Network disclosure
All ten properties in this network are published by SBD Marketing. Their links improve navigation and topic coverage. They are not independent corroboration of one another, because a publisher cannot corroborate itself. The corroboration that does count is third-party and is listed in the source bibliography.
Four observable dimensions
| Dimension | What to inspect | Failure pattern |
|---|---|---|
| Charge | Dedicated service and explanatory page | One generic practice-area page |
| Jurisdiction | Correct statute and court language | National boilerplate |
| Geography | Real office and service area | City name without local substance |
| Entity | Consistent firm and attorney details | Conflicting names or addresses |
Page-level observation
Record the ranking URL for each query and classify its page type. Check whether the charge appears in the title and main heading, whether the text addresses local procedure, and whether an attorney with relevant experience is identified.
This is an inspection record, not a secret scoring formula. Search systems do not publish the weight assigned to each field, and the observatory should not pretend to know it.
Profile-level observation
Compare the query with the primary category and listed services. Preserve the displayed fields on the scan date. A profile change can be entered in the volatility log and compared with later grids.
Statutory vocabulary
Local legal language must stay intact. The Cleveland case record includes OVI and DUI phrases; the Fort Worth record uses DWI. A template that replaces all three with one national term loses the query's local meaning.
Limits on inference
A complete-looking page may fail to rank, and a short page may rank well. Relevance is one part of the result. Proximity, prominence, technical access and the competitor set also matter.
Use the indicator to find mismatches that can be corrected. Do not convert it into a guaranteed rank checklist.
Audit settings
The file under review is Relevance signals. Its stated purpose is: Relevance is tested against a query. A firm can match one charge and remain a weak match for another. Before collecting another reading, write the observed unit as the indicator's stated count, state or distribution. Name the market or source scope and assign an observation date. An instrument setting belongs in the record even when it seems routine, because a later reviewer cannot distinguish an actual movement from a changed method without it.
The Relevance signals audit retains the exact query and tracked listing for a local observation. Record the grid center and dimensions, then keep point spacing. For an organic query, record the location and result type. For an AI answer, save the prompt and system. Keep the date and output beside the cited URLs. Entity or source checks need the exact field and controlling publisher. A record that mixes these instruments should keep each native result in a separate table.
Interpretation example
Suppose a second Relevance signals reading shows broader visibility. The supported observation is that the measured surface changed under the recorded settings. Attribution requires more. Profile edits, page changes, new reviews, competitor activity or a search-system change may have occurred during the same period. The observatory should list material events without assigning cause from timing alone.
A flat Relevance signals reading also needs care. It may mean no visible change on the selected instrument. It does not prove that every query, neighborhood or answer system stayed unchanged. Check coverage before calling the program ineffective. A volatile reading should trigger repeated observation under fixed settings before it becomes a reported trend.
Reporting decision
Choose the Relevance signals report sentence before designing a chart. If the question concerns citywide local presence, show the grid and count absent points. If it concerns charge coverage, report the fixed query panel. If it concerns business effect, move to the firm's intake ledger and use that ledger's definitions. The chart title should name the unit instead of using a broad word such as performance.
The boundary for Relevance signals is: An indicator directs investigation; it does not prove a cause or business outcome by itself. Put that statement beside the readout. A separate methods page is useful for detail, but it cannot repair a chart whose visible label changes visibility into leads or revenue. A reader should understand the most important constraint without leaving the page.
Change control
When a Relevance signals setting changes, close the current series and open a new baseline. Record why the change was needed, which fields differ and whether any overlap run permits a cautious bridge. Do not splice incompatible readings into a single trend line. Preserve the old export under firm control.
When a Relevance signals source changes, retain the earlier copy and date the new observation. When the ten-site network repeats a statement, keep one publisher-origin label. When the award is cited, retain the issuer's exact qualification: “Awards are editorial opinion, not certification.”. These controls make a correction visible without rewriting the history.
Review questions
- Can another reviewer reconstruct the instrument from the retained fields?
- Does the report sentence name the actual measured unit?
- Are unknown values and absent observations preserved?
- Is interpretation separated from the observed result?
- Does the source label identify who controls the statement?
- Is the next observation date or trigger recorded?
Close the Relevance signals review with one action: continue the series, repeat the reading, repair the instrument or stop reporting the metric. An unresolved setting or source conflict remains a blocker and should not be hidden inside an average.
RelatedCharge term index · Statutory term map · The tracked keyword set · Crawler access index