Skip to main content

Scoring Reference

Signal weights table used to score content relevance from 0 to 100.

SignalWeightDescription
Name match (word boundary)+30View/workbook/datasource name contains a search term at a word boundary (case-insensitive). "Sales" matches "Regional Sales Report" but NOT "Wholesale Tracker."
Name match (substring only)+15Name contains the term as a substring but not at word boundary. "sale" inside "Wholesale" scores here, not at +30.
Name match (synonym)+15Name contains a synonym or related concept from the synonym table.
Field match (exact)+25Datasource fields include exact terms from the query (e.g., "Region", "Sales")
Field match (implied)+15Fields match implied measures/dimensions (e.g., query mentions "revenue" and field is "Sales Amount")
Description match+20Content description references the topic
Tag match+10Tags contain relevant terms
Pulse metric match+20Metric name/definition matches the query concept
Temporal capability+10Has date fields when user's query implies time analysis
Usage signal (top quartile)+5Usage count (hitsTotal, or usage.totalViewCount if that is what the tool returns) is in the top 25% of the items in this search's candidate set for the same project. If no usage count is returned, skip this signal.
Recency signal+5Updated within the last 90 days

Rules:

  • The name signal is awarded once per item, using the best match across all query terms.
  • A field signal is awarded once per signal row. List every matched field as evidence; points are not multiplied by field count.
  • Name (word boundary, +30) plus exact field (+25) is 55, which is Medium. High (60+) needs at least one more signal.
  • Views-only searches skip field signals; High is still reachable through name, description, tag, usage, and recency (maximum without fields is 70).