Analytics and audit

The analytics endpoints expose the same rollups that power the dashboard: an org-wide overview, deliverability posture, warmup progress, per-campaign performance (with daily and hourly breakdowns and side-by-side comparison), mailbox health, and account usage. The audit endpoint returns your organization's activity trail ("who did what, when, from where"). All of these are read-only.

Every analytics route shares one auth gate: Scope READ_ANALYTICS · Org permission view_analytics. The audit route uses the same org permission with a dedicated scope. See permissions for the scope reference and authentication for how to present credentials.

Dates are parsed as YYYY-MM-DD unless noted. Errors follow the standard {error, message, code, request_id} envelope documented in error codes.

Get dashboard analytics

GET /analytics/dashboard

Returns the main dashboard overview for the active organization: aggregate stats, recent activity, top campaigns, account health, and a daily trend series. Sent and engagement totals include email steps only; completed wait and action steps do not inflate them. Account-health buckets are mutually exclusive and use the most severe current connection, sync, or warmup-reputation state. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Response

{
  "period": "7d",
  "overall_stats": {
    "total_emails_sent": 1240,
    "total_opens": 612,
    "machine_opens": 88,
    "total_clicks": 143,
    "machine_clicks": 6,
    "total_replies": 57,
    "total_bounces": 9,
    "open_rate": 49.35,
    "click_rate": 11.53,
    "reply_rate": 4.6,
    "bounce_rate": 0.73,
    "active_campaigns": 4,
    "active_accounts": 12
  },
  "recent_activity": [
    {
      "type": "replied",
      "campaign_id": "b1f2c3d4-0000-0000-0000-000000000001",
      "campaign_name": "Q2 outbound",
      "contact_email": "[email protected]",
      "contact_id": "c0ffee00-0000-0000-0000-000000000002",
      "timestamp": "2026-06-11T14:02:11Z",
      "sender_id": "5e11d0c4-0000-0000-0000-000000000003",
      "sender_email": "[email protected]"
    },
    {
      "type": "opened",
      "campaign_id": "b1f2c3d4-0000-0000-0000-000000000001",
      "campaign_name": "Q2 outbound",
      "contact_email": "[email protected]",
      "contact_id": "c0ffee00-0000-0000-0000-000000000002",
      "timestamp": "2026-06-11T13:40:02Z",
      "origin": { "client": "Apple Mail", "client_type": "app", "device_type": "mobile", "os": "iOS", "country_code": "US", "city": "Austin" }
    }
  ],
  "top_campaigns": [
    {
      "campaign_id": "b1f2c3d4-0000-0000-0000-000000000001",
      "name": "Q2 outbound",
      "status": "active",
      "emails_sent": 820,
      "open_rate": 51.2,
      "click_rate": 12.1,
      "reply_rate": 5.0
    }
  ],
  "account_health": {
    "total_accounts": 12,
    "healthy_accounts": 10,
    "warning_accounts": 1,
    "error_accounts": 1
  },
  "daily_trend": [
    { "date": "2026-06-05", "sent": 160, "opens": 79, "clicks": 18, "replies": 7 }
  ]
}

An opened or clicked entry in recent_activity carries origin when the person's first open or click on that step was logged: the same client, device and location object a contact timeline event carries.

Every entry names the mailbox the step was sent from in sender_id and sender_email (its send-as alias when it has one). A reply credits that mailbox even when it landed in the shared reply inbox the sender's reply-to names.

Get deliverability dashboard

GET /analytics/deliverability

Returns the organization's deliverability posture for a time window: bounce, complaint, open, click, and reply counts and rates, suppression and dead-letter pressure, reply-intent breakdown, seed inbox-placement (overall and per recipient provider), warmup-derived placement per recipient mail host, an overall health band and a 0-100 composite score (both from the documented thresholds), a daily timeseries, and per-mailbox and per-campaign breakdowns. Requires an organization context. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Response

{
  "from": "2026-06-05T00:00:00Z",
  "to": "2026-06-12T00:00:00Z",
  "events_total": 1380,
  "bounce_count": 9,
  "complaint_count": 1,
  "unsubscribe_count": 4,
  "reply_count": 57,
  "open_count": 612,
  "click_count": 143,
  "suppressed_recipients": 21,
  "dlq_pending": 0,
  "intent_positive": 18,
  "intent_negative": 6,
  "intent_out_of_office": 11,
  "intent_question": 9,
  "intent_neutral": 13,
  "intent_automated": 22,
  "emails_sent": 1240,
  "bounce_rate": 0.73,
  "complaint_rate": 0.08,
  "open_rate": 49.35,
  "click_rate": 11.53,
  "reply_rate": 4.6,
  "spam_placement_rate": 6.5,
  "inbox_placement_rate": 93.5,
  "placement_samples": 40,
  "band": "healthy",
  "score": 91,
  "timeseries": [
    {
      "date": "2026-06-05",
      "sent": 160,
      "bounces": 1,
      "complaints": 0,
      "opens": 79,
      "clicks": 18,
      "replies": 7,
      "unsubscribes": 1
    }
  ],
  "by_mailbox": [
    {
      "email_account_id": "a0a1a2a3-0000-0000-0000-000000000003",
      "email": "[email protected]",
      "sent": 420,
      "bounces": 2,
      "complaints": 0,
      "bounce_rate": 0.48,
      "complaint_rate": 0.0,
      "band": "healthy"
    }
  ],
  "by_campaign": [
    {
      "campaign_id": "b1f2c3d4-0000-0000-0000-000000000001",
      "name": "Q2 outbound",
      "sent": 820,
      "bounces": 5,
      "complaints": 1,
      "bounce_rate": 0.61,
      "complaint_rate": 0.12,
      "band": "watch"
    }
  ],
  "by_provider": [
    {
      "provider": "gmail",
      "label": "Gmail",
      "samples": 24,
      "inbox": 21,
      "promotions": 1,
      "spam": 1,
      "other": 0,
      "missing": 1,
      "inbox_rate": 87.5,
      "spam_rate": 4.17
    }
  ],
  "warmup_placement": [
    {
      "provider": "microsoft365",
      "label": "Microsoft 365",
      "delivered": 132,
      "spam": 6,
      "inbox_rate": 95.45,
      "spam_rate": 4.55
    }
  ]
}

The intent_* counters and reply_count are two different things and do not add up to each other. reply_count counts reply deliverability events, which are the ones a provider or your own integration reports to record an event. The intent_* counters are the replies Warmbly classified as they arrived in a connected mailbox, one per reply, including the machine ones: a vacation notice lands in intent_out_of_office and an autoresponder, ticket acknowledgement or bounce notice in intent_automated. A workspace that reports no reply events sees reply_count at zero with intents counted normally.

spam_placement_rate and inbox_placement_rate are omitted when there are no seed samples in the window. Seed samples are the copies of the workspace's placement tests that got a verdict in the window: inbox, promotions, other (another Gmail tab), spam, or missing (never arrived); copies that failed to send or were cancelled are not samples. by_provider rolls the same samples up per seed's provider family, where provider is the family id (gmail, google_workspace, microsoft365, outlook, yahoo and so on) and label its display name, and the rates are percentages of samples; warmup_placement is the continuous warmup signal per recipient mail host, where provider is the host id (google_workspace, microsoft365, zoho, hostinger, other and so on, or the provider group google, microsoft, yahoo or other for an arrival whose host was not detected), label its display name, delivered counts verified warmup arrivals and spam the subset the recipient's provider filed into junk. Warmup recipients are mostly other workspaces' mailboxes, so no recipient address or domain is reported. score starts at 100 and subtracts saturating penalties for bounce rate (up to 40 points, maxed at 10%), complaint rate (up to 30 points, maxed at 0.30%), and spam placement (up to 40 points, maxed at 40%).

Get warmup analytics

GET /analytics/warmup

Returns warmup send and reply statistics over a date range, with a summary and per-day series. Optionally scoped to a single mailbox. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Results are scoped to the selected workspace, so every member with access sees the same mailbox history. Workspace-wide results combine all mailboxes into one row per date. average_daily uses active days (dates with a warmup statistics row) as its denominator. target_progress is total sends divided by total planned target volume across those active days and may exceed 100 when sends beat the plan.

total_received and each day's emails_received count verified warmup mail that arrived from pool partners, bucketed on the UTC day it was verified. A day with arrivals but no warmup plan is listed with zero sends and does not count towards days_active.

Response

{
  "email_account_id": "a0a1a2a3-0000-0000-0000-000000000003",
  "email": "",
  "date_range": {
    "from": "2026-06-01T00:00:00Z",
    "to": "2026-06-12T00:00:00Z"
  },
  "summary": {
    "total_sent": 210,
    "total_replied": 84,
    "total_received": 196,
    "average_daily": 17.5,
    "reply_rate": 40.0,
    "target_progress": 87.5,
    "days_active": 12
  },
  "daily_stats": [
    { "date": "2026-06-01", "emails_sent": 12, "emails_replied": 5, "emails_received": 11, "target_volume": 12 }
  ]
}

email_account_id is the zero UUID when no email_id filter is supplied.

Get warmup placement

GET /analytics/warmup/placement

Returns where warmup mail landed in partners' mailboxes over a date range: the primary inbox, a Gmail category tab (Promotions, Updates, Social or Forums), or spam. Counts are per UTC day, per recipient provider, and, on the workspace report, per mailbox. Requires an organization context. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Every figure is measured, not estimated: a delivery is counted when the recipient's own sync finds the warmup email and records the folder it arrived in. rescued counts spam placements the recipient's mailbox was told to move back to the inbox; the move itself is not confirmed back, so it is the rescues requested, not a verified count. unconfirmed counts completed sends more than 24 hours old that no recipient has reported seeing, bucketed on the day they were sent; a provider filter cannot split it, so it is only reported unfiltered.

rate is the headline deliverability figure the dashboard shows next to each mailbox: inbox_rate is inbox plus tabs over what was delivered in the trailing 7 UTC days, and stays null until min_sample (20) deliveries are in. scope is always major: the rate is taken over Google, Microsoft and Yahoo recipients only, the providers that filter on sender reputation. A mailbox whose warmup mail reached only other hosts has no rate (delivered 0, band none). Other hosts are always in summary, daily and providers, and other_delivered and other_inbox_rate give their share of the same window (0 and null with none). band is good at 90 or above, fair from 80, poor below 80, collecting below the sample floor and none with no deliveries. Each day's rolling_inbox_rate is the trailing 7-day figure at the same three providers ending on that day; the day's counts cover every host, so filter by provider to read one. Rates are percentages with two decimals.

Recipient providers are grouped as google (Gmail and Google Workspace), microsoft (Outlook.com and Microsoft 365), yahoo (Yahoo and AOL) and other; each group lists its hosts, where an empty host means the recipient's host was not detected yet.

A range longer than 366 days, from after to, a malformed date or an invalid email_id returns 400; an email_id outside the workspace returns 404.

Response

{
  "email_account_id": "a0a1a2a3-0000-0000-0000-000000000003",
  "date_range": { "from": "2026-06-01T00:00:00Z", "to": "2026-06-30T00:00:00Z" },
  "summary": {
    "sent": 820, "delivered": 791, "inbox": 702, "tabs": 41, "spam": 48,
    "rescued": 44, "unconfirmed": 12, "inbox_rate": 93.93, "spam_rate": 6.07
  },
  "rate": {
    "window_days": 7, "min_sample": 20, "scope": "major", "delivered": 196, "inbox": 180, "tabs": 9, "spam": 7,
    "inbox_rate": 96.43, "band": "good"
  },
  "daily": [
    {
      "date": "2026-06-30",
      "sent": 30, "delivered": 29, "inbox": 27, "tabs": 1, "spam": 1, "rescued": 1, "unconfirmed": 0,
      "inbox_rate": 96.55, "spam_rate": 3.45, "rolling_inbox_rate": 96.43,
      "groups": [
        { "group": "google", "inbox": 15, "tabs": 1, "spam": 0, "rescued": 0 },
        { "group": "microsoft", "inbox": 12, "tabs": 0, "spam": 1, "rescued": 1 }
      ]
    }
  ],
  "providers": [
    {
      "group": "google",
      "sent": 0, "delivered": 402, "inbox": 351, "tabs": 41, "spam": 10, "rescued": 9, "unconfirmed": 0,
      "inbox_rate": 97.51, "spam_rate": 2.49,
      "hosts": [
        { "host": "google_workspace", "sent": 0, "delivered": 310, "inbox": 271, "tabs": 33, "spam": 6, "rescued": 6, "unconfirmed": 0, "inbox_rate": 98.06, "spam_rate": 1.94 },
        { "host": "gmail", "sent": 0, "delivered": 92, "inbox": 80, "tabs": 8, "spam": 4, "rescued": 3, "unconfirmed": 0, "inbox_rate": 95.65, "spam_rate": 4.35 }
      ]
    }
  ]
}

The workspace report (no email_id) omits email_account_id and adds mailboxes, one row per mailbox with activity in the range, ordered worst rate.inbox_rate first, then mailboxes still collecting. Each row carries the window counts, its own rate, and daily_inbox_rate, one entry per day of the range (null on a day with no deliveries).

Placement history is kept for the life of the mailbox, like warmup volume history, and moves with a workspace export. Deliveries recorded before this report existed, and any arrival the live count missed, are counted by a background pass from the receipts still on file within minutes; those read as inbox or spam only, with category tabs and rescues as 0.

Get campaign analytics

GET /analytics/campaigns/:id

Returns a single campaign's performance summary plus per-sequence-step stats, for the campaign's whole history or for the emails sent in a date range. total_contacts counts enrolled leads, and emails_pending is the remaining planned email-step sends across those leads. The campaign must belong to the selected workspace. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Leave out both from and to for the campaign's whole history. Supplying only one of them, a date in another format, or a from after to returns 400 with code bad_request.

A period is a send cohort: it selects the emails sent on its days, from the start of from to the end of to in UTC, and counts every open, click, reply and bounce those emails earned, whenever it arrived. An email sent on September 5 that is replied to on September 10 counts in a September 1 to 7 period; a reply that arrives on September 3 to an email sent in August does not. That keeps each rate's numerator and denominator the same emails. The period scopes emails_sent, the open, click, reply and bounce counts and rates, every row in steps and every bucket in engagement (an open or click counts when the email it answers was sent in the period). total_contacts and emails_pending describe the campaign as it stands now and are the same whatever the period.

date_range is the period the figures cover, as midnight UTC on its first and last day: the from and to you sent, or for the whole history the day of the campaign's first send (its creation day before anything is sent) through today.

Response

{
  "campaign_id": "b1f2c3d4-0000-0000-0000-000000000001",
  "name": "Q2 outbound",
  "status": "active",
  "date_range": { "from": "2026-05-04T00:00:00Z", "to": "2026-06-12T00:00:00Z" },
  "summary": {
    "total_contacts": 500,
    "emails_sent": 820,
    "emails_pending": 60,
    "unique_opens": 410,
    "machine_opens": 52,
    "unique_clicks": 99,
    "machine_clicks": 4,
    "replies": 41,
    "bounces": 5,
    "unsubscribes": 3,
    "open_rate": 50.0,
    "click_rate": 12.07,
    "reply_rate": 5.0,
    "bounce_rate": 0.61
  },
  "steps": [
    {
      "step_id": "5e9e0001-0000-0000-0000-000000000004",
      "name": "Intro",
      "position": 1,
      "emails_sent": 500,
      "opens": 260,
      "machine_opens": 31,
      "clicks": 61,
      "machine_clicks": 2,
      "replies": 28,
      "bounces": 3,
      "open_rate": 52.0,
      "click_rate": 12.2,
      "reply_rate": 5.6,
      "bounce_rate": 0.6
    }
  ],
  "engagement": {
    "countries": [{ "key": "US", "opens": 120, "clicks": 31 }, { "key": "DE", "opens": 44, "clicks": 9 }],
    "clients": [{ "key": "Gmail", "opens": 98, "clicks": 20 }, { "key": "Outlook", "opens": 51, "clicks": 12 }],
    "devices": [{ "key": "desktop", "opens": 140, "clicks": 37 }, { "key": "mobile", "opens": 24, "clicks": 3 }],
    "surfaces": [{ "key": "hidden", "opens": 98, "clicks": 0 }, { "key": "desktop_app", "opens": 51, "clicks": 12 }, { "key": "mobile", "opens": 0, "clicks": 3 }]
  }
}

engagement groups the campaign's human opens and clicks by country (ISO code), mail client, device type (desktop, mobile, tablet) and surface, counting distinct contacts per bucket. A clients key is the mail client (Gmail, Apple Mail, Outlook), Webmail in <browser> for an open read in a browser that named no client, or for a click the browser the link opened in. surfaces joins device and app or webmail: mobile_app, desktop_app, tablet_app, webmail, the bare device type when neither is known (a click, unless a mail app made the request itself), and hidden when a mailbox provider's image proxy fetched the email and the device cannot be known. Each list holds the busiest eight; an empty key is unknown. Country needs the GeoLite2 database on the consumer, otherwise every country row is unknown.

unique_opens, opens, total_opens and every open rate count a person's opens only. machine_opens is the separate count of steps whose only open came from an automated fetcher (UA-less clients, known scanner networks, and opens inside the instance's automated-open window, which starts when the step is dispatched to a worker); it is not part of the open counts. Apple Mail Privacy Protection's relay (a user agent of just Mozilla/5.0) and the stripped WebKit signature Mail on a Mac and the new Outlook for Windows share are counted as a prefetch only inside that window; by themselves they do not prove automation. A person's later open on a machine-opened step moves it from machine_opens into the open count. machine_clicks counts the contacts whose only clicks on a step were automated (a security gateway walking the links); those are not part of unique_clicks or total_clicks, which only ever count a person's click.

steps lists the campaign's email steps in sequence order, the order the steps list returns them. position is the step's 1-based number among those email steps, the N the canvas shows as Email N. It is not the step's own position field: action steps are not counted and a deleted step leaves no gap. Match a row to its step by step_id.

Each entry in steps carries that step's own rates: open_rate, click_rate, reply_rate and bounce_rate are percentages of that step's emails_sent, not of the campaign's, so a follow-up that reached a tenth of the contacts still compares against the first touch. They are 0 while the step has sent nothing. machine_opens and machine_clicks follow the same rules as their summary counterparts, scoped to the step.

Get campaign daily stats

GET /analytics/campaigns/:id/daily

Returns per-day send, open, click, and reply counts for one campaign over a date range. The campaign must belong to the caller. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Each row is one UTC day with at least one send, and its opens, clicks and replies are those earned by that day's sends, whenever they arrived. The rows for a period add up to the campaign analytics for the same from and to.

Response

The series is returned under a data envelope.

{
  "data": [
    { "date": "2026-06-05", "sent": 120, "opens": 61, "clicks": 14, "replies": 6 },
    { "date": "2026-06-06", "sent": 110, "opens": 58, "clicks": 12, "replies": 5 }
  ]
}

Get campaign hourly stats

GET /analytics/campaigns/:id/hourly

Returns per-hour stats for one campaign on a single day. The campaign must belong to the caller. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Response

The series is returned under a data envelope, with the resolved date echoed back. Each item's hour is the UTC hour, 0-23, so a day's rows add up to that day's row in the daily stats.

{
  "data": [
    { "hour": 9, "sent": 22, "opens": 11, "clicks": 3, "replies": 1 },
    { "hour": 10, "sent": 30, "opens": 16, "clicks": 4, "replies": 2 }
  ],
  "date": "2026-06-11"
}

Compare campaigns

GET /analytics/campaigns/compare

Returns side-by-side performance for up to 10 campaigns over a date range. Every requested campaign must belong to the caller. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Each campaign's figures are the emails it sent from the start of from to the end of to, with every open, click, reply and bounce they earned, the same send cohort as campaign analytics for that period.

Response

{
  "campaigns": [
    {
      "campaign_id": "b1f2c3d4-0000-0000-0000-000000000001",
      "name": "Q2 outbound",
      "status": "active",
      "emails_sent": 820,
      "open_rate": 50.0,
      "click_rate": 12.07,
      "reply_rate": 5.0,
      "bounce_rate": 0.61
    },
    {
      "campaign_id": "b1f2c3d4-0000-0000-0000-000000000005",
      "name": "Reactivation",
      "status": "paused",
      "emails_sent": 410,
      "open_rate": 44.1,
      "click_rate": 9.8,
      "reply_rate": 3.4,
      "bounce_rate": 1.2
    }
  ],
  "period": {
    "from": "2026-06-01T00:00:00Z",
    "to": "2026-06-12T00:00:00Z"
  }
}

List account statuses

GET /analytics/accounts

Returns the health and usage status of email accounts in the selected workspace, one bounded page at a time. The per-mailbox reads are batched across the page, so a page's cost does not grow with the size of the whole inventory. The request fails if a page's usage read cannot be completed, so clients never receive a silently incomplete page. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Query parameters

Response

Returned under the standard data plus pagination envelope. pagination.next_cursor is an opaque token for the next page, or null on the last page. Each item in data has the same shape as get account status. A workspace with more mailboxes than one page follows next_cursor to reach the rest, rather than having the overflow silently dropped.

{
  "data": [
    {
      "id": "a0a1a2a3-0000-0000-0000-000000000003",
      "email": "[email protected]",
      "provider": "google",
      "status": "active",
      "last_synced_at": "2026-06-12T08:00:00Z",
      "health": { "status": "healthy", "score": 100, "issues": [] },
      "errors": [],
      "daily_usage": {
        "date": "2026-06-12",
        "campaign_sent": 18,
        "campaign_limit": 50,
        "warmup_sent": 22,
        "warmup_limit": 40
      },
      "in_campaign": true
    }
  ],
  "pagination": {
    "total": 1,
    "next_cursor": null,
    "has_more": false
  }
}

Get account status

GET /analytics/accounts/:id

Returns the detailed status for one email account: a combined health score (folding in warmup-pool reputation and measured warmup placement), active errors, today's UTC usage, warmup status, warmup-pool health, and the trailing warmup inbox rate. The account must belong to the selected workspace. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

Response

{
  "id": "a0a1a2a3-0000-0000-0000-000000000003",
  "email": "[email protected]",
  "provider": "google",
  "status": "active",
  "last_synced_at": "2026-06-12T08:00:00Z",
  "health": {
    "status": "warning",
    "score": 90,
    "issues": ["Warmup reputation needs watching"]
  },
  "errors": [
    {
      "id": "e1e1e1e1-0000-0000-0000-000000000006",
      "error_code": "IMAP_AUTH",
      "severity": "WARNING",
      "title": "Mailbox reconnect recommended",
      "message": "Token nearing expiry",
      "created_at": "2026-06-11T22:14:00Z"
    }
  ],
  "daily_usage": {
    "date": "2026-06-12",
    "campaign_sent": 18,
    "campaign_limit": 50,
    "warmup_sent": 22,
    "warmup_limit": 40
  },
  "warmup_status": {
    "enabled": true,
    "paused": false,
    "started_at": "2026-05-20T00:00:00Z",
    "current_volume": 22,
    "target_volume": 33,
    "max_volume": 40,
    "reply_rate": 35,
    "days_active": 23
  },
  "warmup_health": {
    "state": "watch",
    "score": 78,
    "spam_score": 0,
    "partner_mailboxes_7d": 14,
    "partner_domains_7d": 11,
    "partner_organizations_7d": 9,
    "received_7d": 12,
    "senders_7d": 8,
    "evaluated_at": "2026-06-12T06:00:00Z"
  },
  "warmup_placement": {
    "window_days": 7,
    "min_sample": 20,
    "scope": "major",
    "delivered": 142,
    "inbox": 126,
    "tabs": 5,
    "spam": 11,
    "inbox_rate": 92.25,
    "band": "good"
  },
  "in_campaign": true
}

warmup_status is present only when warmup has ever been enabled; warmup_health is present only when the mailbox is in a warmup pool. in_campaign reports whether the mailbox currently backs a live campaign.

warmup_placement is the mailbox's measured warmup inbox rate over the trailing 7 UTC days, the same figure Get warmup placement reports as rate, and is absent when nothing was delivered in that window. Once inbox_rate is present, health.score is capped at it (rounded down), so a mailbox landing any mail in spam no longer reads 100, and a rate below 90 also sets health.status to at least warning with an issue naming the rate.

daily_usage.warmup_sent and campaign_sent both count sends the mailbox completed on that date, which is what the caps are enforced against, so neither can read above the target in force.

The three partner_*_7d counters are the distinct partners reached by confirmed warmup sends over the last seven days: mailboxes, their domains, and the workspaces behind them. Failed and still-pending attempts do not count. partner_organizations_7d of 1 over a full week means the mailbox is only warming inside one workspace, which is expected on a self-hosted instance that is not linked to Warmbly Cloud and worth investigating anywhere else.

received_7d and senders_7d are the other direction over the same window: verified warmup arrivals and the distinct partners they came from. Only mail that the mailbox's own sync verifies counts. A mailbox with many partners on the sending side and few on the receiving side is being written to less than it writes; the pool favours it on every draw until the two even out.

Get usage overview

GET /analytics/usage

Returns account, campaign, and contact usage counters for the selected workspace. campaigns.emails_sent counts sent email steps in the requested period; wait and action steps do not count. Auth: Scope READ_ANALYTICS · Org permission view_analytics.

The api object is a compatibility placeholder. For real API-key request totals, error rates, latency, and endpoint breakdowns use GET /api-keys/usage/summary and the API-key analytics endpoints.

Response

{
  "user_id": "11111111-0000-0000-0000-000000000007",
  "period": "day",
  "email_accounts": { "total": 12, "active": 11, "in_warmup": 8, "with_errors": 1 },
  "campaigns": { "total": 9, "active": 4, "paused": 2, "draft": 3, "emails_sent": 124 },
  "contacts": { "total": 8200, "subscribed": 8050, "added_today": 120 },
  "api": { "total_calls": 0, "daily_limit": 50000, "top_endpoints": [] }
}

List audit logs

GET /audit-logs

Returns the organization-wide activity trail for the caller's current organization ("who did what, when, from where"). The organization is always taken from the session and never from a client parameter, so one organization can never read another's trail. Auth: Scope READ_AUDIT_LOGS · Org permission view_analytics.

Response

A data plus pagination envelope with an opaque cursor. actor is null when the acting user has since been deleted; entity_id, changes, and metadata are omitted when empty.

{
  "data": [
    {
      "id": "9c9c9c9c-0000-0000-0000-000000000008",
      "org_id": "0a0a0a0a-0000-0000-0000-000000000009",
      "user_id": "11111111-0000-0000-0000-000000000007",
      "actor": {
        "id": "11111111-0000-0000-0000-000000000007",
        "first_name": "Ada",
        "last_name": "Lovelace",
        "email": "[email protected]"
      },
      "action_date": "2026-06-12T08:14:00Z",
      "action": "update",
      "entity_type": "campaign",
      "entity_id": "b1f2c3d4-0000-0000-0000-000000000001",
      "ip_address": "203.0.113.10",
      "user_agent": "Mozilla/5.0",
      "changes": { "status": "paused" },
      "timestamp": "2026-06-12T08:14:00Z"
    }
  ],
  "pagination": {
    "next_cursor": "c1_b3BhcXVlLWN1cnNvcg",
    "has_more": true
  }
}

Secret values (API key material, webhook secrets, passwords) are never recorded in changes or metadata; the trail records only that a field changed.