YundaDeskWhen buyers chase shipment, they are not asking for a perfect apology. They want certainty: Is my order fine? When will it ship? If it cannot ship, who is handling it?
Split chase-shipment messages into states: still on time, close to late, overdue, preorder, out of stock, or escalated. AI checks order status and answers first when facts are clear. If timing or risk crosses the line, it hands off with context.
Check Four Fields Before Using Any Script
A good where is my order script is built on facts, not softer wording. Before an agent or AI replies, it should check four fields:
| Field | Purpose | How to use it |
|---|---|---|
| Payment time | See whether the order is still inside the promise | Tell the buyer which processing day it is |
| Product status | Separate in-stock, preorder, custom, and out-of-stock items | Explain why it has not shipped |
| Warehouse status | See picking, carrier pickup, or exception state | Give the next action |
| Promised window | Compare against the order email or policy | Decide whether to escalate |
Without these fields, support can only comfort from memory. YundaDesk lets AI look up the order and knowledge base, answer when facts are clear, and hand off when the case is overdue or risky.
Once those fields are available, a chase-shipment script can move from “please wait” to “here is the current state, and here is when it escalates.” The illustrative model below shows how human first-response pressure can fall week by week when AI handles low-risk shipment questions first.
Human first-response trend for chase-shipment cases (illustrative)
Scenario One: The Order Is Not Late Yet
The order is still within the promised window, but the buyer may have missed the shipping note or need the item for a specific date.
Reply script:
I checked your order, and it is still within our normal processing window. You placed it on {order date}, and this item usually ships within {promised shipping window}. The current status is {warehouse status}.
Once it ships, we will send the tracking number to your email or message thread. If there is still no update by {specific date}, I will escalate it to our human support team for follow-up.
Do not only say “it is not late.” Explain why and give a date. This low-risk case can usually be handled by AI, as long as it does not invent warehouse activity.
Scenario Two: The Order Is Close to the Promised Date
When the order is close to the promised shipping date, the buyer’s patience is thinner. Asking them to wait can still be reasonable, but the next step must be explicit.
Reply script:
I can see that your order is expected to ship by {promised date}. It is now close to that timeline, and the current status is {warehouse status}.
If there is no shipment update by {date or time}, the conversation will be automatically escalated to a human teammate. You do not need to send the same message again. We will check whether the warehouse can prioritize it, split the shipment, or offer a clearer next option.
The key line is “You do not need to send the same message again.” With an Omnichannel inbox, messages from the website widget, email, WhatsApp, Messenger, Instagram, TikTok, LINE, and other channels land in one workspace.
Scenario Three: The Order Is Already Overdue
Once the order is overdue, the tone should move from explanation to ownership. Buyers can accept delays more easily than silence.
Reply script:
I am sorry. This order has passed the shipping window we originally promised. The current status is {warehouse status or exception reason}. I have escalated this to a human teammate and included your order details and previous conversation history.
Next, we will check three things first: whether the order can be prioritized, whether a split shipment is possible, and what options we can offer if it cannot ship right away. Any refund, compensation, or price change will be reviewed by a human before anything is processed.
Do not promise “it will ship today” unless the warehouse has confirmed it. Refunds, compensation, and price changes are high-risk actions, so humans should approve them.
Scenario Four: Preorder, Custom, or Out-of-Stock Orders
Preorder and custom items are easy to misunderstand. The buyer may not remember the shipping window; they remember paying.
Preorder script:
I checked your order, and this item is a preorder. The estimated shipping window is {date range}. Your order is not lost; it is still inside the preorder shipping plan. Once it ships, you will receive the tracking number. If you do not want to keep waiting, reply here and we will hand this to a human teammate to review your available options.
Out-of-stock script:
I checked your order, and it is currently held because {product name} is waiting for a stock update. To avoid leaving you waiting without a clear answer, I will escalate this to a human teammate. They will check whether we can ship available items first or offer another option.
Preorder explanations can often be handled by AI when the knowledge base and order status are clear. Out-of-stock cases should usually escalate because they may involve split shipment, replacement items, or refunds.
Scenario Five: The Buyer Keeps Chasing or the Tone Escalates
If the same buyer asks repeatedly in a short window, or uses words like scam, complaint, bad review, refund, or compensation, stop using the normal script.
Reply script:
I understand that this has already been frustrating. To avoid making you repeat the same details, I will send this conversation and your order information to a human teammate now.
They will review the current shipment status and give you a clear next step. If the case involves refund, compensation, or any money-related action, it will be confirmed by a human before anything is processed.
Use four escalation rules: hand off when the buyer asks for a human, when the same order gets repeated chase messages, when refund or complaint intent appears, or when AI lacks order evidence. YundaDesk keeps the boundary simple: AI answers first, humans back up.
Turn Scripts Into AI Rules
A script library that lives only in a spreadsheet still depends on agents copying and pasting the right line. Break it into knowledge base content and workflow rules instead.
| Intent | Condition | AI action | Escalation trigger |
|---|---|---|---|
| Chase shipment | Still inside promised window | Check order and reply with expected timing | Buyer is unhappy or asks for a human |
| Chase shipment | Near promised date | Reply with current status and set a reminder | No update by the promised time |
| Chase shipment | Already overdue | Comfort buyer and hand off | Immediate escalation |
| Preorder shipment | Inside preorder window | Explain the preorder window | Buyer asks to cancel or refund |
| Out of stock | Inventory exception | Collect context and hand off | Immediate escalation |
The knowledge base provides shipping policy and FAQs. Order status provides real-time evidence. The shared workspace lets AI and human agents switch in one click. When an agent corrects AI, YundaDesk turns that into a learning suggestion you confirm. It only becomes new knowledge or a new skill after review, and every step is traceable, testable, and revertible.
Pre-Launch Checks
Before using these chase-shipment scripts live, test them with real cases:
- Run 20 historical chase-shipment conversations and check whether AI identifies order status correctly.
- Make sure every answer includes a specific date or next step.
- Test out-of-stock, refund, and complaint messages to confirm handoff.
- Test English and key market languages to make sure the tone feels natural.
- Check whether channel messages merge into the same customer profile, so buyers do not repeat themselves.
If you are rebuilding shipping and delivery knowledge, read how to build a knowledge base that feeds AI support and where to draw the AI-first, human-backed boundary.
A good where is my order script reduces uncertainty: answer clearly when facts are available, set a reminder near the deadline, and escalate to humans when the order is overdue.