Open any ecommerce support inbox and count the messages. A big share of them say some version of the same thing: "Where's my order?" It's the most repetitive question in online retail, it rarely needs a human's judgment, and it eats hours that could go to customers with real problems.
The frustrating part is that the answer almost always exists already. The order shipped, there's a tracking number, and the status is sitting in your system. The customer just can't get to it easily, or doesn't know how long to expect, so they message support, and a person copies and pastes a tracking link or a delivery timeline. A chatbot on your website can't see your order system, but it can answer the timeline and where-to-find-tracking part instantly, all day, and collect the order number for the rest.
Why order-status questions pile up
Customers ask about their orders because the waiting is anxious and the information is hidden. They bought something, they're excited or nervous about it, and the status lives behind an account login they've forgotten or in an email they can't find. Asking a human feels easier than hunting.
The volume spikes at the worst times, too. After a sale, during the holidays, or when a shipment is delayed, everyone wants an update at once. That's exactly when your team is already stretched, so response times slow, which makes more people message to ask why no one's replying. It feeds on itself.
What a chatbot actually does with these
The mechanics are straightforward. A store-integrated bot identifies the order and looks up its status. A website assistant like SpideyChat doesn't connect to your store, so it works one step earlier: it answers from your published shipping policy and help pages, and passes the specific lookups to a person.
Here's the typical flow:
- The customer asks about their order.
- The bot explains your dispatch and delivery times and where the tracking link is sent.
- If the order is still inside the normal window, that usually settles it.
- If it isn't, or the customer wants a specific check, the bot collects the order number and email.
- It opens a ticket or hands over to a person with those details attached.
For many customers whose orders are simply in transit, steps two and three are the whole conversation. They hear "orders ship within two days and arrive three to five days later, and your tracking link is in the confirmation email," and they're satisfied in ten seconds. No ticket, no wait, no agent time.
Handle the exceptions, don't hide them
The mistake to avoid is a bot that treats every order the same. Most orders are fine, but the ones that aren't are exactly where a customer needs a person, and a bot that keeps chirping "it's on its way" about a package lost two weeks ago makes things worse.
Build in the exceptions:
- An order that's overdue or hasn't moved in days should route to a human.
- A tracking status of "delivered" when the customer says they didn't get it needs a person.
- A request that shifts from status to refund or complaint should hand off.
- Anything the bot can't confidently identify should collect details and escalate.
The bot's job is to clear the routine 80 percent and flag the messy 20 percent, not to stonewall people whose orders actually went wrong. Done right, your team stops answering "where is it" and starts spending that time on the genuinely stuck shipments that need a human touch.
A quick example
Take a fictional home goods store, Marigold & Oak. During their fall sale, support volume tripled overnight, and most of it was order-status checks. Their two-person support team fell a full day behind, so customers who had real issues, like a damaged item, waited a day for a reply behind a wall of routine questions.
They added a chatbot trained on their shipping policy, their sale dispatch times and the help page explaining tracking emails. Within a week the pattern shifted. Most "when will it arrive" checks got answered instantly and never became tickets, and the rest arrived with the order number attached. The two humans went from drowning in "where's my order" to handling the damaged items, the address changes, and the genuinely late shipments, which were the conversations that actually needed them. Response time on real issues dropped from a day to under an hour, not because they hired anyone, but because the noise stopped reaching the inbox.
Here's the rough before and after in shape, framed as illustration rather than a promise:
| Before the bot | After the bot | |
|---|---|---|
| Delivery-time questions | Handled by a person, one by one | Answered instantly, no ticket |
| Real issues | Stuck in the queue behind status checks | Reached a human quickly |
| Team focus | Copy-pasting tracking links | Solving actual problems |
| Nights and weekends | No coverage | Bot answers timelines anytime, takes order details |
What it takes to set up well
A WISMO bot is only as good as the information it has and the honesty of its answers. A few things make the difference between one that helps and one that frustrates.
First, give it accurate, specific policy content. SpideyChat doesn't connect to your store or carrier, so it can't look up an individual order. What it can do is quote your real dispatch cut-offs, delivery windows by region and where the tracking email comes from, which answers more of these questions than you'd expect, because many customers are really asking "is this normal?" rather than "where exactly is it?"
Second, make the handoff easy. When a customer needs a specific check, have the bot ask for the order number and accept the checkout email as a fallback, because plenty of customers don't have the number handy. The cleaner the details, the faster your team can answer.
Third, be honest about delays. If an order is past your stated window, the bot should say so plainly and pass it to a person, not paper over it with a cheerful non-answer. Customers can handle bad news; they can't handle being ignored.
Look past status to the whole post-purchase experience
Once the bot handles the timing questions, the same setup can absorb the questions that cluster around it. Returns, exchanges, "can I change my address," "how do I cancel." These live in the same post-purchase moment and follow the same pattern: routine rules the bot can explain, with a handoff for the odd cases.
Start with delivery timing because it's the loudest and easiest win, then widen from there as you see what else people ask. Read the transcripts, notice the next most common question, and give the bot an answer for it.
There's a customer-experience upside beyond the ticket count, too. People often ask "where's my order" not because they're impatient but because they're anxious, and a slow reply makes the anxiety worse. An instant, accurate answer about what's normal calms that in the moment it strikes, which quietly builds trust in your brand. The same automation that saves your team time also happens to give customers the fast, self-serve answer they actually prefer over waiting on a person.
The result is a support team that isn't buried under the same question a hundred times a day, and customers who get an instant answer instead of joining a queue. The order was always trackable. You're just making sure people know how long to expect and where their tracking link is, at the moment they care, and that the ones who need a person reach one with the details ready.