new_releases 更新:查看
版本資訊瞭解新功能和產品更新。
虛擬服務專員數據分析
透過集合功能整理內容
你可以依據偏好儲存及分類內容。
代理程式開始傳送訊息後,您可以追蹤代理程式的訊息和對話指標。Business Communications Developer 控制台會回報已傳送訊息的分析資料,您也可以追蹤代理程式對話的事件和其他屬性,自行開發分析資料。
Business Communications 開發人員控制台中的數據分析
Business Communications 開發人員控制台支援三個與服務專員訊息相關的指標:
- 已傳送:服務專員傳送的訊息數量。
- 已送達:成功傳送給使用者的訊息數量。
- Read:使用者讀取的訊息數量。
控制台會在每天早上重新整理指標,並以世界標準時間 (+0) 的時區為所有指標加上時間戳記。
所有指標都會依據特定日期傳送的訊息進行分組,也就是說,與訊息相關聯的所有事件 (「傳送」、「已送達」、「已讀」) 會顯示在服務專員傳送訊息的日期 (以世界標準時間為準)。如果在後續幾天發生「已送達」或「已讀取」事件,指標會回填至相關聯訊息的「傳送」日期。主控台會回填八天的數據分析資料。
舉例來說,如果服務專員在 1 月 1 日傳送訊息,且使用者的裝置在 1 月 1 日收到訊息,則 1 月 1 日的「Sent」和「Delivered」指標在隔天更新時都會增加 1 次。如果使用者在 1 月 2 日讀取訊息,則在隔天更新指標時,1 月 1 日 (代理人傳送訊息的日期) 的「Read」指標會增加 1 次。不過,如果使用者在 1 月 10 日 (超過八天的回補時間範圍) 讀取訊息,1 月 1 日的「Read」指標就不會增加,且不會回報「Read」事件。
查看服務專員數據分析
- 開啟 Business Communications 開發人員控制台,使用 RBM Google 帳戶登入,然後點選您的服務專員。
- 在左側導覽列中,按一下「Analytics」。
- 指定資料的時間刻度、開始日期和結束日期。
- 指定要查看的指標。
如要將目前的數據分析匯出為 CSV 檔案,請在數據分析窗格中開啟選單,然後按一下「匯出資料」。
評估自有數據分析
如果您有不同的數據分析偏好設定,或是需要比 Business Communications 開發人員控制台提供的更深入資訊,可以自行擷取數據分析資料。
資料擷取和儲存需求會因機構和用途而異。遵守所有適用政策和法規。
以下是您可以擷取的部分指標:
- 傳送訊息:每次客服專員傳送訊息時,記錄詳細資料。追蹤
messageId
可讓您將訊息與後續事件和回應建立關聯。
- 送達回條:擷取
DELIVERED
事件。
- 讀取回條。擷取
READ
事件。
- 使用者參與。擷取
IS_TYPING
事件、使用者回應數,以及 READ
事件和後續使用者回應之間的時間差異。
- 互動類型。瞭解使用者如何回應您的訊息。使用有意義的
postbackData
追蹤建議的回覆和動作,並監控回覆格式,以識別使用者是否傳送檔案、位置或純文字訊息。
- 服務專員回應速度:記錄服務專員回覆使用者訊息所需的時間。
擷取的數據分析資料可供您剖析並整理,以符合業務需求。
除非另有註明,否則本頁面中的內容是採用創用 CC 姓名標示 4.0 授權,程式碼範例則為阿帕契 2.0 授權。詳情請參閱《Google Developers 網站政策》。Java 是 Oracle 和/或其關聯企業的註冊商標。
上次更新時間:2024-11-22 (世界標準時間)。
[null,null,["上次更新時間:2024-11-22 (世界標準時間)。"],[[["\u003cp\u003eThe Business Communications Developer Console provides analytics for agent messages, tracking the number of messages sent, delivered, and read by users.\u003c/p\u003e\n"],["\u003cp\u003eMetrics in the console are refreshed daily, grouped by the date the message was sent, and backfilled for up to eight days to account for delayed delivery or read events.\u003c/p\u003e\n"],["\u003cp\u003eUsers can access and view agent analytics in the console by navigating to the 'Analytics' section, specifying the desired time frame, and selecting the metrics to view, with an option to export as a CSV file.\u003c/p\u003e\n"],["\u003cp\u003eFor more customized or in-depth analysis, developers can capture their own analytics data, including message details, delivery and read receipts, user engagement, interaction types, and agent responsiveness.\u003c/p\u003e\n"],["\u003cp\u003eCapturing message \u003ccode\u003emessageId\u003c/code\u003e and events like \u003ccode\u003eDELIVERED\u003c/code\u003e, and \u003ccode\u003eREAD\u003c/code\u003e, can help provide a more in-depth analysis.\u003c/p\u003e\n"]]],[],null,["# Agent analytics\n\nAfter your agent starts sending messages, you can track metrics for your\nagent's messages and conversations. The Business Communications Developer\nConsole reports analytics of sent messages, and you can develop your own\nanalytics by tracking events and other attributes of your agent's conversations.\n\nAnalytics in the Business Communications Developer Console\n----------------------------------------------------------\n\nThe Business Communications Developer Console supports three metrics surrounding agent messages:\n\n- **Sent**: The number of messages sent by an agent.\n- **Delivered**: The number of messages successfully delivered to users.\n- **Read**: The number of messages read by users.\n\nThe console refreshes metrics each morning and timestamps all metrics\nin the UTC (+0) timezone.\n\nAll metrics are grouped by messages sent in a given day, meaning that all events\nassociated with a message (**Sent** , **Delivered** , **Read** ) appear on the date\n(in UTC) that the agent sent the message. If a **Delivered** or **Read** event\noccurs on a subsequent day, the metrics are backfilled to the associated\nmessage's **Sent** date. The console backfills analytics for eight\ndays.\n\nFor example, if an agent sends a message on January 1st, and the user's device\nreceives it on January 1st, the **Sent** and **Delivered** metrics for January\n1st both increase by one when the metrics refresh the next day. If the user\nreads the message on January 2nd, the **Read** metric for January 1st (the date\nthe agent sent the message) increases by one when the metrics refresh the next\nday. However, if the user instead reads the message on January 10th, which is\noutside the eight-day backfill window, the **Read** metric for January 1st\ndoesn't increase, and the **Read** event isn't otherwise reported.\n\n### View agent analytics\n\n1. Open the [Business Communications Developer\n Console](https://business-communications.cloud.google.com?utm_source=/business-communications/rcs-business-messaging/guides/measure&utm_medium=devsite&utm_campaign=rcs-business-messaging), sign in with your RBM Google account, and click your agent.\n2. In the left navigation, click **Analytics**.\n3. Specify the time scale, start date, and end date for your data.\n4. Specify the metrics you want to view.\n\nTo export the current analytics as a CSV file, open the menu in the analytics\npane, then click **Export data**.\n\nMeasuring your own analytics\n----------------------------\n\nIf you have different analytics preferences or require more in-depth information\nthan is available in the Business Communications Developer Console, you can\ncapture your own analytics data.\n\nData capture and storage requirements vary by organization and use-case.\nFollow all policies and regulations that you are subject to.\n\nHere are some metrics that you might capture:\n\n- **Messages sent** . Capture details each time your agent [sends a\n message](/business-communications/rcs-business-messaging/guides/build/messages/send). Tracking the `messageId` lets you associate a message with the subsequent events and responses.\n- **Delivery receipts** . Capture [`DELIVERED`](/business-communications/rcs-business-messaging/guides/build/events#delivered) events.\n- **Read receipts** . Capture [`READ`](/business-communications/rcs-business-messaging/guides/build/events#user_reads_agent_message) events.\n- **User engagement** . Capture [`IS_TYPING`](/business-communications/rcs-business-messaging/guides/build/events#user_starts_typing) events, the number of user responses, and the time difference between `READ` events and the following user responses.\n- **Interaction type** . Identify how users respond to your messages. Use meaningful `postbackData` to track suggested [replies](/business-communications/rcs-business-messaging/guides/build/messages/send#replies) and [actions](/business-communications/rcs-business-messaging/guides/build/messages/send#actions), and monitor [response\n formats](/business-communications/rcs-business-messaging/guides/build/messages/receive#examples) to identify if users send files, locations, or plain text messages.\n- **Agent responsiveness**. Capture how long it takes your agent to respond to user messages.\n\nWith your analytics captured, you can parse and organize them however best suits\nyour business needs."]]