration: Success factor systematization and reproducibility assurance - Prediction Accuracy Improvement: Model improvement and prediction accuracy enhancement - Adaptive Improvement: Learning and improvement system evolution according to environmental changes
Culture and Organizational Transformation: - Learning-Emphasis Evaluation: Learning contribution as important indicator in individual and team evaluation - Failure Active Utilization: Making failures learning resources rather than hiding them - Knowledge Sharing Promotion: Actively evaluating individual knowledge organizational assetization - Long-term Perspective Management: Prioritizing long-term organizational learning capability over short-term results
Phase 3: Learning Organization Completion (Ongoing)
Sustainable Improvement System: - PDCA Self-improvement: Continuous improvement of PDCA process itself - Learning Efficiency: Learning process efficiency and advancement - External Learning Integration: Latest learning integration from industry and academia - Next Generation Preparation: Learning system evolution responding to future changes
Expected Effects: - Clinical Trial Success Rates: Phase II 28% → 45%, Phase III 45% → 65% - Development Period: Average 12 years → 9 years (25% reduction) - Development Cost: 18 billion yen per drug → 14 billion yen (efficiency effect) - Approval Acquisition: 2 cases annually → 5 cases annually (success probability improvement)
Investment Plan: - PDCA Learning System Construction: 2.5 billion yen annually - Expected Benefits: 12 billion yen annually (efficiency + success probability improvement) - Investment Recovery Period: 3 months
"What's important is not fearing failures but continuing to learn from them. PDCA is the perpetual learning engine for organizations to continue growing."
24 months later, a report arrived from BioPharmaTech Europe.
R&D Transformation Results through Learning-Accumulation PDCA Introduction:
Dramatic R&D Efficiency Improvement: - Clinical Trial Success Rates: Phase I 65% → 78%, Phase II 28% → 47%, Phase III 45% → 68% - Development Period: Average 12 years → 8.5 years (30% reduction) - Development Cost: 18 billion yen per drug → 12.5 billion yen (efficiency and early decision effects) - Approval Acquisition: 7 cases in past 2 years (3.5x from previous 2 annually)
Learning System Success:
Plan Stage Innovation: - Failure Prediction Accuracy: 85% (advance discovery of similar patterns) - Plan Accuracy Improvement: 50% reduction in gap between initial plan and actual results - Early GO/NO-GO Decision: 40% reduction in wasteful investment - Risk Advance Countermeasures: 70% prevention of serious problems
Do Stage Innovation: - Real-time Monitoring: 75% reduction in problem discovery time - Adaptive Modification: 20% success rate improvement through interim result design modifications - Preventive Response: 80% early detection and countermeasures for adverse effects and efficacy problems - Parallel Learning: 30% efficiency improvement through simultaneous project learning effects
Check Stage Innovation: - Multi-faceted Analysis: 90% discovery of success and failure factors previously overlooked - Prediction Model Accuracy: 85% next-stage success prediction accuracy (from previous 60%) - Competitor Comparison: Clarification of own position and differentiation factors across industry - Learning Value Evaluation: Future value creation through learnings from failed projects
Action Stage Innovation: - Knowledge Accumulation: 95% of all learnings available as organizational assets - Standardization: 80% standardization of success patterns into processes - Education Effect: Average 40% improvement in researcher skills and judgment - Next Integration: 100% reflection of past learnings in next project planning
Fundamental Organizational Culture Transformation:
Learning Organization Culture Establishment: - Failure Consciousness: "Things to hide" → "Treasure trove of learning" - Knowledge Sharing: 95% of individual knowledge shared as organizational knowledge - Continuous Improvement: "Status quo maintenance" → "Continuous improvement" as organizational standard - Long-term Perspective: Emphasis on long-term learning and growth over short-term results
Researcher Consciousness and Capability Changes: - Learning Motivation: "Avoid failures" → "Learn from failures" consciousness - Analytical Capability: Individual analysis → Cross-organizational, multi-faceted analytical capability - Predictive Capability: Experience-dependent → Scientific prediction based on data and learning - Collaborative Capability: Individual work → Active contribution to organizational learning
Specific Success Cases:
Alzheimer's Disease Treatment Drug Development Project: - Traditional Failure Pattern: Past 3 similar projects failed in Phase II - Learning Utilization: Improved biomarkers and administration methods through failure pattern analysis - Result: Phase II success rate 65% (from previous 20%), Phase III progressing with favorable results
Cancer Immunotherapy Drug Development Project: - Learning-Integrated Planning: Advance identification and countermeasures for past hepatotoxicity patterns - Adaptive Execution: Adverse effect prevention through early biomarker monitoring - Result: Significantly improved safety profile, efficacy also exceeding expected values
Rare Disease Treatment Drug Development Project: - Cross-sectional Learning: Applied success elements from other diseases to rare diseases - Prediction Model: High-accuracy success prediction from small-scale trial data - Result: Development period reduced from 6 years to 4 years, approval acquisition certain
Evaluation Changes from Industry and Investors:
Industry Evaluation Improvement: - Pharmaceutical Industry Evaluation: "High investment low efficiency company" → "Learning-type efficiency company" - Regulatory Authority Evaluation: "Problem-frequent company" → "Preventive quality control company" - Academic Evaluation: "Individual research" → Highly evaluated as "organizational research capability" - Competitor Evaluation: "Technology equivalent but efficiency advantage" → "Differentiated by learning capability"
Investor Evaluation Transformation: - Investment Decision: "High risk high return" → "Stable growth through learning" - Corporate Value: Sustainable competitive advantage through organizational learning capability evaluated - Long-term Investment: Emphasis on long-term learning capability over short-term new drug success - ESG Evaluation: Highly evaluated as sustainable R&D model
Researcher Voices:
Principal Researcher (Pharmacology, 15 years experience): "Previously regretted 'why couldn't we predict' when failures occurred, but now think 'what can we learn from this failure.' Realize failures are precious assets leading to future success."
Clinical Development Director (Medical background, 10 years experience): "Through past similar trial learnings, can provide safer and more effective treatments to patients. Can detect and respond to problems early through real-time monitoring, dramatically improving patient safety."
Biostatistician (Statistics, 8 years experience): "Data analysis changed from one-time work to part of organizational learning. Thinking my analysis contributes to future project success, I conduct more careful and deeper analysis."
New Researcher (PhD graduate, 2 years with company): "Having accumulated organizational learning from joining allows efficiently learning senior experiences. Can challenge without fearing failures in environment, feel growth speed is dramatically faster."
Social Impact Expansion:
Contribution to Patients and Healthcare: - New drug development success rate improvement increases treatment options - Development period reduction improves patient access to therapeutic drugs - Safety improvement significantly reduces adverse effect risks - Rare disease drug development efficiency improves unmet need response
Industry-Wide Impact: - PDCA learning model spreading as pharmaceutical industry best practice - Industry-wide R&D efficiency improvement and success rate improvement contribution - Approval process efficiency improvement contribution through regulatory authority collaboration - Practical application promotion of basic research through academia strengthening
Next Generation Impact: - R&D personnel development model referenced by other companies - Learning organization management model expanding to other industries - Culture of failure assetization spreading to society overall - Sustainable R&D system establishment
Sustainable Growth Foundation: - Learning System: Continuously evolving through self-improvement - Organizational Culture: Learning-emphasis culture inheritance across generations - Competitive Advantage: Differentiation through inimitable organizational learning capability - Social Value: Sustainable growth through patient and society contribution
Helmut's letter contained deep gratitude and conviction in organizational transformation:
"Through PDCA learning system introduction, we evolved from 'individual excellent researcher collective' to 'organization that keeps learning.' Most importantly, rather than hiding failures, we learned from them and shared and utilized those learnings organization-wide. Now every failure becomes foundation for next success, and all researchers continue challenging without fearing failures. Acquired 7 new drug approvals in past 2 years, development period also reduced by 30%. Above all, being able to deliver safer and more effective therapeutic drugs to patients is the greatest result. PDCA is not just improvement method but life force for organizations to perpetually learn and continue growing."
That night, I pondered deeply about organizational learning and continuous improvement.
BioPharmaTech Europe's case clearly showed that even in organizations with high expertise, individual excellence and organizational learning capability are separate things. No matter how excellent researchers exist, if their knowledge isn't organizationally accumulated and utilized, similar failures repeat and growth remains limited.
PDCA cycle's true value, beyond mere improvement methods, lies in constructing systems for organizations to continuously learn and evolve. By accumulating learnings at each Plan→Do→Check→Action stage as organizational wisdom and utilizing them in next cycles, overall organizational capability improves exponentially.
In Volume 19 "New Frontiers of Analysis," while the previous seven cases demonstrated various analytical method powers, Case 248's PDCA analysis proved the importance of organizational evolution through continuous improvement. Analytical methods demonstrate true value not through one-time utilization but within continuous learning cycles.
"True competitive advantage lies not in initial success but in the ability to continue learning from failures."
The next case will surely depict another moment when analytical methods realize organizational continuous evolution.
"Improvement is an endless journey. PDCA cycles are the perpetual engine that reliably advances that journey." — From the Detective's Notes
Solve Your Business Challenges with Kindle Unlimited!
Access millions of books with unlimited reading.
Read the latest from ROI Detective Agency now!
*Free trial available for eligible customers only