傳送時間最佳化 send-time-optimization

貴組織應先在Journey Optimizer中使用「電子郵件」動作，至少30天後再在電子郵件中使用「傳送時間最佳化」 ，以收集一些電子郵件傳送、開啟和點按事件。

貴組織應先在Journey Optimizer內使用「推播」動作30天以上，再使用「推播」內的「傳送時間最佳化」來允許收集某些推播傳送和開啟事件。

如果您的組織已使用電子郵件及/或推播動作型別至少30天，則貴組織在Adobe啟用傳送時間最佳化後，不需要等候更久即可使用。 隨著您的組織收集資料長達16週，結果將會持續改善。

傳送時間最佳化可在組織最佳化的所有郵件中，提高電子郵件點按率和推播開啟率，幅度約為2%至10%。

例如，如果組織傳送電子郵件時未將傳送時間最佳化，則平均點按率為5.0%，而同一組具有傳送時間最佳化的電子郵件平均可產生高達5.5%的點按率(5.0% * (1+10%) = 5.5%)。

由於小型樣本大小中的可變性，在傳送單一訊息時，可能無法觀察到傳送時間最佳化的好處。

組織在以下情況下更有可能從使用傳送時間最佳化中獲得更大的好處：

• 現有歷程使用的傳送時間是固定的且未妥善最佳化
• 客戶行為的可變性（點按和開啟）會對應至客戶地點和客戶偏好設定
• 組織對大部分電子郵件和推播訊息使用傳送時間最佳化
• 組織選擇建議的6-12小時範圍內的最大等待時間

發生此問題的原因有多種：

• 您的訊息已選取為「探索」訊息傳送時間，而非「最佳化」訊息傳送時間。
• 相似使用者的行為影響了模型以建議另一個傳送時間。

傳送時間最佳化可在下列情況下，於使用者當地時區的夜間將推播訊息傳送給使用者：

• 當使用者表現出行為表示他們很可能會與夜間傳送的訊息互動時
• 模型選擇「探索」傳送時間時

為避免在夜間時段向客戶傳送推播訊息，請排程批次推播訊息傳送以在早上或午後發生，並選擇較短的持續時間以進行傳送時間最佳化。 （例如，上午9點的傳送時間與最長8小時的等候時間。）

This section contains structured knowledge intended to support interpretation, retrieval, and question answering related to this topic.

For complete understanding, this information should be combined with the documentation on this page. Neither source is intended to stand alone; the page describes the feature, while this section provides additional context that helps disambiguate terminology, intent, applicability, and constraints.

• TL;DR: This page explains how to configure and use Send-Time Optimization in Adobe Journey Optimizer, an AI-powered feature that predicts the best time to send email or push messages to each individual to maximize engagement.

Intents:

• Enable Send-Time Optimization on an email or push action in a journey
• Choose whether to optimize for opens or click-throughs on email messages
• Set the maximum wait window (Send within next) for delayed delivery
• Understand how the AI model predicts optimal send times using behavioral data
• Determine whether Send-Time Optimization is appropriate for a given message type
• Use Send-Time Optimization within a Wait activity to delay before any downstream activity, decoupled from the message send

Glossary:

• Send-Time Optimization (STO): An AI-powered feature that delays message delivery to each profile until the predicted optimal engagement hour within a configured time window (product-specific)
• Journey AI: Adobe’s AI services powering Send-Time Optimization within Journey Optimizer (product-specific)
• Exploration send time: A randomly selected send time (used for 5% of sends) to test different times and improve model accuracy (product-specific)
• Optimized send time: A model-predicted send time selected to maximize click or open rates (used for 95% of sends) (product-specific)
• Send within next: The maximum number of hours (2–100) the system will wait before sending the message to a given profile (product-specific)

Guardrails:

• Send-Time Optimization must be enabled by Adobe for the organization; contact Adobe Customer Care or your Adobe representative to activate it.
• Send-Time Optimization applies to Email and Push notification channels within Journeys, and to the Wait activity; it is not available for Campaigns or custom actions.
• Send-Time Optimization has no visibility into quiet hours rules; a Send-Time Optimization Wait activity can select a time inside a quiet-hours window for a downstream channel action, which may then queue or discard the message depending on the quiet hours rule configuration.
• The organization must have used Email or Push actions in Journey Optimizer for at least 30 days before Send-Time Optimization produces meaningful results.
• Do not use Send-Time Optimization for urgent or time-sensitive operational messages (e.g., order confirmations, password resets, flight gate changes).
• Maximum wait time range is 2–100 hours; recommended range is 6–24 hours for best results.
• Model scores are stored in profile attributes at _experience.intelligentServices.journeyAI.sendTimeOptimization and are not human-readable.
• Models are trained weekly initially, then retrained and rescored monthly after 16 weeks.

Terminology:

• Canonical name: Send-Time Optimization — Acronym: STO — variants: best send time, send time AI, intelligent send time
• Synonyms: “Send-Time Optimization” = “optimal send time” = “AI send time”
• Do not confuse: “Exploration send time” ≠ “Optimized send time” (exploration is random for model testing; optimized is model-predicted for engagement)

FAQ:

• Q: Which channels support Send-Time Optimization? — Email and Push notification channels within Journeys, and the Wait activity; Campaigns and custom actions are not supported.
• Q: Does Send-Time Optimization know about quiet hours? — No. Quiet hours are only evaluated when a profile reaches a message action, so a Send-Time Optimization Wait activity can pick a time inside a quiet-hours window. Depending on the quiet hours rule, the message is then queued until quiet hours end, or discarded and the profile exits the journey. Learn more.
• Q: Should I optimize for opens or clicks on email? — Optimize for Clicks for most emails. Choose Opens when the message is informational and not intended to drive a specific action.
• Q: How long does the organization need to wait before enabling STO? — At least 30 days of Email or Push usage in Journey Optimizer is needed to collect sufficient behavioral data. Results continue to improve for up to 16 weeks.
• Q: Can STO send push notifications at night? — Yes, if a user’s behavior suggests night-time engagement or if an exploration send time is selected. To avoid this, use a morning send time with a short maximum wait window.
• Q: What is the expected benefit of Send-Time Optimization? — Approximately 2–10% improvement in email click rate or push open rate across all optimized messages, though benefits may not be observable on individual small-volume sends.