See which number changed, where it came from, and what decision it supports.
Experts define the supported sources, mappings, refresh method, and decision rule. Each view keeps source, freshness, and gaps visible—so your team can act on evidence instead of a blended number it cannot explain.
The Analytics Contract
One decision view. Every number keeps its receipt.
01
Use configured inputs
A view can use approved manual entries, exports, or customer-specific feeds only when the engagement supports the source and delivery path.
02
Keep the receipt
Source, period, units, as-of time, generated time, and mapping status stay attached. A transformed record is not automatically an accurate business fact.
03
Expose the gap
Stale, missing, partial, rejected, and unmatched records remain visible. The decision view does not silently turn an incomplete input into success.
Signal To Decision
Many noisy inputs. One inspectable decision.
The workflow makes incompatibility and uncertainty visible before a number reaches the person making the call.
Illustrative managed flow. Inputs, cadence, mappings, calculations, and actions depend on the configured engagement.
01
Name the question
Start with the decision: what changed, which source disagrees, or what needs review?
02
Receive the records
Approved manual entries, exports, or configured feeds arrive through the path defined for that source.
03
Validate the receipt
Definitions, units, period, timestamp, and required fields determine what can continue and what needs review.
04
Map without hiding
Compatible records enter the agreed schema. Partial, stale, failed, and unmatched records branch into an exception state.
05
Build one decision view
Only compatible evidence is compared. Observed values, derived metrics, and unknowns stay visibly distinct.
06
Decide with context
A named owner reviews the evidence and chooses the next step. Any later outcome requires its own source trail.
Source Ledger
If the source is unknown, the metric is not ready.
A clean chart can still hide a stale export, mismatched unit, missing location, or failed transform. The ledger keeps those conditions in the decision.
- Source
- The named system, approved export, manual record, or configured feed.
- Freshness
- The source period, as-of time, delivery cadence, and report generation time.
- State
- Accepted, transformed, partial, stale, failed, rejected, missing, or unmatched.
- Owner
- The person responsible for access, definition, exception review, and the decision.
Exception, Not Success
A partial source should create a review task—not a confident green number.
Illustrative State
REVIEW REQUIRED / SOURCE STALE
The decision owner sees which input failed, when it was last valid, and which comparison remains unavailable.
Evidence Ladder
Observed, normalized, derived, and verified are different claims.
Observed
A source record exists for a named property and period.
Normalized
The record passed the agreed field, unit, identity, and period mapping.
Derived
A disclosed calculation uses eligible records and a named method.
Verified outcome
A downstream first-party record supports the claimed business result.
Access And Handoff
Record who controls the source—and what the agreement assigns.
The engagement records source-account control, access, copied data, refresh method, retention, and export terms. It separately identifies assigned customer-specific dashboards, mappings, integration code, reports, source files, and documentation. Provider history, licensed inputs, third-party services, hosting, and Connected Spaces shared code remain separate; continued function may require access, export, migration, or replacement.
Proof Requires Provenance
No fabricated dashboards, profitability, or automatic source success.
Any product capture or performance example requires permission or a clear demo label, redaction, named source, property, date range, time zone, filters, units, definitions, refresh method, and attribution context. Manual, provider-reported, derived, and first-party verified values stay distinct.
Know what the view can prove—and what still needs configuration.
Connected Spaces can use approved manual entries, exports, and customer-specific feeds when the source and delivery path are configured. Current platform analytics include integration-pipeline health, agent activity, task progress, configured marketing-project signals, and configured inventory item and low-stock counts. Availability depends on the modules, permissions, and records actually in use.
No. A connection is live only after supported access, configuration, and end-to-end verification. Current campaign provider sync methods are not wired as default live feeds. Until a source is configured, the engagement may use approved manual or imported records instead.
Normalization maps supplied fields, identities, units, and periods into an agreed schema. It does not make uncertain data true. Failed, partial, stale, rejected, and unmatched records remain visible for review.
Only when the required revenue, cost, budget, item-mapping, allocation, and period rules are configured and reconciled for that engagement. The current general analytics service does not provide a universal POS-to-cost profitability or always-current budget-versus-actual view.
Freshness is source-specific. Each decision view should identify the source period, as-of time, delivery method, and report generation time. A manual value is only as current as its last approved entry, and a failed or missing source is shown as a gap rather than an observed zero.
The engagement records which source accounts your business controls, who can grant or revoke access, what data is copied into Connected Spaces, and the export and retention terms. Customer-specific dashboards, mappings, integration code, reports, source files, and documentation transfer only as identified in the agreement. Provider history, licensed inputs, third-party services, hosting, and Connected Spaces shared code remain separate.
Continue The Decision