Home · Business · Customer Service · Tier-1 support & chatbots
As of 13 August 2026, AI can only partly let customers track their order.
Most people should hand this to a purpose-built tool.
Can you do it?
1 hourto a draft.
2 hoursto something you’d act on.
Cost, all in£0/month
Skill neededdeveloper
Who has to check ita colleague
What the alternative costsA purpose-built support platform such as Botpress is the software alternative; the supplied tool data gives no price for it.
If this goes wrong, a customer receives a false status, sees another customer's information or contacts your team after a missed delivery.
What to actually do
Use a tool built for this
The route this page recommends
Hand it to a person
Second choiceA person who owns the outcome does this end to end, worth it when the failure is dear.
Do it yourself
The distant thirdA chat interface, developer skill, and roughly 2 hours until you can act on the result.
How to actually do it
- Open your order-management and delivery systems and record the available API or webhook fields for order ID, customer verification, status, tracking link, dispatch date and estimated delivery date.
- Choose the customer verification fields and write the hand-off route for orders that cannot be found, are delayed, lost, cancelled, refunded or need changing.
- Paste the prompt into an AI chatbot and add your business name, verification fields and support contact, then ask it to produce the integration specification and conversation flow.
- Give the specification to the person configuring your chatbot platform, connect it to a test environment or sample orders, and restrict responses to the authenticated customer's order.
- Run the generated test cases against known orders, comparing every status, date and tracking link with the order-management system and carrier record.
- Send failed tests and privacy issues back to the AI for corrected fallback messages, then have a colleague approve the final flow before making it available to customers.
Prompt
Design a customer order-tracking chatbot for [BUSINESS NAME]. It must use our live order system or delivery API rather than inventing status information. For each customer, require [ORDER IDENTIFIER] and [SECOND VERIFICATION FIELD] before showing only that customer's order. Return these fields where available: order status, items, dispatch date, tracking link, estimated delivery date and the latest carrier update. If the order cannot be found, the data is stale, the delivery is delayed, the parcel is lost, the customer requests a refund or the request involves changing an order, explain the limitation and hand the conversation to a human at [SUPPORT CONTACT]. Never guess, expose private data, promise a delivery date that the source system does not provide, or claim that an order has shipped without a live source. Produce: the required integration fields, the conversation flow, authentication and privacy rules, fallback messages, test cases for normal and failed orders, and a checklist for comparing each answer with the order system before launch.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot provide live order status without a working connection to your order or delivery system.
- AI cannot decide whether a customer has supplied enough information to access an order unless you define and enforce the verification rules.
- AI cannot guarantee that a carrier's estimated delivery date is current or accurate.
- AI cannot take responsibility for exposing private order data or giving a customer a misleading delivery promise.
- AI does not remove the work of testing delayed, lost, cancelled, refunded and unrecognised orders.
What caps this at PARTLY: private data access, real time truth and stakes of error.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 1 |
| Verification | 1 |
| Liability | 1 |
| Effort delta | 1 |
| Total | 6 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI track my customers' orders?
- Partly. AI can collect an order reference, query a connected order or delivery system and explain the returned status, but it cannot track anything from a chat interface alone.
- Can I connect an AI chatbot to my order system?
- Yes, if your order system provides a suitable API, webhook or other supported connection. You still need to configure customer verification, restrict the data returned and test the connection against real system records.
- Can AI tell customers when their order will arrive?
- It can repeat an estimated delivery date supplied by your order or carrier system. It must not invent a date or present an estimate as a guarantee, and delayed or conflicting data should go to a person.
- Is it safe to let a chatbot show order details?
- Only with a proper customer verification step and a connection that returns data for the authenticated order. Test wrong order numbers, shared devices, stale records and requests for another person's information before launch.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.