Organization API: Org Staff Insight

Base URL: https://api.coritan.com/api/v1. Paths below are complete.

To try these requests in the browser, open the interactive Organization API reference.

GET /api/v1/orgs/{org_slug}/staff/activity

The org's audit log, newest first. Admins read all of it. Tier 3 and billing must name the customer, order or server they are looking at, or ask for their own actions, which is all Tiers 1 and 2 may see.

GET /api/v1/orgs/{org_slug}/staff/activity/actions

The action names that have occurred, for the filter, with how many times each occurred, most often first and by name when two are equal. At most 100 names are listed. Below Tier 3 the list holds only the actions the caller took, and it is current. Tier 3 and above read the whole brand's list, which is normally under a minute old and can be a few minutes old on the first request after a quiet spell.

GET /api/v1/orgs/{org_slug}/staff/activity/export.csv

The filtered audit log as a CSV, newest first. For org admins only.

Takes the filters of GET /staff/activity. Carries at most 50,000 entries, with times in UTC as ISO 8601 (2026-09-28T12:00:00Z). The columns are time, action, actor_type, actor_id, actor_name, resource_type, resource_id, details (JSON) and ip_address. A text that would open as a formula in a spreadsheet starts with an apostrophe.

GET /api/v1/orgs/{org_slug}/staff/activity/facets

Counts of the log's entries by area, for the chips above the list.

Takes the filters of GET /staff/activity and the same role rules: Tiers 1 and 2 count only their own actions, Tier 3 and billing must name a customer, order or server, and admins count the whole log. An area is the text of an action before its first dot (customer for customer.suspended), worded as the console words actions. Each area's prefix is the action filter that narrows the list to it (customer.). total is the number of entries the filters match, so it equals the sum of the area counts. With a filter the counts are current. With no filter they cover the whole log, are normally under a minute old, and can be a few minutes old on the first request after a quiet spell.

GET /api/v1/orgs/{org_slug}/staff/insights

Money, orders, customers and support over a period, by day, and set against another period.

Needs billing or admin. The period is the last days days (30 by default), or the days from from to to in the brand's own time zone, at most 366. compare sets it against previous (the same number of days just before it; a calendar month compares to the prior month), week, month or year (the same days earlier), custom (the days from compare_from to compare_to) or none. A bad word, a period over 366 days or a custom comparison without both days answers 422.

The answer keeps its daily series (revenue, orders, customers and support) and adds period (from, to, days and the bucket a chart groups by), compare (the other period's days, or null), totals and previous_totals. Money is never added across currencies: totals.currency is the one with the most payments, the money figures are in it, and totals.by_currency lists every currency. mrr is what is recurring now, so it is null for a period that ended before today and for previous_totals. New customers are accounts: placeholders for moved forum authors are left out.

GET /api/v1/orgs/{org_slug}/staff/insights/acquisition

Where the organization's visits, sign-ups, orders and revenue came from, over a period and against the one before it. Needs billing or admin.

The period is from to to in the brand's time zone, the last 30 days when left out, and at most 366 days. Every list holds at most limit rows, ranked by orders, then revenue, then visits. The totals always cover every row.

An order is a paid storefront order that is not free, counted in currency. Orders and revenue never mix currencies: currencies names every currency with an order in the period, and currency picks the one the answer is in. Revenue is the invoice total without tax, in that currency's minor units. A sign-up is a customer made through the storefront, so a customer the staff create is not one. A visit is a stored session with bots left out. Sessions are kept for 180 days, so visits_complete is false when the period starts earlier and the visit figures are partial.

The answer holds:

• totals: visits, signups, orders, new_customers (orders from a customer's first paid order), revenue_minor and conversion_rate_bp (orders over visits). previous holds the same figures for the earlier period, or null when compare is none. Every figure ending in _bp is basis points, so 123 is 1.23 percent.
• channels: each channel of the last touch with its visits, signups, orders, new_customers, revenue_minor and share_bp (its share of the orders). first_touch ranks the channels of the first touch, and matrix counts orders by first channel and last channel.
• sources, campaigns, landing_pages and devices: the same figures by the last touch's source and medium, its campaign, the page it landed on (the path, with no query and no language prefix) and the device.
• click_ids: how many orders carry an ad click ID, and for which ad network.
• heard_about: the answers to "How did you first hear about us?" with their share of the answered orders.
• return: the ad spend return report for the period in currency (rows, totals, truncated), or null when the organization has no ad spend in the period. truncated is true when the return report read only its first 50,000 orders.

A from after to, a period over 366 days, a compare other than previous or none, a currency that is not three letters and a limit outside 1 to 50 answer 422.

GET /api/v1/orgs/{org_slug}/staff/insights/export.csv

The daily series as one CSV: a column per measure, a row per day, for the same period as GET /staff/insights. Needs billing or admin.

GET /api/v1/orgs/{org_slug}/staff/insights/report

One figure split by one dimension over a period, against another period. Needs billing or admin.

The answer is { metric, dimension, period, compare, rows, total, previous_total, series }. A row is { key, label, value, previous, filter }: filter is the list behind the row, { path, query } for the console (/staff/orders with plan=Paper), so a row or a chart point opens what it counts. series is the figure over the period by its bucket (hours up to 2 days, days up to 92, weeks up to 220, then months), each point { key, x, label, value, previous, filter } (x repeats key, the bucket's start) with previous from the comparison period at the same place. dimension.bucket is the grain of a time split and null for a category split, whose series follows period.bucket. period, compare and the period words are those of GET /staff/insights; the period is the last 30 days when left out.

Money is per currency and never added across them: currency is the one with the most money, every money figure is in it, and currencies and a row's by_currency carry the rest. active_services is what is active now, so it has no previous and no series. A metric the dimension does not suit answers 422 naming the dimensions it takes. truncated is true when a median read only the newest 20,000 tickets.

The attribution dimensions (channel, first_channel, source, medium, campaign, landing, device, heard_about and coupon_campaign) split revenue, orders and new customers by where the buyer came from: channel, source, medium, campaign and landing read the last touch. A payment, order or account with no attribution record is the group unknown. visits and conversion_rate (paid orders over visits, in basis points, null when there were no visits) take time and the visit words. A row whose list has no page to open, such as every visit row or a landing row, has a filter with a null path.

GET /api/v1/orgs/{org_slug}/staff/insights/report/registry

What the report builder can count and split by.

metrics lists each figure with its format (money, count, duration in seconds or percent in basis points) and the dimensions it may be split by, so a page offers only pairs GET /staff/insights/report accepts. dimensions lists the splits: a time one gives a row per day, week or month, a category one gives the biggest groups. Needs billing or admin.