Metrics and methodology
Definitions, scope and limitations for AI answers, website Health, crawler activity and imported Ads results.
On this page
Overview
This reference explains what the displayed numbers and evidence types measure. Always identify the site or workspace, observation dates, successful coverage and applicable filters before comparing values. A missing observation is not a measured zero.
All numerical examples below are illustrative arithmetic, not customer outcomes. Counts have no percentage denominator unless one is explicitly stated.
AI Visibility: selected collection
The following rows use one selected site/project and collection. Each successful provider response or repetition counts as one response. Providers are pooled without separate weighting. Failed and unfinished runs are excluded from answer metrics. These shared rules apply to every row in this table.
| UI label or observation | Meaning / numerator or count | Denominator | Time window | Source evidence | Example | Limitation |
|---|---|---|---|---|---|---|
| Visibility | Successful target recommended or target mentioned responses | All successful responses | Selected collection | run status and target records | 3/4 = 75%, rounded to whole percent | Neither unique-question coverage nor population-wide visibility; with no successes, no usable percentage exists |
| Recommended | Successful responses classified as target recommended | None; count | Selected collection | target recommendation records | 2 responses = 2 | Classification is recorded evidence, not a provider approval contract |
| Mentioned | Successful responses naming target without a recommendation | None; count | Selected collection | target mention records; recommendation takes precedence | 1 mention-only response = 1 | Other views may use mention inclusively or count other subjects |
| Competitors | Distinct non-target subject strings plus configured competitor names with an applicable explicit domain citation | None; count | Selected collection | non-target mention subjects and configured competitor domain matches | Two collected names = 2 | Not guaranteed to be fully alias-normalized unique businesses; not share of voice |
| Sources cited | Distinct explicit-citation URL strings | None; count | Selected collection | explicit citation relationships | A, B, B, C = 3 | Exact URL strings, not unique domains or citation-bearing responses |
| From your site | Distinct explicit URL strings marked owned | None; subset count | Selected collection | relationship plus stored ownership | One owned URL among three cited URLs = 1 | A third-party page about you remains third-party |
Completion, history and prompt cells
| UI label or observation | Meaning / numerator | Denominator | Scope and time | Provider treatment | Incomplete/failed evidence | Source | Example | Limitation |
|---|---|---|---|---|---|---|---|---|
| Responses completed | Succeeded runs | Expected runs | One collection | Each planned response/repetition is a position | Failed/pending remain outside numerator; show their states separately | job counts | 4/6 successful | Does not identify both remaining runs as failed |
| Platform completion | Distinct providers with at least one success | None; count; compare separately with expected platforms | Latest Monitoring collection | A partial provider can count as completed at platform level | One success does not erase its other failures | loaded run provider/status | 3 providers with a success | Platform completion is less precise than response completion |
| Expected platforms | Distinct provider identifiers in loaded runs | None; count | Latest Monitoring collection | One per loaded provider ID | Missing run rows may omit planned providers | loaded run roster | Four provider IDs = 4 | Not an independent schedule roster |
| Prompt matrix result | Strongest recorded target outcome: recommended, then mentioned, then missed | None; categorical | Exact prompt text/provider in selected collection | Repetitions collapse to the strongest outcome | Verify underlying run status; unavailable/failed cannot establish absence | grouped run result | One mention and one recommendation show recommendation | Does not show the distribution across repetitions |
| Explicit citation | URL cited in a recorded answer | None unless a separately defined report counts it | Individual answer and collection date | Provider-specific normalized evidence | No usable answer cannot prove absence of citations | explicit relationship | Answer explicitly links URL A | Distinct URL count differs from occurrence count |
| Returned source | Source returned alongside an answer | None | Individual response and date | Depends on provider payload | Missing source metadata may be a coverage limit | returned relationship | Provider returns A without citing it in final answer | Not automatically an explicit citation |
| Searched source | Source recorded as search evidence | None | Individual response/search record and date | Depends on provider evidence | An unavailable search trace is not proof of no search | searched relationship | Search evidence contains B | Not proof of final-answer citation |
| Owned source | Source assigned to the site's ownership context | None | Source record and relevant collection | Ownership is separate from provider/relationship | Unknown ownership should remain unknown | ownership field | Explicit URL on approved owned domain | Ownership alone does not make a returned source an explicit citation |
| Competitor appearance | Applicable non-target mention/recommendation or explicit configured-domain citation | None unless report defines a count | Inspect question/provider/date; competitor view may use retained history | Keep provider context | Failures are not negative competitor observations | subject/domain evidence | A competitor named in one successful answer | Tracked configuration and detected candidates are different sets |
| Monitoring history mentions | Loaded runs with a mention or recommendation record of any subject | None; count | One listed historical collection | Loaded provider runs pooled | Mapper does not explicitly filter success; inspect underlying statuses | history mapping | A competitor-only answer may contribute | Broader than target-only Overview; reconcile label before equating |
| Monitoring history sources | Distinct URL strings across all loaded citation relationships | None; distinct count | One listed historical collection | Provider records pooled | Mapper does not explicitly filter success; coverage depends on loaded evidence | history mapping | Returned A and explicit A count as one URL; searched B adds another | Not Overview's successful-response, explicit-only URL metric |
Trend and manual reporting calculations
| UI label or observation | Meaning / numerator | Denominator | Scope and time | Provider treatment | Incomplete/failed evidence | Source | Example | Limitation |
|---|---|---|---|---|---|---|---|---|
| Visible percentage in a trend bucket | Sum of target recommended plus target mentioned responses | Sum of successful recommended, mentioned and missed responses | Selected project; UTC day buckets for 7/30 days, UTC months for 12 months | Pooled response counts, no equal-provider weighting | Jobs with no success absent; gap is not measured zero | trend counts and frontend buckets | 1/2 plus 9/10 = 10/12 = 83% | A point may include several collections |
| Not visible percentage | Successful target-missed responses | Same successful response denominator | Same trend scope/window | Same pooled counts | Failed runs excluded | Recorded evidence | 2/12 = 17% | Only the observed successful responses |
| Range percentage | All visible responses in range | All successful responses in range | Available history in selected range | Pooled successes | Retention and missing collections limit coverage | Recorded evidence | 10/12 = 83% | Selector does not guarantee that much history exists |
| Client-report owned-citation rate; manual field, not a claim about a PB card | Successful answers with at least one explicit owned citation, once per answer | Successful answers in the report's defined comparable subset | Manually specified site/questions/period | State included provider profiles | Disclose exclusions; zero denominator is unavailable | approved answer evidence plus report method | 2/10 = 20% | Different from the distinct owned-URL count |
| Educational share of voice; manual formula only | Whatever eligible brand appearances the report explicitly defines | Defined eligible appearances across its selected brand set | Explicit manual scope and period | Disclose weighting and provider set | State exclusions and coverage | reviewed evidence and written formula | 3 of 10 defined appearances = 30% | Do not label this a shipped PB metric without verifying the implemented formula |
Target matching in the Overview uses the configured brand/aliases. Historical trends use stored target records and exclude rejected records. Changes in configuration or review decisions can affect comparisons. Retain the method used for each report.
Search & AI Visits and portfolio activity
The inspected page loads a 30-day aggregate per available site, up to 100 recent requests and 250 page rows per site, plus a separately requested 365-day trend. These are different scopes. Site selection applies across those records; a chart's time selector is not a global time filter for every card and table.
With all platforms and no specific path filter (including the default slash value), cards use the aggregate. Platform or specific-path filtering switches relevant values to the bounded loaded request sample. Providers here are user-agent classifications of recorded requests, not AI answer providers and not authenticated proof of a crawler operator's identity.
| UI label or observation | Meaning / numerator | Denominator | Scope and time | Incomplete/failed evidence | Source | Example | Limitation |
|---|---|---|---|---|---|---|---|
| Visits / Requests | Recognized crawler activity log rows | None; count | Selected available sites; 30-day aggregate or filtered recent sample | Failed requests remain included; unavailable site fetches can be omitted | qualifying logs | 120 matching logs = 120 visits | Not human visits, unique users or AI answer appearances |
| Pages visited | Aggregate: distinct stored full URL strings per site, summed across sites; filtered sample: distinct site/path pairs | None; count | Same aggregate/sample scope | Missing coverage is not zero pages | URL/site/path | Three qualifying URLs = 3 in aggregate | Aggregate and sampled path definitions can differ |
| Cache hits card | Logs with HIT cache status | Aggregate: HIT, MISS, STALE or COALESCED logs; sampled UI: loaded rows with a displayed cache value | Same card scope; rounded percentage | Status filtering is not identical to successful delivery; empty denominator has no usable rate | Recorded evidence | 80 hits / 100 eligible aggregate cache responses = 80% | Sample denominator may include Bypass; reconcile before comparison |
| Fresh renders card | MISS logs with HTTP status 200–399 | None; count | Same card scope | Missing status and failed status excluded | Recorded evidence | Five successful MISS rows = 5 | Includes redirects; does not by itself establish expected page content or billable usage |
| Failed requests | Logs with null status, status below 200 or at least 400 | None; count | Same card scope | Null status counts as unsuccessful classification | Recorded evidence | Two 403s plus one unknown = 3 | Inspect cause; not all failures are crawler blocking |
| Portfolio hits/cache/last crawl | Per-site activity count, cache-hit ratio and most recent activity time | Cache ratio only: eligible cache responses | Loaded 30-day rollup, with recent-row fallback where used | No available data cannot establish inactivity | site augmentation | 80/100 = 80% cache | Do not equate it with a billing-cycle allowance |
| Chart Cache hits | HIT log counts | None; count | Selected sites; 7/30-day or 12-month view of loaded history | Missing site retrieval can reduce coverage | UTC-day source trend | Two HIT rows in bucket = 2 | A displayed zero in a filled bucket does not prove complete observation |
| Chart Fresh | Aggregate trend: all MISS logs, without card's successful-status condition; sampled fallback: non-Hit rows | None; count | Same chart view; filtered views can use recent sample | Aggregate includes failed MISS; fallback can include other cache states | Recorded evidence | One failed MISS may increase chart Fresh but not Fresh renders card | Known definition mismatch; not a substitute for fresh-render billing usage |
| Visiting platforms | User-agent-classified Google, Bing and AI request counts | None; count | Same chart window/site or filtered sample | Request failure does not imply no recorded request | Recorded evidence | Three classified request rows = 3 observations | Classification does not prove training, indexing, citation or downstream use |
| Delivered / Failed row | Delivered when HTTP status is 200–399; otherwise Failed | None; categorical | One request timestamp | Null code is unsuccessful; rendering provenance may remain unverified | status and delivery fields | 302 can display Delivered | Follow redirect and inspect delivery/content; Delivered is not verified semantic success |
| Fresh / Stale row | Successful non-STALE row is labeled Fresh; successful STALE is Stale | None; categorical | One request timestamp | Failed status displays Failed | cache/status mapping | A successful cache HIT may be Fresh | Fresh label does not prove the content reflects the latest CMS change |
| Rendered content access / Delivery | Stored rendered-content flag and delivery source | None; categorical | One request timestamp | Unknown flag remains unverified | provenance fields | Cached rendered content versus origin response | A response can succeed without establishing rendered-content delivery |
Health, Readiness and page monitoring
These observations concern recorded page evidence, not AI answer coverage. AI monitoring provider weighting is not applicable. Evidence can include different check sources, so a report timestamp does not establish that every external system was queried at that moment.
| UI label or observation | Meaning / numerator | Denominator | Scope and time | Incomplete/failed evidence | Source | Example | Limitation |
|---|---|---|---|---|---|---|---|
| Checks by category | Checks with Clear status in the category | Available checks in category | Selected site's latest report | Unavailable comparisons may be Warning; inspect details | report checks | 3/4 clear | Four checks, not every page |
| Platform access summary | Clear Access and Rendering checks | Available checks in those groups | Same Readiness report | Missing evidence is not positive verification | Recorded evidence | 3/5 clear | Not an indexing or traffic score |
| Content understanding summary | Clear Entities and Content checks | Available checks in those groups | Same Readiness report | Inspect unavailable evidence and source dates | Recorded evidence | 4/6 clear | Not measured human or AI comprehension |
| Critical / Warning | Checks with that severity | None; count | Same report | Warning can represent uncertainty | Recorded evidence | Two warnings = 2 | Triage classification, not probability of harm |
| Open issues | Non-Clear checks | None; count | Same report | Distinguish actionable problem from unavailable comparison | Recorded evidence | One critical plus two warnings = 3 | Do not claim every open check is a verified defect |
| Monitoring cadence / configured URLs | Schedule and configured page set, not a performance metric | None | Current configuration; actual run dates separate | A schedule is not proof each check ran successfully | configuration and run/incident evidence | Weekly schedule, two configured URLs | Scheduled monitoring is not continuous uptime coverage |
AI Ads and allowance terms
Ads results summarize imported provider-account performance within the workspace. The inspected implementation selects snapshots whose window end is at or after the requested rolling cutoff (default 30 days, allowed 1–90). Chart grouping uses the database day boundary; confirm the deployed timezone. Do not assume the beginning of a selected day, a site filter or exact agreement with a provider UI timezone.
| UI label or observation | Meaning / numerator | Denominator | Provider/scope and time | Incomplete/failed evidence | Source | Example | Limitation |
|---|---|---|---|---|---|---|---|
| Impressions | Sum of imported impression counts | None; count | Connected supported provider accounts, workspace/window above | No import is unknown coverage, even if displayed as 0 | performance snapshots | 100 + 200 = 300 | Aggregate ad delivery, not exact prompt or unique-user count |
| Clicks | Sum of imported click counts | None; count | Same results scope/window | Import failures leave incomplete evidence | Recorded evidence | 2 + 3 = 5 | Not conversions or organic referrals |
| Spend | Sum of imported spend micros; one currency unit is 1,000,000 micros | None; monetary total | Same results scope/window; preserve account currency | Currency is absent from inspected results payload; confirm display and account comparability | Recorded evidence | 2,500,000 micros = 2.5 units in the verified currency | Do not sum different currencies or invent conversion into dollars |
| Conversions field | Stored field is summed, but inspected importer supplies fixed zero rather than importing conversions | None; not a supported observed conversion metric in this flow | Same results scope/window | Treat as unavailable until measurement/import support is verified | Recorded evidence | A displayed 0 is not evidence of zero conversions | Do not calculate conversion rate, CPA or ROAS from this field |
| Creative/variation results | Imported totals grouped by linked creative ID, displayed with current version title; up to 50 groups | None; counts and spend above | Workspace/window; provider external ad linkage | Unlinked ads may remain; exact text/URL matching can be ambiguous | Recorded evidence | A linked ad contributes five clicks to creative C | Not immutable version attribution; not proof the displayed current title earned those clicks |
| Last imported | Latest imported-at value among selected snapshots | None; timestamp | Results scope/window | Null/no snapshots means no confirmed import in that scope | Recorded evidence | Import time after reporting window | Not the same as provider event time or connection last-synced time |
| Fresh render allowance | Effective account render consumption | Effective allowance for that usage period | Workspace/account period; not AI provider scope | Reconcile reservations, failures and diagnostics with usage rules | account usage and render contract | 50 used of an effective 500 = 10% of allowance | Crawler Fresh card is a different population; cache hit does not use fresh-render allowance |
| Tracked prompt allowance | Active monitored questions | Effective prompt capacity | Current account/site setup | A failed response does not remove the prompt slot | Recorded evidence | Two questions across three providers still use two question slots | Not a response or API-call count |
| Page-monitoring URL allowance | Configured monitored public URLs | Effective URL capacity | Current account configuration | Run completion is separate | Recorded evidence | Three configured URLs use three URL positions | Not prompt capacity |
| Generated content allowance | Recorded generation usage under effective account rules | Effective content-generation allowance | Applicable account usage period | Confirm failure/reservation treatment; saving is not publication | content/usage records | One generated draft is not one published article | Do not infer unused allowance from deleting a draft |
| Ads credits | PB operation credits consumed/reserved under current tariff | Effective credit allowance and period | PB workspace; separate from ad-provider spend | Completed Proof can consume credit for any result; operational exception releases its reservation in inspected flow | operation ledger and account display | Copy 1 + optional image 1; completed Proof 1 in inspected revision | Reverify tariff; opening saved local preview is not generation; no provider spend equivalence |
Read the account's effective limit and period before interpreting usage. A legacy subscription, a trial and a current plan can have different entitlements. Do not derive them from a public plan name alone.