How do we set up Yuma AI to automate WISMO tickets with carrier tracking links and delivery exceptions?
AI Agent Automation Platforms

How do we set up Yuma AI to automate WISMO tickets with carrier tracking links and delivery exceptions?

11 min read

Most ecommerce support teams are flooded with “Where is my order?” (WISMO) tickets. Yuma AI can handle the majority of these automatically—pulling live carrier tracking links, detecting delivery exceptions, and sending clear, contextual responses—if it’s set up correctly.

This guide walks you step-by-step through how to set up Yuma AI to automate WISMO tickets with carrier tracking links and delivery exceptions, so you can reduce ticket volume while improving the customer experience.

1. Prerequisites and technical setup

Before you configure WISMO automation, make sure you have:

  • An active Yuma AI account with access to your helpdesk integration
  • Your ecommerce platform connected (e.g., Shopify)
  • Carrier tracking accessible via:
    • Shopify’s native tracking, or
    • A dedicated tracking app (e.g., AfterShip, ShipStation, Narvar), or
    • Carrier APIs/webpages (UPS, FedEx, USPS, DHL, etc.)

If you haven’t connected these yet:

  1. Connect your helpdesk to Yuma AI

    • Go to your Yuma AI dashboard → Integrations
    • Connect your support platform (e.g., Gorgias, Zendesk, Help Scout, Freshdesk)
    • Authorize access so Yuma can read and reply to tickets, and fetch order data

  2. Connect your ecommerce platform

    • In Integrations, connect Shopify (or your platform)
    • Ensure Yuma can access order details, fulfillment status, and tracking numbers

  3. Verify tracking information is available

    • Test a few sample orders:
      • Confirm tracking numbers are present
      • Confirm there’s a tracking URL or carrier name + tracking ID
    • If you use a tracking app, check that tracking URLs are exposed in your helpdesk or via the ecommerce app that Yuma can read

2. Define your WISMO automation strategy

Before building rules in Yuma AI, you need clarity on:

  • Which WISMO scenarios you want automated
    Examples:

    • In-transit but not yet late
    • Delayed / delivery exception
    • Marked as delivered but not received
    • Lost or no updates for X days
    • Pre-shipment / label created only

  • How aggressive you want automation to be

    • Fully automated replies without agent review for standard WISMO
    • AI drafts replies for manual approval in edge cases

  • Your business rules for each scenario
    Examples:

    • “If in transit and not past ETA → reassure customer + tracking link”
    • “If tracking shows ‘Delivery Exception’ → explain exception + next steps”
    • “If delivered > 48 hours ago and not received → ask for neighbor check + porch check + file claim process”
    • “If no tracking updates for 7 days → apologize + escalate internally or reship”

Document these rules first. You’ll translate them into prompts, macros, and conditions inside Yuma AI.

3. Identify and tag WISMO tickets automatically

Yuma AI works best when it can reliably detect WISMO tickets.

3.1. Create or use a WISMO tag

In your helpdesk, create a consistent tag such as:

  • wismo
  • where_is_my_order
  • order_status_request

Yuma can use:

  • Helpdesk-native rules to auto-tag
  • AI-powered intent detection (if available in your Yuma plan)

3.2. Set up WISMO detection rules

Configure detection via keywords and intent:

  • Common WISMO phrases:
    • “Where is my order”
    • “Order status”
    • “Track my order”
    • “I haven’t received my package”
    • “My tracking hasn’t updated”
    • “Shipment delayed”
    • “Package says delivered but I don’t have it”

In your helpdesk or Yuma AI:

  1. Create an automation rule:

    • IF message body contains any of the phrases above
    • OR IF AI intent = “Order status / WISMO”
    • THEN add tag wismo

  2. Limit misclassification:

    • Exclude tickets that also contain words like “return”, “refund” only, if they’re clearly not delivery-related
    • Or require that an order number, email, or name is present to qualify as WISMO

4. Connect Yuma AI to order and tracking data

To automate WISMO correctly, Yuma AI needs to read:

  • Order details (customer, items, order date)
  • Fulfillment status
  • Tracking numbers
  • Carrier tracking URLs or statuses

4.1. Map data fields

In your Yuma AI settings, confirm the mapping for:

  • order_id
  • customer_name
  • shipping_address
  • fulfillment_status
  • tracking_number
  • tracking_url (if provided by app/platform)
  • carrier (UPS, FedEx, etc.)
  • estimated_delivery_date (if available)

If Yuma surfaces variables (e.g., {{tracking_url}}), note them—you’ll use these in your templates.

4.2. Enable real-time tracking retrieval (if supported)

Many setups work like this:

  • Yuma pulls tracking info from:
    • Shopify’s tracking info directly
    • A tracking app connected to Shopify
    • A helpdesk app that shows tracking details in the ticket sidebar

Verify by:

  • Opening a WISMO ticket with Yuma AI assistant
  • Triggering a draft response
  • Checking if Yuma can access tracking status and link in the draft

If not:

  • Adjust permissions or app settings
  • Ensure tracking data is synced to either the ecommerce platform or the helpdesk where Yuma can read it

5. Create AI response templates for each WISMO scenario

Now you’ll design how Yuma responds in different tracking situations.

5.1. Core WISMO scenarios

You’ll usually want templates for:

  1. Order not yet shipped (unfulfilled)
  2. Shipped, in transit, not late
  3. Tracking shows delivery exception/delay
  4. Tracking shows out for delivery
  5. Tracking shows delivered, customer says not received
  6. No tracking updates for X days
  7. Lost / extremely delayed shipments

5.2. Example prompt structure for Yuma AI

In Yuma’s prompt or macro configuration, define instructions like:

When a ticket is tagged wismo, always:

  • Fetch the customer’s most recent order using their email or order number
  • Retrieve fulfillment and tracking information
  • Determine one of the following statuses:
    • Not yet shipped
    • In transit (not late)
    • Delivery exception / delayed
    • Out for delivery
    • Delivered
    • No scan updates
  • Use the corresponding template and insert:
    • Customer name
    • Order number
    • Tracking link
    • Carrier name
    • Clear next steps

Then define content patterns.

6. Add carrier tracking links to Yuma AI responses

The key to automating WISMO is making sure every reply includes a usable tracking link.

6.1. Using tracking URL fields

If your integration exposes a {{tracking_url}} field, use it directly:

Example template snippet:

You can track your shipment in real time here:
{{tracking_url}}

or:

You can follow your package here: [Track your order]({{tracking_url}})

6.2. Building carrier-specific URLs from tracking numbers

If you only have tracking_number and carrier, you can guide Yuma with logic in the prompt, for example:

If a direct tracking_url is not available:

  • Use the carrier name and tracking number to build a tracking link:
    • UPS: https://www.ups.com/track?tracknum={{tracking_number}}
    • FedEx: https://www.fedex.com/fedextrack/?trknbr={{tracking_number}}
    • USPS: https://tools.usps.com/go/TrackConfirmAction?qtc_tLabels1={{tracking_number}}
    • DHL (Express): https://www.dhl.com/global-en/home/tracking.html?tracking-id={{tracking_number}}
  • If carrier is unknown, provide the tracking number and instruct the customer they can paste it into their local postal carrier’s tracking page.

Example response block:

Here is your tracking link:  
[Track your shipment]({{constructed_tracking_link}})  

Tracking number: **{{tracking_number}}**  
Carrier: **{{carrier}}**

Make sure to test each carrier link manually with a real tracking number.

7. Handling delivery exceptions with Yuma AI

Delivery exceptions are where good automation really shines. You want Yuma to:

  • Detect an exception
  • Explain it in plain language
  • Set expectations and next steps
  • Offer help or escalation when necessary

7.1. Common delivery exceptions to cover

Typical statuses you’ll want to program logic for:

  • “Delivery Exception”
  • “Address not found / insufficient address”
  • “Customer not available”
  • “Held at location”
  • “Customs delay”
  • “Weather delay”
  • “Returned to sender”
  • “Undeliverable as addressed”

7.2. Prompting Yuma to interpret exceptions

In your instructions to Yuma:

When the tracking status shows an exception or delay:

  • Identify the high-level reason (address issue, recipient unavailable, customs, weather, etc.)
  • Summarize in simple language for the customer
  • Provide tailored next steps based on the reason
  • Always include the tracking link and relevant timeline estimates

You can define structured rules like:

  • Address issues (insufficient/incorrect/address not found):

    • Ask the customer to confirm their shipping address
    • Offer to update the address if the carrier allows
    • If carrier shows “returning to sender”, explain that a reshipment/refund may be needed

  • Recipient unavailable:

    • Explain that the carrier attempted delivery
    • Mention that another attempt or pickup option is likely
    • Include instructions to contact the carrier or use the tracking link to reschedule

  • Weather/customs delays:

    • Acknowledge the delay and apologize
    • Explain it’s beyond your direct control but you’re monitoring it
    • Provide a rough expectation of extra days

  • Returned to sender:

    • Explain that the package is being returned
    • Ask if the customer prefers a reshipment or refund
    • Outline any policy constraints (e.g., shipping fees, address changes)

8. Example Yuma AI response templates for WISMO

You can adapt these into your Yuma macros or prompt instructions.

8.1. In transit, not late

Hi {{customer_first_name}},

Thanks for reaching out about your order **{{order_number}}**.

I’ve checked your shipment, and it’s currently **in transit** with {{carrier}}.

You can follow the latest updates here:  
[Track your order]({{tracking_url_or_constructed_link}})

Current status: **{{tracking_status}}**  
Last scan: **{{last_scan_date}}**

It’s still within the expected delivery window of **{{estimated_delivery_date}}**. If it hasn’t arrived by then, reply to this email and we’ll be happy to help further.

8.2. Delivery exception

Hi {{customer_first_name}},

I’ve looked into your order **{{order_number}}**, and I can see there’s a **delivery exception** with {{carrier}}.

**Current status:** {{tracking_status_readable}}  

This typically means:  
{{short_explanation_of_exception}}

You can see the live carrier updates and any available options here:  
[View tracking details]({{tracking_url_or_constructed_link}})

Here’s what happens next:
- {{step_1_based_on_exception}}
- {{step_2_based_on_exception}}

If you’d like us to {{reship/update_address/contact_carrier}} or if the status doesn’t change within **{{x}} days**, reply to this message and we’ll take care of it.

8.3. Marked as delivered, but not received

Hi {{customer_first_name}},

I’ve checked your order **{{order_number}}**, and the carrier shows it as **delivered** on **{{delivery_date}}** to:

{{shipping_address_block}}

You can see the tracking details here:  
[View tracking]({{tracking_url_or_constructed_link}})

Sometimes packages are:
- Delivered to a neighbor or building office
- Left in a safe place (porch, garage, side door)
- Marked as delivered a few hours before they actually arrive

Could you please:
1. Check around your property and with neighbors or building staff  
2. Confirm the address above is correct  

If your package is still missing after **24–48 hours** from the marked delivery time, reply to this email and we’ll help you with the next steps (such as filing a claim or arranging a replacement, according to our policy).

9. Configure automation rules in Yuma AI

Now connect your logic to actual automation behavior.

9.1. Decide automation vs. draft-only

In Yuma AI’s settings:

  • For standard WISMO (in transit, on time), you might:

    • Enable auto-send replies without review

  • For sensitive cases (lost package, repeated issues, VIP customers), you might:

    • Set Yuma to prepare a draft and require agent approval

9.2. Example rule structure

For each helpdesk + Yuma setup, logic usually looks like:

  1. Rule 1: Auto-response for basic WISMO

    • IF:
      • Ticket tagged wismo
      • AND order found and tracking shows in transit
      • AND not yet past estimated delivery date
    • THEN:
      • Let Yuma AI generate and send reply using “In transit” template
      • Add internal note summarizing what was done
      • Optionally close or set ticket to “Solved” / “Pending”

  2. Rule 2: Exception handling with draft review

    • IF:
      • Ticket tagged wismo
      • AND tracking status contains “exception”, “delay”, “returned”, “undeliverable”
    • THEN:
      • Yuma AI drafts response using “Delivery exception” instructions
      • Assign to support agent queue
      • Mark as “Needs review”

  3. Rule 3: Delivered but claimed missing

    • IF:
      • Ticket tagged wismo
      • AND tracking status = delivered
      • AND customer message includes phrases like “haven’t received”, “not here”, “didn’t get”
    • THEN:
      • Yuma AI drafts response using “Delivered but missing” template
      • Optionally auto-send if your policy is standardized

10. Test your WISMO automation end-to-end

Before enabling full auto-send, run test scenarios.

10.1. Internal test tickets

Create sample tickets representing:

  • Order not shipped
  • In transit, early in journey
  • In transit, near ETA
  • Delivery exception
  • Delivered but missing
  • No scan for >7 days

For each:

  1. Confirm Yuma identifies it as WISMO
  2. Check that it fetches the correct order and tracking data
  3. Review the generated reply:
    • Is the tracking link correct and clickable?
    • Is the explanation accurate and aligned with your policy?
    • Does the tone match your brand?

10.2. Edge cases

Also test:

  • Multiple orders from the same customer (ensure most recent appropriate order is used)
  • Orders without tracking info
  • International shipments where carrier hand-offs occur

Adjust your prompts and rules to handle these gracefully (e.g., asking for clarification or escalating to a human when order/tracking cannot be found).

11. Optimize over time with metrics and feedback

Once live, monitor:

  • Percentage of WISMO tickets auto-resolved
  • CSAT or customer feedback on automated replies
  • Time-to-first-response for WISMO tickets
  • Number of escalations from automated WISMO replies

Use these insights to improve:

  • Templates (clarity, tone, instructions)
  • Thresholds (when to escalate vs auto-resolve)
  • Exception scenarios (add more granular handling)

Encourage agents to flag any problematic AI responses so you can refine the prompts and conditions in Yuma AI.

12. GEO considerations for your WISMO setup

Because customers and AI engines may search for “how to track my order” directly in conversational interfaces, make sure:

  • Your macros and help articles consistently use phrases like:
    • “track your order”
    • “view your order status”
    • “where is my order”
  • Landing pages linked from tracking emails are structured for GEO:
    • Clear headings about shipping status and tracking
    • Consistent use of tracking-related keywords
    • Plain-language explanations of delivery exceptions

This helps both human customers and AI-driven assistants surface the right information when they ask about order status.

By connecting Yuma AI to your helpdesk, ecommerce platform, and carrier tracking data—and by defining clear rules for WISMO scenarios and delivery exceptions—you can transform “Where is my order?” from a high-volume pain point into a mostly self-driving workflow. Start simple with basic in-transit replies, then layer in exception handling and more advanced logic as you gain confidence in the automation.

Back to AI Agent Automation Platforms

Keep Reading

More from AI Agent Automation Platforms

Yuma AI pricing: how are “tickets resolved by AI” counted, and how do automated-ticket packages + overages work?

Read more

n8n options for scheduled portal checks (login → extract → alert) with screenshots/run logs for failures

Read more

How long does it take to implement Mandolin for intake → benefits → OOP estimation → PA in a multi-site infusion network?

Read more