docs: mark B2b Geo-Experiment tutorial as done in ROADMAP

[igerber]

Summary

PR #289 shipped Tutorial 18 (18_geo_experiments.ipynb) but ROADMAP.md still listed B2b as "In progress" with a description mentioning the GeoLift/CausalImpact comparison. That comparison was explicitly dropped from Tutorial 18 in commit 96d2eea and remains scoped under the separate B2c row ("diff-diff vs GeoLift vs CausalImpact comparison page").

• Flip B2b status to "Done (Tutorial 18)"
• Drop the "comparison with GeoLift/CausalImpact" mention from B2b's description (B2c still tracks that work independently)

Methodology references (required if estimator / math changes)

• Method name(s): N/A - no methodology or estimator changes
• Paper / source link(s): N/A
• Any intentional deviations from the source (and why): None

Validation

• Tests added/updated: No test changes (single-line roadmap status update)
• Backtest / simulation / notebook evidence (if applicable): N/A

Security / privacy

• Confirm no secrets/PII in this PR: Yes

---

Generated with Claude Code

[igerber]

/ai-review

[github-actions]

🔁 AI review rerun (requested by @igerber)

Head SHA: 5460c58e2e6b407155ec4128a0aa3b468fcb8662

---

Overall Assessment

Looks good (highest unmitigated severity: P3).

Executive Summary

• No methodology-affecting code changed; the diff is a roadmap/status update only in ROADMAP.md:L90-L91.
• The new B2b wording matches the shipped Tutorial 18 materials, which describe a SyntheticDiD geo-experiment walkthrough rather than a dedicated GeoLift/CausalImpact comparison page: docs/tutorials/18_geo_experiments.ipynb:L8-L10, docs/tutorials/README.md:L73-L79, docs/practitioner_getting_started.rst:L343-L344.
• No code quality, performance, security, or tech-debt concerns are introduced by this one-line docs change.
• One minor documentation consistency issue remains: docs/business-strategy.md:L237-L240 still frames the geo-experiment tutorial itself as the GeoLift/CausalImpact comparison.

Methodology

No findings. The PR only updates roadmap text in ROADMAP.md:L90-L91. SyntheticDiD methodology remains documented separately in docs/methodology/REGISTRY.md and referenced from docs/tutorials/18_geo_experiments.ipynb:L44-L48; no estimator, weighting, variance/SE, identification assumption, or default behavior changed.

Code Quality

No findings. The change is limited to roadmap copy and status.

Performance

No findings. No executable code path changed.

Maintainability

No findings in the changed file. The revised B2b row is clearer because it separates the shipped tutorial from the still-pending comparison page in ROADMAP.md:L90-L91.

Tech Debt

No findings. Nothing here needs TODO tracking.

Security

No findings. This is a documentation-only change with no secrets or executable-surface impact.

Documentation/Tests

• Severity: P3. Impact: After ROADMAP.md:L90-L91 makes the B2b tutorial a completed SyntheticDiD walkthrough and leaves the GeoLift/CausalImpact comparison as separate B2c work, docs/business-strategy.md:L237-L240 still says the tutorial’s unique value is a direct GeoLift/CausalImpact comparison. That leaves planning docs inconsistent about scope. Concrete fix: Update docs/business-strategy.md:L237-L240 to describe the geo-experiment tutorial as the SDiD walkthrough, and keep comparison language only on the dedicated comparison-page item.