Early Google Search evidence from two owned Markets

Read property totals as the top of a stack

For the completed 2–29 September 2026 GSC window, agency1001.com reports 7 clicks and 869 impressions, while search1001.com reports 4 clicks and 704 impressions at property level. These totals establish early reported Search activity for two owned Markets. They do not tell us that all activity went to participant profiles, or that a reported click became a commercial lead.

1. State the property, search type and inclusive date window.
2. Show exact example rows without treating them as the whole property.
3. Keep suppressed or unmatched query data unassigned.

Two-property GSC evidence table with limits

Use rows without overstating coverage

A query-by-page row can support an exact attribution claim for that pair. The Lumos row reports 2 clicks and 21 impressions; it is a subset, not a reconciliation of the property total. Google may omit query rows for privacy or reporting thresholds. Compare page and query cuts as separate views, retain the export window, and leave the unassigned remainder unassigned.

To carry this decision forward, understand the profile discovery path; separate totals from attributed rows; build the measurement worksheet.

A two-property evidence reading

Agency1001 and Search1001 report different property totals and different public Market structures. Put the four property metrics in a table with the same completed period and search type, then keep profile-specific rows beneath it rather than adding them to a total again. PageOptimizer Pro and Scalenut each have page-level observations in Search1001, but those are page cuts, not query-by-page attribution. The distinction matters when an article names a particular participant. Early data can justify a measurement practice without proving that one Market design outperforms another.

When the same report is viewed by page and by query, totals may not reconcile by simple addition because dimension cuts and privacy suppression differ. Preserve the raw export and request parameters in the evidence package. In the public article, quote only values whose reporting level is explicit. A reader should be able to tell which numbers describe an entire property, which describe a page and which describe a specific query-page pair without reverse-engineering the table.

Property totals are not profile totals

GSC can anonymize queries and aggregate cuts can differ. Do not sum unlike cuts or attribute property-level clicks to a visible query without a matching query×page row.