Campaign Analytics

Campaign performance analysis with multi-touch attribution, funnel analysis, and ROI calculation. Three Python CLI tools — standard library only, no external dependencies.

Typical Workflow

# Step 1 — Attribution: which channels drive conversions
python scripts/attribution_analyzer.py campaign_data.json --model time-decay

# Step 2 — Funnel: where prospects drop off
python scripts/funnel_analyzer.py funnel_data.json

# Step 3 — ROI: profitability and benchmarks
python scripts/campaign_roi_calculator.py campaign_data.json

All scripts support --format text (default) and --format json for integrations.

---

Scripts

1. attribution_analyzer.py

Five attribution models:

Model	Description	Best For
First-Touch	100% to first interaction	Brand awareness
Last-Touch	100% to last interaction	Direct response
Linear	Equal credit all touchpoints	Balanced evaluation
Time-Decay	More credit to recent	Short sales cycles
Position-Based	40/20/40 split	Full-funnel marketing

2. funnel_analyzer.py

Stage-to-stage conversion rates, bottleneck identification, segment comparison.

3. campaign_roi_calculator.py

ROI, ROAS, CPA, CPL, CAC, CTR, CVR — flags underperforming campaigns against industry benchmarks.

---

Input Requirements

All scripts accept a JSON file. Use python -m json.tool your_file.json to validate syntax.

For complete input schemas (journey, funnel, campaign), sample data, and validation rules, see references/attribution-models-guide.md.

---

Reference Guides

Guide	Location	Purpose
Attribution Models	references/attribution-models-guide.md	Models with formulas, selection criteria, input schemas
Campaign Benchmarks	references/campaign-metrics-benchmarks.md	Benchmarks by channel and vertical
Funnel Optimization	references/funnel-optimization-framework.md	Stage-by-stage strategies, bottlenecks

---

Best Practices

1. Use multiple attribution models — Compare 3+ to triangulate channel value
2. Match time-decay half-life to average sales cycle length
3. Segment funnels by channel, cohort, geography
4. Benchmark against your own history first — industry benchmarks provide context only
5. Include all costs — creative, tooling, labor alongside media spend

---

Limitations

• Descriptive metrics only (no statistical significance testing)
• Standard library only (not optimized for >100K journeys)
• Offline analysis (static JSON snapshots)
• Single-currency; simplified time-decay; no cross-device tracking

Related Skills

• analytics-tracking: For setting up tracking (not analyzing data)
• ab-test-setup: For designing experiments from analytics insights
• marketing-ops: For routing insights to execution