Sandbox Usage Tracking

Instrument a sandbox environment to capture granular usage events that feed engagement scoring, intervention triggers, and close-rate prediction models.

When to Use

After provisioning a sandbox via sandbox-environment-provision. This fundamental sets up the event taxonomy specific to sandbox evaluation — distinct from general product analytics because the audience is prospects, not customers, and the goal is deal progression, not retention.

Instructions

1. Define the sandbox event taxonomy

All sandbox events follow the naming convention sandbox_{action} with standard properties:

| Event | Fires When | Key Properties | |-------|-----------|----------------| | sandbox_provisioned | Environment is created | deal_id, industry, persona, sample_data_records | | sandbox_first_login | Prospect logs in for the first time | deal_id, hours_since_provisioned, login_method | | sandbox_session_started | Any login after first | deal_id, session_number, days_since_provisioned | | sandbox_feature_used | Prospect interacts with a product feature | deal_id, feature_name, feature_category, is_differentiator | | sandbox_workflow_completed | Prospect completes a multi-step workflow end-to-end | deal_id, workflow_name, steps_completed, duration_seconds | | sandbox_data_uploaded | Prospect uploads their own data | deal_id, data_type, record_count | | sandbox_error_encountered | An error occurs during usage | deal_id, error_type, feature_name, severity | | sandbox_help_requested | Prospect clicks help, opens docs, or contacts support | deal_id, help_type, context_feature | | sandbox_milestone_achieved | Prospect completes a success checklist item | deal_id, milestone_name, milestone_index, total_milestones | | sandbox_expired | Sandbox reaches expiry date | deal_id, total_sessions, total_features_used, milestones_completed |

2. Implement event capture

In your product frontend, wrap sandbox-specific tracking:

// Detect sandbox context
const isSandbox = window.location.hostname.includes('sandbox-') || currentWorkspace?.sandbox === true;

if (isSandbox) {
  // Enrich every event with sandbox context
  posthog.register({
    sandbox_id: currentWorkspace.sandboxId,
    deal_id: currentWorkspace.dealId,
    prospect_company: currentWorkspace.prospectCompany,
    industry: currentWorkspace.industry,
    days_since_provisioned: daysSince(currentWorkspace.provisionedAt)
  });
}

For feature usage tracking, instrument each product area:

function trackFeatureUsage(featureName, category, isDifferentiator) {
  posthog.capture('sandbox_feature_used', {
    feature_name: featureName,
    feature_category: category,
    is_differentiator: isDifferentiator,
    session_duration_so_far: getSessionDuration()
  });
}

3. Build the engagement score

Create a PostHog computed property or n8n workflow that calculates a sandbox engagement score:

| Signal | Points | Max | |--------|--------|-----| | First login within 24 hours of provisioning | 10 | 10 | | Each session after first | 5 | 25 | | Each unique feature used | 3 | 30 | | Each differentiator feature used | 5 | 15 | | Each workflow completed | 8 | 24 | | Own data uploaded | 15 | 15 | | Each milestone achieved | 10 | 30 | | Help requested (positive — shows engagement) | 2 | 6 | | Error encountered (negative) | -3 | -15 |

Score ranges: 0-30 Cold, 31-60 Warm, 61-90 Hot, 91+ Champion.

4. Configure real-time alerts

Set up PostHog actions or n8n triggers for sales-critical events:

• High intent: sandbox_data_uploaded or 3+ sandbox_workflow_completed in one session → Slack alert to deal owner: "Prospect is actively evaluating — consider reaching out."
• Stall signal: No sandbox_session_started for 3+ days after first login → Trigger re-engagement flow.
• Error risk: 2+ sandbox_error_encountered in one session → Alert support to investigate and deal owner to follow up.
• Milestone completion: All milestones achieved → Alert deal owner: "Prospect completed full evaluation — ready for proposal."

5. Build the sandbox analytics dashboard

Create a PostHog dashboard with:

• Sandbox provisioning rate (sandboxes created / deals at Connected stage)
• Time to first login (histogram)
• Session count distribution
• Feature usage heatmap (which features get used most)
• Milestone completion funnel
• Engagement score distribution across active sandboxes
• Sandbox-to-proposal conversion by engagement score tier

Error Handling

• Missing deal_id: If a sandbox session fires without a deal_id, log a warning and attempt to resolve via the sandbox_id → deal_id mapping in Attio.
• Duplicate events: Use PostHog's deduplication by including a unique event ID for milestone and workflow completion events.
• Clock skew: Use server-side timestamps for duration calculations, not client-side.

Output

A complete sandbox usage event stream in PostHog, an engagement scoring model, real-time sales alerts, and a sandbox analytics dashboard.