Strategy View & Scenario Analysis

The Strategy view provides a visual map of how your metrics interconnect and a scenario analysis tool for stress-testing strategic decisions. It is accessed via the Performance / Strategy toggle at the top of the dashboard.

Tier requirement: The Strategy toggle is visible to all users, but the full Strategy view requires the Strategic tier. Professional and Free tier users see an upgrade prompt when they click the Strategy button.

Admin vs Viewer: Admins see the live interactive network map built from current data. Viewers see a frozen, read-only map captured at the time of the most recent published snapshot.

Layout Modes

The Strategy view offers two layout modes, selectable via a toggle in the top-left corner:

• Network -- A force-directed network graph showing metrics as nodes and their relationships as connecting edges. This is the default view and the most information-dense.
• Risk Matrix -- A severity-by-likelihood heatmap that plots risks on a traditional 2D risk matrix. Useful for a quick visual assessment of risk concentration.

Network Map

The force-directed network map is the centerpiece of the Strategy view. It visualizes every tracked metric in your dashboard -- OKRs, Financial & Performance Metrics, Risks, and KRIs -- as an interconnected graph.

How to read the map:

• Nodes represent individual metrics. Each node is colored by its RAG status (green, yellow, red, or gray for untracked).
• Node size reflects systemic importance -- larger nodes have more connections and greater influence on the overall system. A large red node is a high-priority concern.
• Edges (lines between nodes) represent relationships discovered by the network analysis pipeline. These include structural relationships (e.g., a KRI belongs to a risk) and AI-discovered cross-domain correlations (e.g., revenue growth linked to employee retention).

Interacting with the map (Admin only):

• Zoom and pan using scroll and drag to navigate the graph.
• Click a node to open the Node Detail Modal (see below).
• Hover over a node to highlight its direct connections.
• The map uses a physics simulation that settles into a stable layout. Closely connected metrics cluster together naturally.

Viewer experience: Viewers see the network map as it was at publish time. The layout, node positions, and all data are frozen from the published snapshot. Interaction is limited to viewing node details (read-only).

Side Panel

The side panel sits to the right of the network map and contains two tabs: Network Analysis and AI Analysis.

Network Analysis Tab

This tab provides algorithmic insights derived from the network graph.

Strategy & Clusters sub-tab:

The network analysis pipeline uses Louvain community detection to identify natural clusters of closely related metrics. Each cluster is given a descriptive strategic name generated by the AI (e.g., "Revenue Growth Engine" or "Operational Resilience Cluster").

• Each cluster card shows the cluster name, the number of metrics it contains, and the dominant RAG status.
• Click a cluster card to highlight its members on the network map.

Influencers & Vulnerabilities sub-tab:

Two ranked tables showing:

• Top Influencers -- Metrics with the highest systemic influence scores. These are the nodes that, if improved, would have the greatest positive ripple effect across the system. Changes to these metrics tend to cascade widely.
• Top Vulnerabilities -- Metrics with the highest exposure scores. These are the nodes most susceptible to negative cascade effects. If a connected metric deteriorates, these are the first to be impacted.

Each row shows the metric name, its RAG status dot, and its influence or vulnerability score.

AI Analysis Tab

This tab surfaces insights from the most recent Performance Intelligence Report. It requires a cached report to exist for the current published snapshot.

Summary sub-tab: The executive summary from the report, providing a high-level health assessment and key findings.

Priority Themes sub-tab: Parsed from Section 1 of the report, showing the major strategic themes identified by the AI across all dashboard domains.

Recommendations sub-tab: The prioritized recommendations from Section 7 of the report, with impact and urgency indicators.

If no cached report exists, these sub-tabs display a prompt to generate a Performance Report first.

Node Detail Modal

Click any node on the network map to open a detail modal for that metric. The modal shows:

• Metric identity -- Name, type (OKR Key Result, Metric, Risk, or KRI), and parent entity (e.g., the objective an OKR belongs to, or the risk category a risk belongs to).
• Current value and target -- The latest value and the configured target or threshold.
• RAG status -- The computed status with a colored indicator.
• Connections -- A list of directly connected metrics in the network graph, showing both structural and AI-discovered relationships.
• Owner -- The assigned owner, if applicable.

Admin view: Full interactive detail with the option to navigate to the metric's edit modal.

Viewer view: Read-only data from the published snapshot. Values, targets, and statuses reflect the point-in-time data captured at publish.

Scenario Analysis (What If)

Scenario Analysis lets you model hypothetical events and see how they would ripple across your organization's metrics. It produces a detailed Strategic Scenario Briefing that you can download as a branded PDF.

Access: Click the What If button in the Strategy view toolbar. The button appears for Strategic tier users when prerequisites are met.

Prerequisites

All three conditions must be met for the What If button to be enabled:

1. Strategic tier -- Scenario Analysis is a Strategic-tier feature.
2. Published snapshot less than 90 days old -- The analysis requires a recent published snapshot as its data foundation.
3. Cached Performance Report -- A Performance Intelligence Report must have been generated for the current snapshot, since the scenario analysis builds on those findings.

When any prerequisite is not met, the button is disabled and a tooltip explains which condition is missing.

Step 1: Category

Choose one of six scenario types:

Category	Use Case
Market Disruption	New competitor entry, technology shift, demand collapse
M&A	Acquisition target, merger scenario, or divestiture
Operational Change	Restructuring, new systems, process overhaul
Regulatory	New regulation, compliance requirement, policy change
Macroeconomic	Interest rate shifts, currency fluctuation, recession
Custom	Any scenario that does not fit the standard categories

Step 2: Context

A set of guided fields tailored to the selected category. Common fields include:

• Description -- A free-text description of the scenario (required).
• Affected departments -- Which parts of the organization would be impacted.
• Estimated cost range -- The expected financial impact range.
• Timeline -- When the scenario would take effect and over what period.
• Probability -- Your estimated likelihood of the scenario occurring.

The specific fields adapt based on the category selected in Step 1.

Step 3: Documents (Optional)

Upload up to 3 supporting documents for additional context. Supported formats are PDF, Word, CSV, and TXT formats, with a maximum size of 10MB per file. These documents are analyzed by the AI alongside your dashboard data to produce more grounded recommendations.

Step 4: Review

A summary card displays all your inputs -- category, description, parameters, and attached documents. Review the details and click Run Analysis to begin.

Analysis

When you run the analysis, RWRD AI interprets your scenario in the context of your published dashboard data. A Monte Carlo simulation propagates the estimated impacts across your metric network, and a strategic briefing is generated.

The report pipeline runs in four phases: content (eight detailed sections — Scenario Definition through Action Plan, generated with forward-only context threading), reviewing for accuracy (a single Editor-in-Chief pass cross-checks every numeric and citation claim against snapshot truth and applies targeted edits deterministically), composing executive summary (the cross-section synthesis is generated last so it actually reflects the detail), and finalising executive TL;DR (the one-page CEO briefing — 3 insights, 3 risks, 2 decisions, 1 critical trade-off — sits on page 1 of the PDF). Analysis typically takes 5-10 minutes. A progress indicator shows which phase is currently running.

Results

The results screen shows:

• KPI summary cards at the top: Severity rating, number of Metrics Affected, Average Impact percentage, and estimated Time Horizon.
• Download card with a button to save the full briefing as a PDF.

Download PDF

The downloadable PDF is a branded Strategic Scenario Briefing containing:

• A cover page with RWRD branding, scenario title, and generation date.
• Charts and data visualizations throughout.
• A one-page Executive TL;DR (Chapter 2) — 3 insights, 3 risks, 2 time-bound decisions with the trade-offs at stake, 1 critical trade-off — designed for board distribution.
• The Executive Summary (Chapter 3) — a 3-4 paragraph cross-section synthesis grounded in the detail beneath it.
• Eight content chapters: Scenario Definition, Research Findings, Quantitative Impact Analysis (with a What Breaks First failure-sequence paragraph), Risk Cascade Assessment (with a Structural Fragility paragraph naming the single most fragile part of the business), Opportunity Mapping (with Opportunity Validation preconditions for every opportunity), Strategic Response (with optional Competitive Implications when your profile lists ≥2 competitors and a target market), 90-Day Action Plan, and Strategic Recommendations (now structured as Decisions Required (Next 30 Days) + Early Warning Indicators).
• A methodology appendix explaining the analysis pipeline (interpretation, Monte Carlo simulation, content generation, Editor-in-Chief pass, summary synthesis).
• A glossary of terms.

Scenarios produced before April 23 2026 do not carry the Executive TL;DR or the Editor-in-Chief pass; they render in their original form for historical accuracy.

The file is named RWRD-Scenario-Briefing-{title}-{date}.pdf.

Previous Analyses

Completed scenario analyses are saved and displayed as pill-shaped buttons at the bottom of the wizard (Step 1). Click any pill to load that analysis instantly without re-running it.

• Hover over a pill to see a tooltip with the scenario title, creation date, and who ran it.
• Admin delete: Admins see a trash icon in the results view header when viewing a saved scenario. Click to open a confirmation dialog.

Previous analyses are available to all Strategic-tier users in the organization, not just the person who created them.

Research Intelligence

Research Intelligence is a Strategic-tier feature that runs automatically when an Admin publishes a dashboard snapshot. It does not require any manual interaction.

What it does: RWRD AI performs a web search and SEC Edgar analysis focused on your industry and competitors, producing structured research findings with citations.

Where findings appear:

• In the Sources & References appendix of your Performance Intelligence Report.
• As background context that informs Scenario Analysis recommendations.

You do not need to configure or trigger Research Intelligence separately -- it runs as part of the publish workflow when the Strategic tier is active.

Export for AI

The Export for AI button (right of the What If button on the Strategy view toolbar) packages a published snapshot, the cached executive report, and every scenario you have run against that snapshot into a single ZIP designed for paste-into-Claude / ChatGPT / other orchestrating AIs.

Tier and role: Strategic tier, admin only. The button is hidden for viewers, contributors, and non-Strategic tiers. The same prerequisites that gate the What If button apply.

Why it exists: RWRD is the intelligence layer between human strategy and AI execution. Export for AI lets you take that intelligence with you -- feed it to whichever AI tool your team prefers, with the full snapshot grounding so the AI does not have to guess at numbers, names, or relationships.

What's in the Bundle

Each export is a single ZIP named rwrd-export-{orgSlug}-{snapshotIdShort}-{YYYYMMDD}.zip containing:

• README.md -- overview, file map, and caveats
• llm-context.md -- single flat Markdown bundle for Claude-scale context windows (200K+)
• llm-context-summary.md -- trimmed version (~30 KB) for ChatGPT / GPT-4o
• snapshot/ -- manifest.json, entities.json, context.json, analysis.json, plus network.json and research.json for Strategic-tier snapshots
• reports/executive.json, reports/executive.md, and reports/executive.pdf (PDF only included if the report has been opened at least once in the app)
• scenarios/{scenarioId}/scenario.json -- one file per scenario run against the snapshot

The full schema is documented in docs/export-bundle-schema.md for integration partners.

How to Export

1. Open the Strategy view.
2. Click Export for AI (right of the What If button).
3. A confirmation popover appears explaining the bundle contents and reminding you that the ZIP persists on whichever machine receives it. Click Export to confirm.
4. The button shows an Exporting... spinner. The ZIP downloads automatically when ready.
5. A success toast confirms the download. If the cached executive PDF has not yet been rendered, the toast notes that.

Rate limit: 5 exports per hour per user. Size limit: 45 MB.

Using the Bundle with External AIs

• Large-context models (Claude 4.x): Open llm-context.md and paste it into a fresh chat. Then ask your questions -- the AI is grounded in the source material.
• Smaller-context models (ChatGPT, GPT-4o): Use llm-context-summary.md instead.
• Programmatic use: Every JSON file is machine-readable. Start with snapshot/manifest.json.

User-supplied text in the bundle is sanitised (control characters stripped, Markdown fence-breakers neutralised). The Markdown header includes an explicit "data, not instructions" directive to the consuming AI.

Security & Audit Trail

Every successful export writes a row to your organization's audit log (action: data_export, target snapshot:{snapshotId}). The bundle is delivered via a 1-hour signed URL -- after that hour the URL expires, but the ZIP itself persists on whichever machine downloaded it. Treat the file as sensitive once it leaves the platform.

Snapshot retraction triggers automatic cleanup: every export ZIP for the retracted snapshot is deleted from storage, and new exports cannot be produced for a retracted snapshot.

Pre-fix scenario flag: Scenarios created before 2026-04-21 are flagged in the bundle as computedAgainstLiveData: true. They were computed against live dashboard data rather than the frozen snapshot they reference, so treat their cascade figures as indicative rather than authoritative.