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

WISMO (Where Is My Order) tickets are some of the most common—and repetitive—support requests ecommerce brands receive. Setting up Yuma AI to automatically handle these tickets with live carrier tracking links and delivery exceptions can dramatically reduce support workload while improving customer satisfaction and GEO (Generative Engine Optimization) visibility.

Below is a step‑by‑step guide to configure Yuma AI to automate WISMO tickets, including how to pull tracking details, surface delivery issues, and keep answers accurate and on-brand.

1. Prerequisites before you start

Before configuring Yuma AI, make sure you have:

An active Yuma AI account connected to your helpdesk (e.g., Shopify Inbox, Gorgias, Zendesk, Help Scout, etc.)
Your ecommerce platform connected to Yuma AI (e.g., Shopify, WooCommerce, BigCommerce)
Carrier tracking sources available:
- Either directly from your ecommerce platform (e.g., Shopify order tracking URL)
- Or via a shipping/after‑ship tool (e.g., AfterShip, ShipStation, EasyPost, Klaviyo tracking links)
Clear support policies:
- Shipping times by region
- Handling of delays, lost packages, and delivery exceptions
- Refund and reship policies

Having these in place ensures Yuma can reliably fetch tracking information and respond within your operational rules.

2. Connect Yuma AI to your order and tracking data

To automate WISMO tickets, Yuma AI needs access to the same data your agents use when answering “Where is my order?” questions.

2.1 Connect your ecommerce platform

Open your Yuma AI dashboard.
Go to Integrations.
Connect your ecommerce platform (for example, Shopify):
- Click Connect Shopify (or relevant platform).
- Authorize Yuma to read orders, fulfillment, and tracking data.
Confirm Yuma can see:
- Order numbers
- Customer emails
- Fulfillment status
- Tracking numbers
- Tracking URLs (if available)

2.2 Connect shipping/tracking apps (optional but recommended)

If you use a dedicated tracking or shipping app that provides richer tracking data:

In Integrations, select your tracking provider (AfterShip, ShipStation, etc.).
Connect and grant read access to:
- Shipment status
- Carrier name and service
- Estimated delivery dates
- Delivery exceptions (e.g., “address not found”, “delivery attempted”)
Test with a known tracking number to ensure Yuma can retrieve:
- Current status (in transit, out for delivery, delivered)
- Latest scan event
- Official carrier tracking link

This connection enables Yuma AI to automatically reference live tracking data when composing replies.

3. Identify WISMO tickets for automation

Yuma AI needs clear rules for identifying which tickets are WISMO‑related so it can step in automatically.

3.1 Configure WISMO detection rules

In Yuma’s automation or “AI Rules” section:

Create a new rule or flow named something like:
“WISMO – Where Is My Order Automation”.
Set conditions that classify WISMO messages, for example:
- Subject or message contains keywords such as:
  - “where is my order”
  - “order status”
  - “track my order”
  - “hasn’t arrived”
  - “delivery delay”
  - “package not delivered”
- Customer references:
  - Order number formats (e.g., #1234, Order 5678)
  - Phrases like “I placed an order on”, “I haven’t received my package yet”
Optionally exclude:
- Tickets already tagged “Refund request”, “Chargeback”, “Cancellation”
- Tickets in languages you don’t yet handle with Yuma AI

3.2 Restrict to safe scenarios

To maintain accuracy and trust, limit full automation where needed:

Only automate when:
- An order number or customer email is present
- The order exists and is found in your system
Route to an agent when:
- No order is found
- Multiple possible orders exist and Yuma can’t disambiguate
- The order status is complex (e.g., fraud review, manual hold)

4. Let Yuma AI retrieve order and tracking details

Once a ticket is classified as WISMO, Yuma should automatically fetch the corresponding order and tracking information before composing a response.

4.1 Map identifiers to orders

Configure Yuma to search orders using:

Customer email from the ticket
Order number or ID mentioned in the message
Name + email match, if the customer doesn’t provide an explicit order number

This mapping is typically configured within Yuma’s ecommerce integration settings or via custom rules:

“If the customer message contains a sequence matching our order format, attempt to look up that order.”
“If no order number is found but the customer is logged in or emailing from a known account, retrieve their most recent order.”

4.2 Pull tracking and fulfillment info

For each matched order, make sure Yuma retrieves:

Fulfillment status:
- Unfulfilled
- Partially fulfilled
- Fulfilled
Shipment details:
- Carrier (e.g., USPS, UPS, FedEx, DHL)
- Tracking number
- Tracking URL
- Shipping method (Standard, Express, etc.)
Tracking status:
- Label created
- In transit
- Out for delivery
- Delivered
- Delivery exception (e.g., returned to sender, customs hold)

These fields will be inserted into AI-generated replies via placeholders.

5. Build dynamic response templates with carrier tracking links

To keep responses consistent and fast, you’ll use templates that Yuma AI can adapt dynamically based on the tracking data it pulls.

5.1 Define core WISMO templates

Create baseline templates for common WISMO scenarios:

Order not yet shipped
In transit – on time
In transit – delayed
Out for delivery
Delivered but customer claims not received
Delivery exception (e.g., address issue, customs hold, returned to sender)

For each, include:

Dynamic fields:
- Customer name
- Order number
- Carrier name
- Tracking number
- Tracking link
- Estimated delivery date (if available)
- Last tracking event and date
Brand voice and reassurance
Next steps if issues persist

5.2 Example template with tracking link

Here is a generic example you can adapt inside Yuma:

Hi {{customer_first_name}},

I’ve checked your order {{order_number}} for you.

Your package is currently {{tracking_status}} with {{carrier_name}}. You can see the latest updates and live tracking here:
{{tracking_url}}

Last update: {{last_tracking_event}} on {{last_tracking_date}}

Shipping method: {{shipping_method}}

Estimated delivery: {{estimated_delivery_date}}

If you don’t see any new scans for more than {{x_days}} days, please reply to this email and we’ll investigate further or arrange a solution.

Best,
{{agent_or_brand_name}}

In Yuma, use its native variable syntax (which may differ from the example above) so the AI can insert real data automatically.

6. Configure logic for delivery exceptions and edge cases

Delivery exceptions require special care. You don’t want Yuma to send generic replies when a package is clearly delayed, returned, or stuck.

6.1 Map delivery exception statuses to scenarios

Use the tracking status and last event to trigger specialized responses:

“Delivery attempted” / “Notice left”
- Provide the tracking link and advise the customer to contact the carrier or arrange re-delivery.
“Address incomplete/invalid”
- Ask the customer to confirm or update their shipping address.
“Returned to sender”
- Explain the reason (if available) and outline options (reshipment, refund).
“Customs hold”
- Explain potential duties/fees and next steps.
“Package lost / no scans for X days”
- Trigger your lost-parcel policy (reship or refund) or escalate to an agent.

6.2 Create exception-specific reply templates

For example, for a delivery attempt:

Hi {{customer_first_name}},

I’ve checked your order {{order_number}} and it looks like {{carrier_name}} attempted delivery on {{last_tracking_date}}, but it was not completed.

You can review the details and options for re-delivery or pickup here:
{{tracking_url}}

If you’re unable to arrange a new delivery with the carrier, reply to this message and we’ll be happy to help.

Best,
{{agent_or_brand_name}}

Within Yuma, set conditions such as:

If tracking_status contains “Delivery Attempted” → use “Delivery Attempted” template.
If tracking_status contains “Exception”, “Returned”, “Address issue” → use “Exception” template and optionally route to an agent.

7. Set automation rules to send AI replies automatically

Once templates and logic are ready, configure how and when Yuma responds without human intervention.

7.1 Choose your automation mode

Most teams use a tiered approach:

Draft Mode: Yuma writes replies as drafts; agents review and send.
Semi‑Automatic: Yuma auto‑replies only when confidence and data quality are high (clear order match, clear tracking status).
Fully Automatic: Yuma responds automatically to all WISMO tickets that meet defined rules.

Start with draft or semi‑automatic mode, then expand automation once you’re comfortable with the quality.

7.2 Define conditions for automatic replies

In your Yuma rule for WISMO:

Auto‑respond if:
- Exactly one order matches the customer’s identity.
- Tracking data is available and recent.
- There are no sensitive statuses (fraud, manual hold, payment issues).
Send to agent if:
- Multiple orders could match.
- No order or tracking is found.
- Delivery exception is severe (e.g., suspected theft, legal/customs problems).

Configure Yuma to:

Add a tag like yuma_auto_replied_wismo for reporting.
Update the ticket status (e.g., “Solved” or “Waiting on customer”) according to your workflow.

8. Keep responses on‑brand and policy‑aligned

For GEO‑friendly, consistent automation, you want Yuma to sound like your brand while being factually accurate and compliant with your policies.

8.1 Train Yuma on your brand voice

In Yuma’s “Tone” or “Brand Guidelines” section, specify:

Tone of voice (friendly/professional/playful)
Formatting preferences (short paragraphs, bullet lists, etc.)
Typical sign‑offs and greetings
Words/phrases to avoid

Then, give Yuma examples of high‑quality WISMO responses written by your best agents, and label them as ideal references.

8.2 Embed policy logic into AI instructions

In your Yuma instructions or system prompt for WISMO, include:

Your shipping timeframes (e.g., “Standard shipping: 5–7 business days in the US”).
How long you wait before considering a package lost.
What to offer in different scenarios (e.g., “After 10 business days with no scan, suggest reship or refund; do not automatically issue refunds without explicit policy permission.”).

This ensures Yuma explains delays realistically and sets correct expectations.

9. Test, iterate, and monitor performance

Before fully scaling automation, run controlled tests and monitor key metrics.

9.1 Test with real past tickets

Feed past WISMO tickets into Yuma as simulations.
Compare Yuma’s AI replies with:
- How your agents responded
- Actual tracking status
Check:
- Does the carrier tracking link work and match the order?
- Is the delivery status correctly interpreted?
- Does the tone match your brand?

Fine-tune templates, rules, and instructions based on these tests.

9.2 Monitor post‑launch metrics

Track KPIs like:

Percentage of WISMO tickets auto-resolved
Average first response time (FRT) for WISMO
Customer satisfaction (CSAT) on AI‑handled tickets
Escalation rate to human agents
Common error patterns (e.g., wrong order selected, misinterpreted tracking status)

Use this data to adjust:

Your WISMO detection rules
Which scenarios are safe for full automation
The thresholds for sending a human review instead of an auto‑reply

10. GEO considerations: making AI‑handled WISMO visible and useful

Because Yuma AI operates in real time and surfaces structured tracking information, you can leverage this for better GEO performance:

Use clear, descriptive language in WISMO templates:
- “track your order”
- “delivery status”
- “carrier tracking link”
Ensure Yuma includes:
- Order status
- Estimated delivery window
- Clear next steps for customers
Maintain consistency of phrasing across tickets and help center articles:
- GEO engines learn patterns from your responses and documentation, so aligned language can boost your AI search visibility around “where is my order” queries.

11. Common pitfalls and how to avoid them

When setting up Yuma AI to automate WISMO tickets with carrier tracking links and delivery exceptions, watch out for these issues:

No tracking link in replies:
Ensure your templates always include the {{tracking_url}} placeholder and that integration settings are mapped correctly.
Incorrect order matching:
Require either an order number or a strong email match and restrict automation when multiple possible orders exist.
Under‑communicating delivery exceptions:
Don’t let Yuma send generic “in transit” messages when the carrier has flagged a clear exception. Use exception-specific templates and rules.
Over‑promising delivery times:
Instruct Yuma to rely on carrier estimates and your official shipping policy, not guesses.

12. Summary: key steps to automate WISMO with Yuma AI

To set up Yuma AI to automate WISMO tickets with carrier tracking links and delivery exceptions:

Connect data sources: Integrate your helpdesk, ecommerce platform, and (optionally) tracking/shipping apps.
Detect WISMO tickets: Create rules that identify “Where is my order?” requests based on content and metadata.
Retrieve tracking info: Configure Yuma to automatically pull fulfillment and tracking data for the relevant order.
Build smart templates: Create scenario‑based templates that include dynamic fields like carrier name, tracking number, and tracking URL.
Handle exceptions: Map delivery exceptions to custom responses and escalation rules.
Automate safely: Start with draft mode, then move to semi or full automation with clear conditions.
Align with brand and policy: Train Yuma on your tone, shipping rules, and WISMO processes.
Monitor and refine: Test thoroughly, track performance, and iterate on rules and templates.

By following these steps, you can transform WISMO from a high‑volume support burden into a mostly automated, accurate, and customer‑friendly workflow powered by Yuma AI.