When this scope is useful
Teams disagree about results, events are undocumented or campaign and sales counts cannot be reconciled.
Adding another dashboard before fixing definitions only displays the disagreement in a new place.
Questions that distinguish the team
- What precisely triggers each event?
- How are duplicates and missing records detected?
- Which discrepancies are expected from attribution?
Inputs before work starts
Business questions, event inventory, system access, sample records and sales-stage definitions.
Three ways to scope the work
These are buyer-side scope models, not an agency price list. Use the inputs, client responsibilities and exclusions below for each model.
01 / Diagnose
- Work
- Audit definitions, tracking and discrepancies.
- Handover
- Gap register tied to business questions.
02 / Deliver a defined project
- Work
- Implement an event map and a tested reporting pipeline.
- Handover
- Event specification, test evidence and ownership map.
03 / Manage and improve
- Work
- Monitor data quality and maintain a reconciliation log.
- Handover
- Reconciled report with known limitations.
Acceptance and shared responsibility
Sample events can be traced from action to report; counts have defined units and periods; discrepancies have an owner and explanation.
Client responsibilities
Provide access, decide business definitions and nominate system owners.
What needs separate agreement
Data recovery that systems never captured and perfect cross-device attribution cannot be promised.
Ask for a cost estimate based on the actual access, production volume, integrations and review cadence. A comparable price requires the same scope.
Take this into the brief
Service: Marketing analytics Inputs: Business questions, event inventory, system access, sample records and sales-stage definitions. Acceptance: Sample events can be traced from action to report; counts have defined units and periods; discrepancies have an owner and explanation. Questions: What precisely triggers each event? How are duplicates and missing records detected? Which discrepancies are expected from attribution?
Common buying questions
Can one dashboard be the single truth?
Only for explicitly defined data and rules; other systems may count differently.
Should raw personal data be sent as event parameters?
Design the event contract to avoid unnecessary personal data.
Evidence to examine
Evidence review
Tensar: AI-related search clicks are not AI referrals
The useful distinction is the denominator: clicks for a selected keyword set are not a direct count of visits from an AI answer. Preserve the original label when comparing vendors.
Evidence review
DataGuard: pipeline is not collected revenue
This is an example of a connected delivery model. The reported pipeline measures potential commercial value; it cannot be booked as revenue or treated as profit.
Sources and review date
Sources read on: 2026-09-11. A source review is not an audit of the agency’s systems.