Value Modeler

Transform agreed trend candidates into customer-specific, ranked solution recommendations by building TIPS relationship networks and generating Solution Templates — the missing link between trend insight and concrete action.

Based on the TIPS Value Modeler methodology (Siemens patent WO2018046399A1, freely usable).

What This Skill Does

The trend-scout skill produces 60 agreed candidates across 4 dimensions. Each candidate lives in its dimension as a standalone item. This skill connects them:

1. Builds TIPS value chains — explicit Trend → Implication → Possibility causal chains across dimensions
2. Consolidates into Investment Themes (Handlungsfelder) — clusters value chains into 3-7 MECE investment domains
3. Generates Solution Templates with Blueprints — concrete enablers per investment theme with multi-dimensional portfolio composition (natural deduplication)
4. Generates SPIs — operational process changes that accompany each Solution Template
5. Defines success Metrics — KPIs that measure value delivery per investment theme
6. Suggests Collaterals — supporting content (case studies, reference architectures) per ST
7. Enables Business Relevance scoring — customer-specific 1-5 ratings per TIP
8. Ranks solutions automatically — using the F1 averaging formula from the patent
9. Generates a Big Block diagram — the customer-specific solution architecture organized by investment theme
10. Curates catalog feedback — promotes pursuit-specific insights back to industry catalogs

The output is a structured strategy with 3-7 investment themes, each containing ranked solutions backed by trend evidence and scored for business relevance. This gives the customer a CxO-ready investment portfolio — not a flat list of 18 solutions, but 5 distinct areas to fund and champion.

Prerequisites

• A completed trend-scout project with workflow_state: "agreed" in .metadata/trend-scout-output.json
• Recommended: If a cogni-portfolio project exists in the workspace, run /bridge portfolio-to-tips before starting value-modeler. This exports your product features, propositions, and pricing so Phase 2 generates Solution Templates grounded in your actual products. Without it, solutions will be abstract and require manual portfolio mapping later. If trend-scout was linked to a portfolio market, this skill automatically picks up that connection — no need to re-discover.
• No portfolio? If no cogni-portfolio project exists, the value-modeler offers a generic B2B ICT portfolio (7 products, 51 features with IS-layer descriptions and taxonomy mappings) derived from the B2B ICT taxonomy. DOES/MEANS propositions are generated dynamically based on the project's research context (industry, subsector, research topic). This provides taxonomy-grounded solution blueprints with coverage data instead of abstract solutions, giving users a tangible view of which ICT capability dimensions each solution requires. The generic portfolio can be replaced with a company-specific one at any time.

Context Independence

This skill reads ALL required state from project files (trend-scout-output.json, portfolio context, catalog data) — it does not depend on prior conversation context. The trends-resume dashboard and any preceding chat are not inputs to the value-modeling pipeline. This means context compaction is safe and recommended before starting.

Before executing Phase 0, run /compact to free working memory. This skill's phases — especially Phase 1 (relationship networks with extended thinking) and Phase 2 (solution blueprint generation) — need substantial context for reading candidate data and building structured output. Compacting early prevents context pressure from accumulating across the 5-phase workflow.

If /compact is unavailable (e.g., non-interactive mode), proceed without it — the skill will still work, but later phases may hit context limits on projects with many investment themes.

Language Support

This skill follows the shared language resolution pattern — see $CLAUDE_PLUGIN_ROOT/references/language-resolution.md.

Interaction language: Read workspace language from .workspace-config.json (via ${PROJECT_AGENTS_OPS_ROOT}/.workspace-config.json or CWD) at startup. Use this for all user-facing messages — scoring prompts, progress updates, phase summaries, AskUserQuestion prompts, and next-step recommendations.

Output language: Inherited from project_language in trend-scout-output.json. Value-modeler does not re-ask — it is a downstream skill that respects the project language set during trend-scout.

All output files use UTF-8 encoding. When the project language is German, use proper umlauts (ä, ö, ü, ß) — never ASCII substitutes (ae, oe, ue, ss). This applies to all generated markdown, JSON string values, and HTML content.

Workflow

Read the phase reference file for each phase before executing it.

Phase 0: Initialize

Reference: references/workflow-phases/phase-0-load.md

Load the trend-scout output, validate prerequisites, discover optional portfolio.

Phase 1: Build Relationship Networks & Investment Themes

Reference: references/workflow-phases/phase-1-relationships.md

Two-pass architecture: First, build granular T→I→P value chains via semantic affinity analysis (bottom-up). Then consolidate chains into 3-7 MECE Investment Themes — the distinct investment domains where this customer should allocate budget and executive attention (top-down). Each investment theme groups 1-4 value chains and represents a CxO-level strategic decision.

Phase 2: Generate Solution Templates

Reference: references/workflow-phases/phase-2-solutions.md

For each Investment Theme, generate Solution Templates — concrete process improvement enablers. Working at the investment theme level naturally deduplicates STs that would otherwise appear redundantly across overlapping chains. Target 2-4 STs per investment theme. If a cogni-portfolio project exists, map templates to existing products/features. When portfolio context v2.0+ is available, Phase 2 starts with portfolio-anchored generation (Step 0.5): features from the portfolio are matched to investment themes and used as delivery anchors for STs with automatic high-confidence mapping. Each ST gets a solution blueprint — a multi-dimensional composition of building blocks mapped to B2B ICT taxonomy categories (connectivity, security, cloud, applications, consulting, etc.) that captures the full solutioning expertise: what portfolio is needed to BUILD and DELIVER the solution, not just which single feature matches. Blueprint readiness scores (0.0-1.0) surface portfolio gaps and feed into the ranking formula. Quality-aware generation (v3.0) flags STs where underlying propositions need improvement. Re-anchoring (Step 2.7): When portfolio context changes or initial mappings need revision, re-anchor rebuilds each ST's solution blueprint from scratch using LLM solutioning intelligence — re-analyzing which taxonomy categories are needed and re-matching against the current portfolio. This is intellectual solutioning work, not mechanical keyword matching. Can be invoked independently (outside a full Phase 2 run) via "re-anchor solutions". See references/workflow-phases/phase-2-solutions.md Step 2.7.

Phase 3: Business Relevance Scoring

Reference: references/workflow-phases/phase-3-scoring.md

Present the TIPS paths and Solution Templates to the user for customer-specific Business Relevance (BR) scoring on a 1-5 scale. Generate an interactive scoring UI.

Phase 4: Rank & Visualize

Reference: references/workflow-phases/phase-4-rank.md

Apply the F1 formula to calculate solution rankings, generate the ranked solution list, and produce the Big Block solution diagram.

Phase 5: Curate (Optional)

Reference: references/workflow-phases/phase-5-curate.md

Review pursuit-specific data (paths, STs, SPIs, Metrics) and generate recommendations for promoting high-value, reusable patterns back to the industry catalog. This creates a learning loop where customer engagements improve the base catalog over time.

Output Files

{project-dir}/
├── tips-value-model.json              # Complete value model (investment themes + chains + solutions + scores)
├── tips-solution-ranking.md           # Human-readable ranked solution list by investment theme
├── tips-big-block.md                  # Big Block solution diagram organized by investment theme
├── value-modeler-scoring.html         # Interactive BR scoring UI grouped by investment theme
└── .metadata/
    └── value-modeler-output.json      # Execution state + metadata

Data Model

Investment Theme

An investment theme groups 1-4 value chains into a distinct investment domain. Investment themes are the primary organizing unit of the value model — they represent CxO-level decisions about where to invest.

{
  "investment_theme_id": "it-001",
  "name": "Health & Nutrition Transformation",
  "strategic_question": "How do we reformulate our portfolio for the health-conscious, GLP-1-era consumer?",
  "executive_sponsor_type": "CPO / Head of Product Development",
  "narrative": "GLP-1 medications and functional food demand are fundamentally reshaping what consumers want. This theme covers reformulation, personalization, and nutritional innovation.",
  "value_chains": ["vc-001", "vc-002"],
  "solution_templates": ["st-001", "st-002", "st-003"],
  "business_relevance_avg": null,
  "ranking_value": null
}

Investment themes must satisfy MECE:

• Mutually Exclusive: Each investment theme answers a different strategic question. Different executive sponsor, different budget line.
• Collectively Exhaustive: Together, investment themes cover ≥80% of linked candidates.

Target: 5 investment themes (ideal). Range: 3 (minimum) to 7 (maximum, Miller's law).

Investment-theme-level Business Relevance is the average ranking_value of the investment theme's Solution Templates (calculated after Phase 4). This represents the investment theme's overall importance to the customer as expressed through its solutions' scores. Use this for investment theme ranking.

Value Chain (TIPS Path)

A value chain connects candidates across dimensions into a coherent causal narrative. Value chains are nested under their parent Investment Theme.

{
  "chain_id": "vc-001",
  "name": "GLP-1 Portfolio Reformulation",
  "investment_theme_ref": "it-001",
  "narrative": "GLP-1 medications reshape consumption (T), requiring AI-driven personalization (I), enabling health-optimized portfolio reformulation (P)",
  "trend": {
    "candidate_ref": "externe-effekte/act/1",
    "name": "GLP-1 Market Impact",
    "business_relevance": null
  },
  "implications": [
    {
      "candidate_ref": "digitale-wertetreiber/act/29",
      "name": "Personalized Digital Experiences",
      "business_relevance": null
    }
  ],
  "possibilities": [
    {
      "candidate_ref": "neue-horizonte/act/14",
      "name": "GLP-1 Portfolio Reformulation",
      "business_relevance": null
    }
  ],
  "foundation_requirements": [
    {
      "candidate_ref": "digitales-fundament/act/41",
      "name": "AI/ML Engineer Demand",
      "relationship": "prerequisite"
    }
  ],
  "solution_templates": ["st-001", "st-002"]
}

A single candidate may appear in multiple chains — the same Trend can drive different Implications depending on context. The chain captures the reasoning, not just grouping.

Solution Template

{
  "st_id": "st-001",
  "name": "Predictive Quality Analytics Platform",
  "description": "Deploy ML-based quality prediction integrated with production line sensors",
  "category": "software",
  "enabler_type": "process_improvement",
  "generation_mode": "portfolio-anchored",
  "investment_theme_ref": "it-003",
  "linked_chains": ["vc-005", "vc-006"],
  "solution_blueprint": {
    "building_blocks": [
      { "role": "lead", "capability": "Predictive analytics engine", "taxonomy_ref": "6.6", "taxonomy_name": "AI, Data & Analytics", "taxonomy_dimension": 6, "coverage": "covered", "feature_slug": "predictive-analytics", "product_slug": "cloud-platform", "delivers": ["ML model training", "anomaly detection"], "gaps": ["edge inference"] },
      { "role": "supporting", "capability": "IoT sensor connectivity", "taxonomy_ref": "1.4", "taxonomy_name": "5G & IoT Connectivity", "taxonomy_dimension": 1, "coverage": "partial", "feature_slug": "iot-gateway", "product_slug": "connectivity-suite", "delivers": ["sensor data collection"], "gaps": ["private 5G"] },
      { "role": "enabling", "capability": "Implementation consulting", "taxonomy_ref": "7.2", "taxonomy_name": "Digital Transformation", "taxonomy_dimension": 7, "coverage": "gap", "feature_slug": null, "product_slug": null, "delivers": [], "gaps": ["domain consulting"] }
    ],
    "readiness": { "covered_count": 1, "partial_count": 1, "gap_count": 1, "unknown_count": 0, "readiness_score": 0.64, "taxonomy_span": [1, 6, 7], "taxonomy_depth": 3 }
  },
  "portfolio_mapping": {
    "product_slug": "cloud-platform",
    "feature_slug": "predictive-analytics",
    "match_confidence": "high"
  },
  "portfolio_anchor": {
    "feature_slug": "predictive-analytics",
    "product_slug": "cloud-platform",
    "investment_theme_needs_delivered": ["ML model training", "anomaly detection"],
    "investment_theme_needs_undelivered": ["edge inference"]
  },
  "quality_flag": null,
  "business_relevance": null,
  "business_relevance_calculated": null,
  "ranking_value": null
}

Solution Process Improvement (SPI)

{
  "spi_id": "spi-001",
  "name": "Establish data governance policy",
  "description": "Define data ownership, quality standards, and access controls for production sensor data",
  "st_ref": "st-001",
  "change_type": "governance"
}

change_type values: governance | training | workflow | organization | measurement

Metric

{
  "metric_id": "met-001",
  "name": "Defect rate reduction",
  "unit": "percentage",
  "direction": "decrease",
  "investment_theme_ref": "it-003",
  "linked_chains": ["vc-005", "vc-006"]
}

direction values: increase | decrease

Collateral

{
  "collateral_id": "col-001",
  "name": "Predictive Maintenance ROI Case Study",
  "type": "case-study",
  "st_ref": "st-001",
  "status": "recommended"
}

type values: case-study | whitepaper | reference-architecture | demo | benchmark status values: exists | recommended

portfolio_mapping is only populated when a cogni-portfolio project is discovered. generation_mode: "portfolio-anchored" when generated from Step 0.5 (feature-first), "abstract" when generated from Step 1 (theme-first). Defaults to "abstract". solution_blueprint captures the multi-dimensional portfolio composition needed to deliver this ST — building blocks mapped to B2B ICT taxonomy categories with coverage assessment. Every ST gets a blueprint (both anchored and abstract). See references/data-model.md for the full SolutionBlueprint and BuildingBlock schemas. portfolio_anchor is derived from the blueprint's lead building block for backward compatibility. It records the primary feature and what it can/cannot deliver. quality_flag: "quality_investment_needed" when v3.0 quality assessment shows a fail on market_specificity or differentiation for a matched proposition. null otherwise. business_relevance is the user override (if set). business_relevance_calculated is computed via formula F1.

Business Relevance Scale

Score	Meaning
1	Secondary process, very little impact on customer activities
2	May bring some limited value in individual business domains
3	Significant benefits in some customer activities, not cross-domain critical
4	Impacts multiple business areas, substantial benefits expected
5	Mission critical, possibility to massively impact company KPIs

Ranking Formula: Enhanced F1

The patent's original F1 is a simple average of Business Relevance scores across linked TIPs. In practice, simple averaging flattens differentiation — a chain with T=5,I=4,P=2 scores the same as T=4,I=4,P=3. Cross-cutting solutions serving multiple chains get pulled toward the mean instead of being rewarded for breadth.

The enhanced formula addresses this with two adjustments:

Step 1: Per-chain score (F1 base)

ChainScore(c) = (sum(BR_T_j) + sum(BR_I_n) + sum(BR_P_k)) / (j + n + k)

Step 2: Peak-weighted aggregation across chains

BR(ST_i) = 0.6 × max(ChainScore) + 0.4 × avg(ChainScore)

For single-chain STs, this equals the original F1. For multi-chain STs, it rewards breadth while still anchoring on the strongest chain. The peak-weighting prevents cross-cutting solutions from being penalized by averaging in weaker chains — the best chain dominates, with breadth as a bonus.

Step 3: Foundation and blueprint readiness adjustment

FinalScore(ST_i) = BR(ST_i) × FoundationFactor × BlueprintFactor

Where FoundationFactor:

• 1.0 if 0-1 foundation dependencies
• 0.95 if 2-3 foundation dependencies
• 0.90 if 4+ foundation dependencies

Where BlueprintFactor (from solution_blueprint.readiness.readiness_score):

• 1.00 if readiness >= 0.8 (well-covered portfolio)
• 0.95 if readiness >= 0.5 (partial coverage)
• 0.90 if readiness >= 0.3 (significant gaps)
• 0.85 if readiness < 0.3 (mostly gaps)
• 1.00 if no blueprint (no penalty for legacy STs)

These mild penalties surface two distinct risk dimensions without dominating the ranking:

• FoundationFactor: prerequisite complexity (how much must be in place first)
• BlueprintFactor: portfolio readiness (can you deliver this with your current portfolio?)

A solution with BR 4.0, moderate dependencies, and partial portfolio coverage scores 4.0 × 0.95 × 0.95 = 3.61 — still high-priority, but both risks are visible.

Fields on each ST:

• chain_scores: array of per-chain F1 scores
• business_relevance_calculated: result of Step 2
• foundation_factor: the foundation readiness multiplier
• blueprint_factor: the portfolio readiness multiplier
• ranking_value: final score after Step 3 (or user override if set)
• business_relevance: user override (bypasses all calculation)

Integration with cogni-portfolio

When a portfolio project is discovered in the workspace:

1. Feature matching — Solution Templates are mapped to existing features by semantic similarity
2. Proposition linking — If a feature+market proposition exists, the ST inherits its IS/DOES/MEANS messaging
3. Solution enrichment — If a portfolio solution exists, the ST inherits implementation phases and pricing
4. Gap identification — STs that don't map to any existing feature represent portfolio gaps worth exploring

This creates a bidirectional bridge: trends inform which features matter most, and existing portfolio data enriches the solution templates with real commercial context.