APQC PCF Analysis & Documentation Project Plan

Project: APQC Process Classification Framework - AI Applicability Analysis Status: Phase 1 Complete (104 documents, 2.1MB) Date: 2026-01-24

---

1. Project Overview

1.1 Objective

Transform the APQC Process Classification Framework (PCF) Cross-Industry Excel dataset into structured markdown documentation for AI applicability analysis, enabling identification of automation opportunities across enterprise processes.

1.2 Source Data

Asset	Version	Format	Size
Excel Dataset	v7.2.1 (April 2023)	XLSX, 19 sheets	~2MB
PDF Reference	v7.4 (January 2025)	PDF, 35 pages	775KB

1.3 Scope

• 1,855 process elements across 13 major enterprise categories
• 1,856 glossary definitions
• 227 benchmarkable processes with APQC metrics
• Version change tracking (v7.2.1 vs v6.1.1)

---

2. Project Phases

Phase 1: Source Data Analysis ✅ COMPLETE

Duration: Session 1

Activities:

1. Read and analyze Excel file structure (19 sheets)
2. Identify data schema and hierarchy levels
3. Map PCF ID system and numbering convention
4. Extract metadata (version, publisher, dates)

Outputs:

• Understanding of 5-level hierarchy (Category → Group → Process → Activity → Task)
• Identification of PCF ID numbering scheme (5-digit unique IDs)
• Change tracking columns identified (+NEW, -removed, cChanged)

---

Phase 2: Master Document Generation ✅ COMPLETE

Duration: Session 1

Activities:

1. Create master overview document (APQC-PCF-OVERVIEW.md)
2. Create 13 monolithic category documents
3. Create README.md with navigation
4. Create CLAUDE.md with context for AI assistants

Outputs:

Document	Purpose	Size
APQC-PCF-OVERVIEW.md	Master index, AI applicability summary	6KB
category-{01-13}-*.md	Full category content	13 files
README.md	Directory documentation	6KB
CLAUDE.md	AI assistant context	4KB

---

Phase 3: Hierarchical Document Breakdown ✅ COMPLETE

Duration: Session 1

Activities:

1. Create 13 category subdirectories
2. Generate README.md for each category
3. Create individual process group documents (72 files)
4. Maintain complete hierarchy in each document

Outputs:

category-{01-13}-*/
├── README.md           # Category overview
├── X.1-*.md           # Process group 1
├── X.2-*.md           # Process group 2
└── ...                 # All process groups

Bug Fixed:

• Regex pattern f"^{cat_num}\\.\d+$" incorrectly matched X.0 headers
• Fixed to f"^{cat_num}\\.[1-9][0-9]*$" for proper process group matching

---

Phase 4: Completeness Verification ✅ COMPLETE

Duration: Session 1

Activities:

1. Build verification script to compare Excel vs generated docs
2. Count elements per category in both sources
3. Identify any missing or orphan elements
4. Generate completeness report

Results:

Category	Excel Elements	Generated	Status
1.0	123	123	✅
2.0	100	100	✅
3.0	202	202	✅
4.0	147	147	✅
5.0	67	67	✅
6.0	104	104	✅
7.0	135	135	✅
8.0	321	321	✅
9.0	270	270	✅
10.0	69	69	✅
11.0	56	56	✅
12.0	54	54	✅
13.0	207	207	✅
Total	1,855	1,855	✅ 100%

---

Phase 5: Enhanced Data Extraction ✅ COMPLETE

Duration: Session 1

Activities:

1. Extract glossary definitions (1,856 terms)
2. Identify benchmarkable processes (227 with metrics)
3. Document version changes (NEW, additions, removals)
4. Integrate definitions into process group documents

Outputs:

Document	Content	Size
APQC-PCF-GLOSSARY.md	All 1,856 definitions	489KB
APQC-PCF-CHANGE-ANALYSIS.md	Version change tracking	33KB
APQC-PCF-BENCHMARKABLE-METRICS.md	227 measurable processes	32KB

Change Statistics:

Change Type	Count
NEW elements in v7.2.1	845
Additions (+)	227
Removals (-)	24
Modifications (c)	74

---

Phase 6: PDF Version Analysis 🔄 IN PROGRESS

Duration: Session 2 (current)

Activities:

1. Locate corresponding PDF document ✅
2. Read and parse 35-page PDF ✅
3. Convert to page-by-page markdown ⏳ PENDING
4. Compare v7.2.1 (Excel) vs v7.4 (PDF) ⏳ PENDING

Status:

• PDF successfully read (35 pages)
• Context limit reached during conversion
• Session compacted, work continuing

---

Phase 7: Classification & Quality Assurance ⏳ PENDING

Activities:

1. Run /classify on all 104 generated documents
2. Verify frontmatter compliance
3. Update moe_confidence scores
4. Review AI applicability assessments

---

3. Deliverables Summary

3.1 Final Output

Metric	Value
Total Documents	104
Total Size	2.1 MB
Process Elements	1,855
Definitions	1,856
Benchmarkable	227 (12%)
Categories	13
Process Groups	72
Hierarchy Depth	5 levels

3.2 Directory Structure

apqc-pcf-enterprise-processes/
├── APQC-PCF-OVERVIEW.md              # Master index
├── APQC-PCF-GLOSSARY.md              # 489KB definitions
├── APQC-PCF-CHANGE-ANALYSIS.md       # Version tracking
├── APQC-PCF-BENCHMARKABLE-METRICS.md # Measurable processes
├── README.md                          # Documentation
├── CLAUDE.md                          # AI context
├── PROJECT-PLAN-APQC-PCF-ANALYSIS.md # This document
├── category-01-develop-vision-and-strategy/
│   ├── README.md
│   ├── 1.1-*.md through 1.4-*.md
├── category-02-develop-and-manage-products-*/
│   ├── README.md
│   ├── 2.1-*.md through 2.3-*.md
... (13 category subdirectories, 72 process group files)

---

4. AI Applicability Analysis

4.1 High-Value Automation Targets

Category	% Benchmarkable	AI Application
9.0 Financial Resources	32%	Transaction processing, reconciliation, reporting
7.0 Human Capital	23%	Resume screening, L&D personalization
3.0 Marketing & Sales	17%	Lead scoring, personalization, analytics
6.0 Customer Service	13%	Chatbots, ticket routing, sentiment analysis

4.2 Automation Categories

Category	AI Fit	Rationale
8.0 IT Management	Full Automation	Rule-based, already digitized
9.0 Financial	High Automation	High-volume transactions
6.0 Customer Service	Chatbots + Routing	Repetitive inquiries
3.0 Marketing	Analytics + Personalization	Data-driven decisions
7.0 Human Capital	Augmentation	AI screens, humans decide
1.0 Strategy	Augmentation	AI provides data, humans judge
4.0 Physical Products	IoT + AI	Hybrid physical-digital

---

5. Technical Implementation

5.1 Tools Used

Tool	Purpose
Python + pandas	Excel parsing and data extraction
openpyxl	Excel file reading
Claude Code	Document generation and analysis
CODITECT Framework	Document standards and classification

5.2 Key Scripts

# Excel reading pattern
import pandas as pd
df = pd.read_excel(file_path, sheet_name='Sheet Name')

# Hierarchy pattern matching
level_1 = df[df['Hierarchy ID'].astype(str).str.match(f"^{cat_num}\\.[1-9][0-9]*$")]

5.3 Bug Fixes

Issue	Cause	Solution
Only READMEs created	Regex matched X.0 headers	Changed to [1-9][0-9]*
Binary file error	Read tool can't read XLSX	Used pandas
Missing pandas	Module not installed	Activated venv, pip install

---

6. Next Steps

6.1 Immediate (This Session)

• Convert PDF v7.4 to page-by-page markdown
• Compare v7.2.1 vs v7.4 changes
• Run /classify on generated documents

6.2 Future Enhancements

• Create process-to-occupation mapping (O*NET integration)
• Build AI applicability scoring model
• Generate automation ROI estimates per process
• Create interactive visualization of process hierarchy

---

7. References

7.1 Source Documents

• K08897_CrossIndustry_v721_vs_v611_April 2023.xlsx - Primary data source
• K014750_APQC Process Classification Framework (PCF) - Cross Industry - PDF Version 7.4_January 2025.pdf - Reference PDF

7.2 Related Research

• O*NET Occupation Database analysis (same parent directory)
• CODITECT AI Applicability research framework
• Process automation ROI methodology

7.3 APQC Resources

• Publisher: American Productivity & Quality Center (APQC)
• Website: apqc.org
• PCF Version: 7.2.1 Cross-Industry

---

Document Version: 1.0.0 Created: 2026-01-24 Author: CODITECT Research Team Status: Phase 1 Complete, Phase 6-7 In Progress