Pancake Statistics API — 7 endpoints (page, user, tag, engagement, ads, new customer)

TL;DR: Pancake exposes 7 Statistics endpoints under https://pages.fm/api/public_api/v1/pages/{page_id}/statistics/... for reporting: pages (page overview by hour), users (agent KPIs), tags (per-tag breakdown), customer_engagements (engagement chart), pages_campaigns + ads (ad reporting), page_customers (new customer growth). All require a page_access_token. Critical: Pancake uses 3 different date param formats across endpoints (since/until Unix seconds, date_range string "DD/MM/YYYY HH:MM:SS - DD/MM/YYYY HH:MM:SS", date_range string "DD/MM/YYYY - DD/MM/YYYY") — wrong format = empty response or 400 with no clear error message.

Once you know how to list conversations, sync customers, and manage tags/assignments, this post covers the remaining Pancake chat endpoint group — Statistics. This is what you reach for when building internal BI dashboards, agent KPI reports, ad ROI tracking, or scheduled email/Slack reports.

1. Use cases

2. Setup + critical warning about date format

All endpoints need a page_access_token with reports permission — see the setup in the Webhook & API guide.

⚠ 3 different date param formats (THE single most important thing to remember)

Pancake is inconsistent about date params across Statistics endpoints. This is the most common bug source when building a pipeline:

// 3 helpers matching the 3 endpoint groups
const toUnixSec = (d) => Math.floor(d.getTime() / 1000)
const toDateRangeLong = (from, to) =>
  `${pad(from, 'DD/MM/YYYY HH:MM:SS')} - ${pad(to, 'DD/MM/YYYY HH:MM:SS')}`
const toDateRangeShort = (from, to) =>
  `${pad(from, 'DD/MM/YYYY')} - ${pad(to, 'DD/MM/YYYY')}`

3. Page — overview by hour

GET https://pages.fm/api/public_api/v1/pages/{page_id}/statistics/pages
  ?page_access_token={token}
  &since={unix_sec}
  &until={unix_sec}

cURL example — last 7 days

SINCE=$(date -u -v-7d +%s)   # macOS; Linux: date -u -d '7 days ago' +%s
UNTIL=$(date -u +%s)

curl -G "https://pages.fm/api/public_api/v1/pages/$PAGE_ID/statistics/pages" \
  --data-urlencode "page_access_token=$TOKEN" \
  --data-urlencode "since=$SINCE" \
  --data-urlencode "until=$UNTIL"

Response shape

{
  "success": true,
  "data": [
    {
      "hour": "2026-06-23T10:00:00",
      "page_inbox_count": 45,
      "page_comment_count": 12,
      "customer_inbox_count": 38,
      "customer_comment_count": 10,
      "new_inbox_count": 5,
      "new_customer_count": 3,
      "phone_number_count": 8,
      "uniq_phone_number_count": 7,
      "inbox_interactive_count": 22,
      "today_uniq_website_referral": 0,
      "today_website_guest_referral": 0
    }
  ]
}

Field meanings

Use case: chart "New conversations per hour this week" — find the peak hours and schedule agent shifts accordingly.

4. User — agent KPIs (response time, conversations handled)

GET https://pages.fm/api/public_api/v1/pages/{page_id}/statistics/users
  ?page_access_token={token}
  &date_range={DD/MM/YYYY HH:MM:SS - DD/MM/YYYY HH:MM:SS}

cURL example

curl -G "https://pages.fm/api/public_api/v1/pages/$PAGE_ID/statistics/users" \
  --data-urlencode "page_access_token=$TOKEN" \
  --data-urlencode "date_range=01/06/2026 00:00:00 - 23/06/2026 23:59:59"

Response shape

{
  "success": true,
  "data": {
    "statistics": {
      "user_uuid_1": [
        {
          "hour": "2026-06-15T10:00:00",
          "average_response_time": 74276,
          "inbox_count": 6,
          "comment_count": 0,
          "phone_number_count": 1,
          "private_reply_count": 0,
          "unique_inbox_count": 2,
          "unique_comment_count": 0
        }
      ]
    },
    "users": {
      "user_uuid_1": {
        "user_name": "John Doe",
        "user_fb_id": "570423649807405",
        "average_response_time": 588569,
        "inbox_count": 637,
        "comment_count": 0,
        "phone_number_count": 5,
        "unique_inbox_count": 30,
        "private_reply_count": 0
      }
    }
  }
}

The response is split into two views:

• statistics — broken down per user × per hour (granular for charts)
• users — aggregated per user (for leaderboards)

Heads-up — average_response_time is in SECONDS

// The value is in seconds — convert to minutes for display if needed
const seconds = user.average_response_time              // 588 seconds
const minutes = (user.average_response_time / 60).toFixed(1)  // 9.8 minutes

Use case: "Top 5 fastest-replying agents this week" dashboard + monthly KPI report for HR/Sales lead.

5. Tag — conversation stats per tag

GET https://pages.fm/api/public_api/v1/pages/{page_id}/statistics/tags
  ?page_access_token={token}
  &since={unix_sec}
  &until={unix_sec}

Response shape

{
  "success": true,
  "data": {
    "categories": ["2026-06-20", "2026-06-21", "2026-06-22"],
    "series": {
      "tag_id_1": [10, 15, 8],
      "tag_id_2": [3, 5, 2]
    }
  },
  "data_today": {
    "categories": ["00:00", "01:00", "02:00"],
    "series": {
      "tag_id_1": [1, 0, 2]
    }
  },
  "tags": [
    { "id": "tag_id_1", "text": "Orders", "color": "#2ecc71" },
    { "id": "tag_id_2", "text": "Complaints", "color": "#e74c3c" }
  ]
}

3 response parts:

• data — time-range series (categories = days, series = each tag has an array of daily counts)
• data_today — today's breakdown by hour (24 entries)
• tags — tag metadata (ID → text + color mapping to render legends)

Use case: "Which tags brought in the most conversations this week" report → decide which ads campaign to scale.

6. Engagement — customer engagement chart over time

GET https://pages.fm/api/public_api/v1/pages/{page_id}/statistics/customer_engagements
  ?page_access_token={token}
  &date_range={DD/MM/YYYY HH:MM:SS - DD/MM/YYYY HH:MM:SS}
  &by_hour={true|false}
  &user_ids={uid1,uid2}

Params

Response shape

{
  "success": true,
  "data": {
    "categories": ["20/6", "21/6", "22/6"],
    "series": [
      { "name": "inbox", "data": [45, 52, 38] },
      { "name": "comment", "data": [12, 18, 10] },
      { "name": "total", "data": [57, 70, 48] },
      { "name": "new_customer", "data": [5, 8, 3] },
      { "name": "new_customer_from_inbox", "data": [4, 7, 2] }
    ]
  },
  "statistics": [
    {
      "hour": "2026-06-22T10:00:00",
      "inbox": 5,
      "comment": 1,
      "total": 6,
      "new_customer": 1,
      "new_customer_from_inbox": 1
    }
  ]
}

2 views:

• data — pre-formatted for chart libraries (categories = X-axis, series[] = multiple line series). Pass straight to ApexCharts/Chart.js and it renders.
• statistics[] — raw data per hour, for custom aggregation or BI tooling

Engagement metrics

• inbox — number of customers interacting via inbox (both new and existing). Each customer counted once per day.
• comment — number of customers interacting via comment (also counted once per day per customer)
• total — total inbox + comment, deduplicated (if a customer interacts on both channels, counted once)
• new_customer — new customers (never interacted before) via message or comment
• new_customer_from_inbox — new customers who first started an inbox conversation (still counted if they previously commented)

Use case: "Customer engagement growth this month" dashboard — render a 5-line chart and email the monthly report to leadership.

7. Ads + Ads Campaign — advertising stats

Pancake has 2 separate endpoints for ads: campaign level (budget overview) and ad level (per-ad detail).

7.1. /statistics/pages_campaigns — campaign level

GET /pages/{page_id}/statistics/pages_campaigns?page_access_token=...&since=...&until=...

curl -G "https://pages.fm/api/public_api/v1/pages/$PAGE_ID/statistics/pages_campaigns" \
  --data-urlencode "page_access_token=$TOKEN" \
  --data-urlencode "since=$SINCE" \
  --data-urlencode "until=$UNTIL"

Response array:

[
  {
    "account_id": "act_123456789",
    "ad_id": "ad_uuid_1",
    "adset_id": "adset_uuid_1",
    "budget_remaining": "150000",
    "currency": "VND",
    "daily_budget": "500000",
    "status": "ACTIVE"
  }
]

Use case: "Which campaigns are running low on budget" dashboard → early alert for top-up.

7.2. /statistics/ads — per-ad detail

GET /pages/{page_id}/statistics/ads
  ?page_access_token=...
  &since=...&until=...
  &type=by_id    # or by_time

curl -G "https://pages.fm/api/public_api/v1/pages/$PAGE_ID/statistics/ads" \
  --data-urlencode "page_access_token=$TOKEN" \
  --data-urlencode "since=$SINCE" \
  --data-urlencode "until=$UNTIL" \
  --data-urlencode "type=by_id"

Response:

[
  {
    "name": "Ad - June 6 Promo",
    "status": "ACTIVE",
    "currency": "VND",
    "reach": "15234",
    "impressions": "45678",
    "spend": "230000"
  }
]

Use case: "Top 10 ads by reach this month" report → decide which ads to scale up.

8. New Customer — new customer growth

GET https://pages.fm/api/public_api/v1/pages/{page_id}/statistics/page_customers
  ?page_access_token={token}
  &date_range={DD/MM/YYYY - DD/MM/YYYY}
  &group_by={day|hour|page_id}

Params

cURL example

curl -G "https://pages.fm/api/public_api/v1/pages/$PAGE_ID/statistics/page_customers" \
  --data-urlencode "page_access_token=$TOKEN" \
  --data-urlencode "date_range=01/06/2026 - 23/06/2026" \
  --data-urlencode "group_by=day"

Response shape — group_by=day/hour

{
  "success": true,
  "data": [
    {
      "date": "2026-06-20",
      "new_customer_count": 4,
      "new_page_inbox_count": 3,
      "new_phone_number_count": 0,
      "phone_number_count": 0
    }
  ]
}

Response shape — group_by=page_id

{
  "success": true,
  "data": {
    "page_id_1": [
      { "date": "2026-06-20", "new_customer_count": 0, "new_page_inbox_count": 1, "new_phone_number_count": 0, "phone_number_count": 0 }
    ],
    "page_id_2": [
      { "date": "2026-06-20", "new_customer_count": 2, "new_page_inbox_count": 1, "new_phone_number_count": 0, "phone_number_count": 0 }
    ]
  }
}

Use case: chart "Customer Acquisition by week" + compare multiple pages if you manage a multi-page setup.

9. Combined pitfalls

10. FAQ

Q: Which endpoint gives the total customers this month?

A: Use /statistics/page_customers?date_range=01/06/2026 - 30/06/2026&group_by=day and sum new_customer_count across all 30 days. Pancake doesn't expose a direct "total" endpoint.

Q: Is there an endpoint that returns CSV/Excel directly?

A: No. The Statistics API only returns JSON. To get CSV → convert client-side (e.g. papaparse in Node) or use /export_data (only for conversations from ads, not for statistics).

Q: What's the rate limit on Statistics endpoints?

A: Same as the rest of the Public API: 5 req/page/second. Statistics is usually called periodically (hourly cron, not realtime), so you rarely hit the limit.

Q: How is inbox_count in /statistics/users (per agent) different from /statistics/pages?

A: /users counts per agent (messages that agent handled); /pages counts per page (total messages the page received/sent). Summing all users does not necessarily equal the page total — auto-replies and unassigned messages also count.

Q: Can I sync Statistics data to Google Looker Studio / Metabase?

A: Yes. Pattern: cron every 1 hour calling all 7 endpoints → write to Postgres/BigQuery → connect Looker Studio/Metabase via standard connector. See the 3 ways to extract Pancake data post for the integration pattern.

Q: Can I filter Statistics by channel (FB / IG / Zalo)?

A: Statistics currently only aggregates at the page level, with no channel breakdown. For per-channel data, you'd have to build it yourself by querying list conversations with a type filter and aggregating.

11. Summary

• 7 Statistics endpoints under /pages/{id}/statistics/...: pages (page overview), users (agent KPIs), tags (per tag), customer_engagements (chart), pages_campaigns (campaign budgets), ads (per-ad detail), page_customers (new customer growth).
• 3 different date param formats across endpoints — pitfall #1, write 3 separate helpers.
• average_response_time is in SECONDS despite the OpenAPI spec saying "milliseconds" — verify with your own page's data before displaying.
• Tag series is an object not an array — use Object.entries().
• /ads requires the type param (by_id | by_time).
• /page_customers Statistics ≠ /page_customers List — same path, different response.
• Main use cases: internal BI dashboards, agent KPIs, ad ROI, scheduled reports.

Related posts

Pancake chat cluster

• The Pancake Ecosystem
• Pancake Webhook & API Send Message Guide
• Get Pancake Conversations List API
• Pancake Page Customers API
• 3 Ways to Extract Pancake Data into Your CRM
• Conversation Workflow API — Tag, Assign, Mark Read/Unread

Pancake CRM cluster

• Introducing the Pancake CRM API
• Records CRUD (lead/account/contact)

Official documentation

• Pancake Developer Docs — full OpenAPI spec
• Pancake User Docs — Statistics (metric definitions in English) — read alongside this post when you need the precise business meaning of a metric

Last updated: 2026-06-23. API version: public_api/v1 for all Statistics endpoints. Check the OpenAPI spec to verify any changes.