CASE.03 / Review / SYNTHETIC

Reconcile China campaign data before assigning credit

A synthetic China report campaign exceeds download targets but misses qualified-lead and follow-up goals. The case reconciles platform, CRM, agency, and execution evidence.

DATA DISCLOSURE

This article uses a fictional scenario and synthetic material to demonstrate a workflow. It does not represent a real client, company, project, or market result.

01 / DECISION BEFORE OUTPUT

Review a China industry-report campaign and decide whether the next cycle needs more traffic, a different conversion path, or repaired lead definitions and follow-up.

Decision to support

Which result is proven, where does the funnel break, what cannot be attributed, and what should the local team change first?

02 / INPUT FILE

Put source material and constraints on the table.

Original goal

900 downloads, 320 qualified leads, and 90% follow-up within 48 hours

Evidence

WeChat, REDnote, Baidu, landing page, CRM, event, agency, and sales records

Data gap

32% of leads lack a complete normalized source

Review rule

Facts, interpretations, hypotheses, and unknowns must stay separate

MINIMUM INPUT
  • Original China campaign brief
  • Platform and first-party exports
  • CRM definitions and lead states
  • Execution timeline and agency or team notes
STARTING PROMPTUse $campaign-retrospective

Review this China campaign across platform, landing-page, CRM, event, agency, and sales evidence. Normalize definitions first, state attribution limits, and return no more than five owned actions with measurement points.

03 / EXECUTION TRACE

Preserve the work trace, not only the answer.

  1. 01

    Restore the original contract

    The review locked the original download, qualified-lead, and follow-up definitions instead of promoting new engagement metrics after results were known.

  2. 02

    Normalize the evidence

    Platform exposure, first-party visits, downloads, deduplicated leads, qualification, and sales follow-up were kept as different metric families with named owners.

  3. 03

    Limit cross-platform attribution

    Because source fields were incomplete and several platforms were walled gardens, the review did not declare a winning channel. It identified tracking repair as a prerequisite for the next decision.

  4. 04

    Create owned next actions

    The final actions covered source normalization, form qualification, lead handoff cadence, and one controlled message test, each with an owner and measurement point.

04 / OUTPUT EXCERPT

The deliverable should make judgment visible.

DOWNLOAD

Top-line target exceeded

Downloads are a clear result but cannot stand in for qualified demand or revenue impact.

QUALIFIED

Lead target missed

The gap points to audience, form, qualification, or data-definition questions that need separate checks.

HANDOFF

Follow-up lagged

The local market and sales handoff weakened the value of front-end acquisition.

ATTRIBUTION

Channel credit is limited

Incomplete source mapping and platform boundaries prevent a defensible single-channel winner.

DELIVERED / COMPLETE PACKAGE
  • Original-goal scorecard
  • Normalized cross-platform evidence table
  • Attribution limits and evidence-backed findings
  • Owned next-action plan
05 / HUMAN CHECKPOINTS

These decisions do not belong to a template.

  • Align the qualified-lead definition with sales
  • Confirm deduplication and agency transformations
  • Secure owner agreement on deadlines and measurement
GOOD FIT
  • Cross-platform China campaign
  • Agency and first-party reports disagree
  • The team needs next actions rather than a victory deck
NOT FOR THIS
  • No original goals or raw records
  • A request for unsupported causal attribution
  • An external performance press release
FIELD NOTE / CORE TAKEAWAY

The retrospective creates value by showing which evidence is comparable, which attribution is impossible, and which operating change should happen before buying more traffic.

REPRODUCE THE METHOD

Reproduce the method with China Campaign Retrospective.

The case demonstrates a reusable operating structure. Bring your own source material, constraints, and decision criteria to a real task.