'WISMO Automation: Let ''Where Is My Order?'' Answer Itself'

Open your support inbox and count how often "where is my parcel?" shows up. If you want to automate "where is my order", that is the best possible starting point: this one question makes up 30 to 50 percent of all tickets in most shops.

Across our own brands it is around 40 percent. And the strange part: the answer has been sitting in your systems all along. The customer only asks because nobody showed it to them.

In this article I show you how to automate WISMO requests. With the data sources, the reply logic and the limits that tool vendors leave out.

What WISMO means and why it is so many tickets

WISMO stands for "where is my order", the question about where the order has got to. The term comes from support jargon, because the request is so frequent that it earned its own abbreviation.

Why are there so many? Three reasons:

• The customer is nervous. They paid and are holding nothing. Every hour without information feels long.
• Tracking emails get lost. Spam folders, overflowing inboxes, typos in the email address.
• Carrier pages are unreadable. "Shipment handed over to delivery partner" does not answer the question "when will it arrive?"

The key point: WISMO is almost never a real problem. Usually the parcel is on its way and everything is running normally. The customer just wants certainty. That is exactly why WISMO is the perfect candidate for automation. The answer always has the same structure, only the data changes.

At nano we understood that early. Response time for this question: up to 24 hours before, about 3 minutes today. Around the clock, weekends included.

The two data sources you need

A good WISMO reply needs two pieces of information, from two different systems.

Source 1: your shop system

From Shopify (or your shop system) you pull:

• The order itself: what was ordered, when, to which address?
• The fulfilment status: has the order shipped or is it still in the warehouse?
• The tracking number, as soon as it exists.

That already covers half the cases. "Your order is being packed and goes out tomorrow" is a complete answer when no shipment exists yet.

Source 2: carrier tracking

For the other half you need the current shipment status from the carrier, so DHL, DPD or whoever drives for you. You pull it through the carrier's tracking API, an interface your system can query automatically.

Important: translate the carrier status into human language. "Arrived at delivery base" becomes "your parcel is at the distribution centre for your region and should arrive tomorrow." That is the real value your automation adds. The customer could look up the raw status themselves. What they want is the translation.

We run both through n8n, an automation tool that connects different systems through their interfaces. Incoming email in, order data and tracking status out, reply written.

The reply logic: recognise first, then decide

The process has three stages.

Stage 1: recognise that it is WISMO

An AI reads the incoming email and classifies it, meaning it assigns it to a category. WISMO requests are easy to spot, luckily. "Where is my parcel", "status of my order", "still nothing here": the wording is very similar every time. Our classification is reliable enough here that we stopped checking it daily.

Stage 2: find the order and check the status

The system looks up the order via the sender address or the order number. Then it decides based on the status:

• Not shipped yet: reply with the processing status and the expected shipping date.
• Shipped, in transit: reply with the translated tracking status, a link and a delivery forecast.
• Delivered according to the carrier: careful. This is no longer a standard case, more on that in a moment.

Stage 3: reply or escalate

Standard cases the system answers immediately and on its own. Everything else goes to a human. This switch is the most important part of the whole setup.

We always escalate when:

• the carrier reports "delivered" but the customer has nothing
• the shipment has shown no new status for several days
• the customer is visibly upset
• the order cannot be identified clearly

How we pre-sort requests in general, and when humans take over, is described in more detail in our setup for email support with AI.

Where automation ends: lost parcels need humans

Now the important part. WISMO automation works for the normal case. For the problem case it does not work, and it should not try.

A parcel that the tracking says was delivered but never arrived is no longer a status question. It is a trust case. The customer paid money and got nothing. If an automatic email arrives now saying "according to DHL your parcel was delivered", the situation escalates. Rightly so.

Cases like that need a human who:

• opens an investigation with the carrier, meaning an official search request for the shipment
• decides whether a replacement goes out immediately or whether you wait for the result
• gives the customer the feeling that somebody is taking care of it

Our rule: when in doubt, the replacement goes out immediately. The investigation runs in parallel. A lost parcel costs us the value of the goods. A lost customer costs more.

Second limit: bundled requests. "Where is my order, and by the way I want to change the colour" is no longer a WISMO request. As soon as several topics sit in one email, a human should answer. Otherwise the automation answers one question and ignores the other.

Third limit: the automation is only as good as your data. If your fulfiller reports tracking numbers two days late, your system answers wrongly for two days. Build the data path cleanly first, then the replies.

How to approach the build

If you want to build this yourself, here is the order that worked for us.

• Week 1: count. Tag every WISMO email for two weeks. Only with the real number do you know whether the build pays off.
• Week 2: check the data path. Does the tracking number reach your shop system reliably and fast? If not, fix that first.
• Week 3: draft the replies. Write a template for every status in your own tone. Translate every carrier status into human language once, properly.
• Week 4: start with approval. Let the system write drafts for two weeks that you check before sending. Only when the hit rate is right do you switch to automatic.
• After that: watch the escalations. Look weekly at which cases land with a human. That is where you learn where your shipping process actually breaks.

The last point is underrated. Our escalation list has become an early warning system on the side. When stuck shipments pile up at one carrier, we see it days before the complaints arrive.

What this actually delivers

Let's run the numbers, with rough orders of magnitude instead of invented precision. Say you get 100 support requests a week. 30 to 50 of them are WISMO. Each one takes you a few minutes manually: read the email, find the order, check the tracking, reply.

That is several hours a week for a question whose answer sits in your systems. After automation what remains are the problem cases, a small fraction.

The second effect matters almost more to me: response speed. A customer who gets a precise answer after 3 minutes writes no second email, no third, and no frustrated comment under your ads. If you want to go deeper: the big picture is in our guide to automating customer service, and where WISMO sits within your overall operation is shown in the overview of e-commerce processes.

Conclusion

Automating WISMO requests is the entry point with the best ratio of effort to effect. The question is frequent, the answer is structured, the data already exists. You need two data sources, a clean switch for problem cases, and the discipline to hand lost parcels to humans every single time.

Don't start with all of customer service. Start with this one question. Once it answers itself, you have cleared the biggest block of tickets.

If you want to know what a setup like this would look like for your shop: get in touch with Flowhouse. We build exactly this, for our own brands too.