01

Start with the decision, not the dashboard

Every metric should change an operating decision. Search impressions can reveal where a property is visible but unconvincing. Visits can show whether the promise earns attention. A completed inquiry can signal useful intent. A partner-confirmed outcome can justify deeper investment.

Problems begin when these stages are collapsed. A click is not a lead. A form submission is not necessarily qualified. A routed inquiry is not revenue. Treating them as equivalents may create a stronger presentation, but it weakens the operating system.

02

The five-stage evidence chain

We use a simple evidence chain that can be adapted to each platform while preserving shared definitions. The purpose is not to make every product identical. It is to make group reporting reconcilable.

  • Discoverability: search impressions reported for queries and indexed pages relevant to the product.
  • Engagement: visits that reach a useful page or interact with a decision tool.
  • Useful action: a meaningful step such as comparing a provider, completing a calculator or starting an RFQ.
  • Qualified intent: a complete, non-spam request that fits the platform's disclosed scope.
  • Commercial outcome: a partner-confirmed transaction or revenue event with an agreed attribution rule.
The farther a measure moves from the website, the stronger its reconciliation and attribution controls must become.
03

Definitions must travel with the number

A metric without a date range, source and definition is not reusable evidence. Monthly reports should preserve all three. If the lead definition changes—for example, by excluding incomplete submissions or adding phone calls—the comparison period should be marked rather than silently restated.

Group totals also need deduplication rules. The same person may use more than one platform, return on multiple devices or submit more than one request. We do not present a group audience number as a count of unique Canadians unless identity resolution supports that statement.

04

Measure the quality of the handoff

Focused platforms create value at the boundary between information and action. That makes handoff quality a first-class metric. A directory should know whether a listing is current and whether the user can reach the provider. A marketplace should know whether the request contains enough information to quote. A seller workflow should explain what happens after the estimate.

Useful quality measures include response time, acceptance rate, invalid-request rate, completion rate and partner disposition. They should be collected with the minimum personal data required for the workflow.

05

A reporting standard that compounds

The long-term advantage is not a prettier dashboard. It is a reliable feedback loop between search demand, product behaviour and real outcomes. That loop reveals which categories deserve deeper data, which pages need clearer intent matching and where a human operating process is failing the product promise.

This is also the foundation for credible public communication. Publish rounded group figures when appropriate, keep underlying definitions internally, and never imply audit, causation or brand attribution that the evidence cannot support.

Sources

References and scope

This paper sets out Boost Commerce's operating measurement standard. It is a framework, not an audit opinion, and does not claim that every portfolio property currently implements every control at the same depth.

  1. Boost Commerce Research methodologyBoost Commerce Research
  2. What are impressions, position, and clicks?Google Search Console Help
  3. Performance report: about the dataGoogle Search Console Help
  4. Creating helpful, reliable, people-first contentGoogle Search Central