Steady Dashboard

Demonstration data

These values test decision behavior. They are not Express Home Services results. Each contract names the future verification source.

KPI-CLI-001

Qualified leads

HealthyDemonstration data

36Target 40High confidence

Decision question: Is demand producing accepted opportunities?
Formula: Count of CRM leads accepted by qualification rules
Numerator: Accepted qualified lead records
Denominator: Not applicable
Source: CRM
Filters: Express Home Services · All tracked services
Inclusion rules: Passes approved qualification rules · Unique lead identity
Exclusion rules: Spam · Duplicates · Unserviceable requests
Window: Last 30 days
Comparison: Previous 30 days
Refresh: Daily
Freshness: 24 hours · observed 2026-07-27T12:30:00.000Z
Confidence rule: High only after source-to-CRM reconciliation passes.
Owner: Marketing Lead
Target: 40
Target rationale: Approved demand target based on service capacity.
Thresholds: Warning 34 · Critical 25
Required response: Review channel quality, routing, and intake rejection reasons.
Limitations: CRM integration is not connected in this pilot.
Last verified: 2026-07-27

KPI-CLI-002

HealthyDemonstration data

25Target 28Medium confidence

Decision question: Are qualified inquiries becoming scheduled work?
Formula: Count of appointments with confirmed date and service
Numerator: Appointments with confirmed date and service
Denominator: Not applicable
Source: CRM and scheduling system
Filters: Express Home Services
Inclusion rules: Confirmed appointment · Service type recorded
Exclusion rules: Cancelled before confirmation · Duplicate bookings
Window: Last 30 days
Comparison: Previous 30 days
Refresh: Daily
Freshness: 24 hours · observed 2026-07-27T12:00:00.000Z
Confidence rule: Medium until CRM and scheduler totals reconcile.
Owner: Client Success Lead
Target: 28
Target rationale: Target protects current technician capacity.
Thresholds: Warning 23 · Critical 18
Required response: Inspect response time, contact rate, and booking objections.
Limitations: Values are illustrative until scheduling is connected.
Last verified: 2026-07-27

KPI-CLI-003

StaleDemonstration data

11Target 5High confidence

Decision question: How quickly does the company respond to qualified demand?
Formula: Median minutes from accepted inquiry to first human response
Numerator: Sum is not used; median of response minutes
Denominator: Qualified business-hour inquiries with a valid response
Source: CRM activity and telephony
Filters: Business-hour inquiries
Inclusion rules: Qualified inquiry · Business-hour receipt · Human response timestamp
Exclusion rules: Spam · Automated acknowledgement · After-hours inquiries
Window: Last 30 days
Comparison: Previous 30 days
Refresh: Hourly
Freshness: 6 hours · observed 2026-07-25T12:00:00.000Z
Confidence rule: High when telephony and CRM timestamps reconcile above 98%.
Owner: CRM Lead
Target: 5
Target rationale: Five minutes is the approved response service level.
Thresholds: Warning 8 · Critical 15
Required response: Inspect routing, coverage, and missed-call recovery.
Limitations: Pilot timestamps are demonstration data.
Last verified: 2026-07-27

KPI-CLI-004

UnavailableUnavailable

UnavailableTarget 50000No confidence

Decision question: Which services create verified collected revenue?
Formula: Collected revenue grouped by completed service
Numerator: Collected revenue assigned to completed service records
Denominator: Not applicable
Source: Accounting system
Filters: Boiler · HVAC · Mini-split · Water heater
Inclusion rules: Collected payment · Completed service · Service classification present
Exclusion rules: Open invoices · Refunded payments · Unclassified deposits
Window: Last 30 days
Comparison: Previous 30 days
Refresh: Daily
Freshness: 24 hours · observed Unavailable
Confidence rule: High only after accounting reconciliation closes.
Owner: Finance Owner
Target: 50000
Target rationale: Illustrative target only; Finance must approve a production value.
Thresholds: Warning 40000 · Critical 30000
Required response: Reconcile accounting access and service attribution.
Limitations: Accounting is not connected.
Last verified: 2026-07-27

KPI-AGT-001

Low confidenceDemonstration data

88Target 85Low confidence

Decision question: Do agent outputs satisfy QA without revision?
Formula: Accepted first reviews divided by all completed reviews
Numerator: Agent outputs accepted on first independent QA review
Denominator: Agent outputs receiving a completed first QA review
Source: Quality review records
Filters: All pilot agents
Inclusion rules: Completed independent QA review
Exclusion rules: Cancelled runs · Tests without review
Window: Last 30 days
Comparison: Previous 30 days
Refresh: Daily
Freshness: 24 hours · observed 2026-07-27T12:00:00.000Z
Confidence rule: Low below 30 representative reviewed outputs.
Owner: Automation & AI Lead
Target: 85
Target rationale: Quality target balances usefulness with independent control.
Thresholds: Warning 75 · Critical 60
Required response: Sample defects by agent, task, source quality, and control.
Limitations: Small pilot sample creates low confidence.
Last verified: 2026-07-27

KPI-QA-001

HealthyDemonstration data

0Target 0High confidence

Decision question: Did an agent or workflow attempt an unauthorized action?
Formula: Count of confirmed permission-envelope violations
Numerator: Confirmed permission-envelope violations
Denominator: Not applicable
Source: Agent and workflow trace
Filters: All environments
Inclusion rules: Confirmed attempt outside approved permission envelope
Exclusion rules: Correctly refused requests · Authorized test simulations
Window: Last 30 days
Comparison: Previous 30 days
Refresh: Immediate
Freshness: 1 hours · observed 2026-07-27T12:45:00.000Z
Confidence rule: High only when every material runtime supplies complete traces.
Owner: Automation & AI Lead
Target: 0
Target rationale: Permission violations have zero tolerance.
Thresholds: Warning 0 · Critical 0
Required response: Pause affected execution, preserve trace, and investigate.
Limitations: Only connected runtimes can provide complete evidence.
Last verified: 2026-07-27