How to Automate Where Is My Order (WISMO) Responses with AI
"Where is my order?" is the most common customer support question in e-commerce. It's also one of the most automatable. With the right setup, an AI agent can read the ticket, extract the order number, look up the shipping status via API, and draft a customer-friendly response — all without a human touching the ticket.
What is a WISMO ticket?
WISMO stands for "Where Is My Order?" — the customer wants to know the shipping status of something they purchased. These tickets typically include an order number (sometimes in the message body, sometimes in a custom Zendesk field) and a straightforward question: when will it arrive?
For most e-commerce support teams, WISMO tickets account for 30–50% of total volume. Each one follows the same pattern: look up the order, check the carrier status, write a response. It's repetitive, time-consuming, and perfectly suited for automation.
The automation flow
A fully automated WISMO response involves four steps:
Step 1 — Read the ticket and extract the order number
The AI agent reads the Zendesk ticket — subject line, description, and custom fields. Order numbers often live in a custom field (like "Bestellnummer" or "Order ID") rather than in the message body. With Macha's custom-fields tool, the agent can read these directly without the customer needing to type the number.
Step 2 — Look up the order via your shop's API
Using a custom API tool, the agent calls your order-management or shipping-status endpoint with the extracted order number. The tool returns the carrier name, tracking events, scheduled delivery date, and tracking link if available.
Custom tools in Macha are configured once (URL template + parameters + auth) and reused across agents. No code required — just the API endpoint and a parameter mapping.
Step 3 — Apply business rules
Not every WISMO response is the same. The agent applies rules based on the data:
- On time or slight delay (1–2 days) — standard status update with tracking info
- Significant delay (3+ days) — apology + compensation offer (e.g., voucher code)
- Already delivered — suggest checking with neighbours or building management
- Still in production — explain the manufacturing timeline and when tracking will be available
Step 4 — Respond and tag
The agent posts the response — either as a public reply (for fully automated flows) or as an internal note (for human-reviewed pilot phases). It also applies tags for reporting: ai_agent_handled, wismo_only, ai_response_internal_note, etc.
Setting it up in Macha
You need three things:
- A Zendesk connector with the right tools enabled — Get Ticket, Get Custom Fields, Add Internal Note (or Add Public Reply)
- A custom API tool pointing at your shipping-status endpoint — just the URL template and one parameter (order number)
- A trigger on "New Ticket" or "Every Customer Message" that fires your WISMO agent
The agent's instructions define the classification logic (is this actually a WISMO question?), the business rules (what to say based on the status), and the response format.
Phased rollout
We recommend a two-phase approach:
- Phase 1 — Internal notes only. The agent reads the ticket, looks up the order, and drafts a response as an internal note. A human reviews it, copies it to a public reply if it's good, or edits before sending. This lets you validate quality on real tickets before going fully autonomous.
- Phase 2 — Public replies. Once you've seen 50+ internal-note drafts that you'd have sent as-is, switch to direct public replies. The agent responds to customers automatically.
Results you can expect
Teams using WISMO automation with Macha typically see:
- Response time: from hours to under 60 seconds
- Deflection rate: 40–70% of WISMO tickets fully automated
- Agent time saved: 2–4 hours per agent per day on high-volume accounts
