CSV Export Schema

This page documents the column names and meanings for each CSV export. Columns are stable within v1: new optional columns may be appended, but existing columns will not be renamed or reordered.

Every export filename follows the shape surfacedby-{workspace_slug}-{section}-{domain}-{YYYY-MM-DD}.csv so a downloaded file is self-describing when you belong to multiple workspaces.

Cited URLs

One row per URL cited in the active analytics window.

Column	Type	Meaning
`URL`	string	Full URL of the cited page.
`Page Type`	string	Page-type classifier (`article`, `product`, `category`, `other`).
`Current Citations`	integer	Citations in the active window.
`Lifetime Citations`	integer	Citations since the domain was first scanned, or `N/A`.
`Citation Rate`	string	Percentage of prompts that cited this URL, e.g. `12.5%`.
`Trend`	string	`up`, `down`, `flat`, or `N/A`.
`Platforms`	string	Pipe-separated list of platforms that cited the URL (for example `chatgpt
`Opportunity Score`	integer	Internal 0 to 100 score; higher means a bigger lift if the URL is improved.
`Cited Prompts`	integer	Number of tracked prompts where this URL was cited.
`Missed Prompts`	integer	Number of tracked prompts where a competitor URL was cited instead.
`Co-Mentioned Brands`	string	Pipe-separated `domain(count)` pairs.
`Health Status`	string	`green`, `yellow`, `red`, or `N/A`.
`Content Score`	integer	Content Checker score (0 to 100), or `N/A`.
`First Cited`	ISO 8601	First scan in which the URL was cited.
`Last Cited`	ISO 8601	Most recent scan in which the URL was cited.
`Sparkline Scope Coverage`	string	Per-day measurement scope breakdown for the row's trend, e.g. `FB:5 DM:2`.

The per-platform citation columns are derived from the platform registry. When a new platform ships (for example Grok or DeepSeek), it appears in the Platforms cell automatically; existing columns stay in place.

Tracked prompts

One row per (prompt, platform) tuple that was actually tested in this scan. Prompts the scanner did not ask, and platforms that did not run for a prompt, do not appear. The Engine State column disambiguates the rows that are present.

Column	Type	Meaning
`Prompt`	string	The prompt as it was sent to AI platforms.
`Country`	string	ISO 3166-1 alpha-2 country code the prompt was scoped to.
`Platform`	string	Display name of the AI platform (for example `ChatGPT`).
`Geo Confidence`	string	`strong`, `medium`, or `weak`.
`Engine State`	string	`cited` if the platform cited you, `not surfaced` if it did not.
`Position`	string	Citation Rank in the response, prefixed with `#`, or `N/A`.
`Citation URL`	string	URL the platform cited, or `N/A`.
`Engine Coverage %`	string	Share of asked platforms that cited the brand for this prompt.
`Status`	string	Tracked prompt status (`active`, `paused`, etc.).
`Priority Score`	integer	Risk and value-weighted score used to rank prompts.
`Priority Level`	string	`High`, `Medium`, or `Low`.
`Relevance`	string	`High`, `Medium`, `Low`, or `N/A`.
`Historical Citation Rate`	string	Trailing brand-coverage percentage for the prompt.
`Pinned`	string	`Yes` if you pinned the prompt to your core set.
`CPC (USD)`	string	Cost per click on this prompt, where known.
`ETV (USD)`	string	Estimated traffic value.
`Monthly Value (USD)`	string	Combined monthly value of the prompt.
`Sentiment Score`	integer	0 to 100, where higher is more positive.
`Sentiment Label`	string	`positive`, `neutral`, `negative`.
`Measurement Scope`	string	`Daily Monitoring` or `Full Benchmark`. See below.

Competitors

One row per competitor in your tracked set.

Column	Type	Meaning
`Competitor Domain`	string	Root domain of the competitor.
`Display Name`	string	Human-readable name, or the domain when no override is set.
`Market Position`	string	Market Position, prefixed with `#` (for example `#3`), or `N/A`.
`Market Position Change`	string	Position change versus the previous comparable scan, or `N/A`.
`Visibility (Position Strength)`	integer	0 to 100 strength score, or `N/A`.
`Trend`	string	`up`, `down`, `flat`, or `N/A`.
`AI Mentions`	integer	Total mentions across all platforms in the scan.
`Mentions Change`	string	Delta versus the previous comparable scan.
`SoV (%)`	float	Share of Voice across your tracked set.
`SoV Change`	string	Delta versus the previous comparable scan.
`Coverage (%)`	float	Brand Coverage for the competitor.
`{Platform} Citations`	integer	One column per platform in the registry.
`Source`	string	Discovery source label (for example `Auto-discovered`, `Manual`).
`Measurement Scope`	string	`Daily Monitoring` or `Full Benchmark`.

Measurement scope

Most analytics exports carry a Measurement Scope column. The two values that ship are:

Daily Monitoring: a focused daily check that picks up recent movement quickly.
Full Benchmark: the complete tracked prompt and platform set used for rank, official reports, and durable competitor comparisons.

The backend stores the raw internal scan_type (pulse, paid_initial, paid_scheduled, console_on_demand, free_audit) on every scan. pulse maps to Daily Monitoring; all other values map to Full Benchmark. The export emits the public scope label so a spreadsheet filter matches the dashboard chip.

Plans

One row per action item across every plan on a domain.

Column	Type	Meaning
`Title`	string	Action item title.
`Plan`	string	Plan title the item belongs to.
`Section`	string	Public section label (for example `Content`, `Technical`).
`Priority`	string	Public impact label.
`Effort`	string	Effort label.
`Effort Hours`	integer	Estimated effort in hours, when set.
`Status`	string	Public status label.
`Source`	string	Where the item came from (opportunity, manual, plan template).
`Source Scope`	string	`Daily Monitoring`, `Full Benchmark`, or empty for manual items.
`Owner`	string	Owner email.
`Country`	string	Country code from the originating evidence.
`Step Progress`	string	`{checked}/{total}` recommended-step progress.
`Due Date`	string	ISO date or empty.
`Notes`	string	Free-form notes.
`Started`	string	ISO 8601 timestamp or empty.
`Completed`	string	ISO 8601 timestamp or empty.
`Reason`	string	Reason for the item (used by Citation gap and Improve rank items).

AI response feed

One row per AI response captured for a domain.

Column	Type	Meaning
`Platform`	string	Display name of the AI platform.
`Query`	string	The prompt sent to the platform.
`Country`	string	ISO country code.
`Geo confidence`	string	`strong`, `medium`, `weak`.
`Cited`	string	`Yes` if the platform cited your domain.
`Position`	string	Citation Rank prefixed with `#`, or empty.
`Sentiment`	string	Sentiment label.
`Score`	integer	Sentiment score (0 to 100).
`Tone`	string	Sentiment tone descriptor.
`Hedging`	string	`Hedged` if the model hedged its answer.
`Answer Preview`	string	First 200 characters of the answer.
`Sources`	integer	Number of source URLs the platform attached.
`Top co-mention`	string	Top co-mentioned competitor domain.
`Competitors Mentioned`	string	Comma-separated list of competitor domains in the answer.
`Measurement Scope`	string	`Daily Monitoring` or `Full Benchmark`.
`Date`	ISO 8601	When the response was captured.

Encoding

All exports are UTF-8 with a BOM so Excel opens them correctly. Field values containing commas, quotes, or newlines are quoted per RFC 4180. Values that begin with =, +, -, @, tab, or carriage return are prefixed with a single quote to defuse CSV-formula injection in spreadsheet apps.