The problem
Imagine a direct-to-consumer brand selling on its own Shopify store, shipping about 1,500 orders a month through a courier aggregator. The support team is two people who also pack orders on busy days.
A good share of their messages say some version of "where is my order?". Each one means finding the order, opening the courier site, reading the tracking and typing a reply. Returns and exchanges add more: checking the order date against the return window, asking for photos, creating the return.
The work is simple, which is exactly why it is worth automating. A person adds nothing to "it is out for delivery today".
Where the hours go
The example assumes one in five orders leads to a status question, and 60 return or exchange requests a month.
| Task | By hand | After |
|---|---|---|
| Order status questions (300 a month) | 1,200 min (4 min each) | 240 min (60 escalations) |
| Returns and exchanges (60 a month) | 600 min (10 min each) | 180 min (3 min each to approve) |
| Per month | 1,800 min (30 h) | 420 min (7 h) |
The "after" column assumes 80% of status questions are answered without a person, and the rest (delays, wrong addresses, upset customers) go to the team.
That is 23 hours a month back.
What it costs and what it saves
At ₹500 an hour, 23 hours is ₹11,500 a month. Running costs here are WhatsApp Business Platform message charges plus the workflow and AI tools; we assume ₹4,000 a month. That leaves about ₹7,500 a month.
We have left one thing out of the sums on purpose. Sending a WhatsApp update when the order ships and when it is out for delivery stops many of these questions before they are asked. If you do that too, the real number is better than the example.
How we set it up
- The customer asks in their own words. On WhatsApp, email or the website chat. The assistant finds the order by phone number, email or order number.
- It reads live data. Order details come from Shopify and the tracking status from the courier or aggregator's API, at the moment of asking, so the answer is never out of date.
- It answers plainly. "Your order shipped on Tuesday and is out for delivery today." No guessing: if tracking has not updated for a while, it says so and passes the conversation to a person.
- Returns run through a short form. The customer picks the item and reason and adds photos. The flow checks the order date against your return window and the item against your rules, then creates the return request for a person to approve.
- Escalations arrive ready to act on. When a person is needed, they get the conversation, the order and the tracking in one message, so they can reply without looking anything up.
What stays with a person
Refunds, exceptions to the policy, complaints and anything where the customer is upset go to the team. The assistant only states what the order and tracking data say, and it never promises a delivery date the courier has not given.
Run the numbers for your business
Look at a week of support messages and count how many are about order status or returns. Multiply by four for the month, then by your minutes per message. The savings calculator does the rest.
Questions people ask
We do not use Shopify. Can this still work?
Yes, as long as your store and courier have an API or a way to export data. WooCommerce, custom stores and marketplaces each need a slightly different connection, which we check in the audit.
What if the courier tracking is wrong?
It happens. The assistant reports what the courier says and tells the customer when it was last updated. If the status has not moved for longer than you set, it hands over to a person.
Can customers still reach a human?
Always. Typing "agent" or anything like a complaint sends the conversation to your team straight away.
Written by Kishan Makani
Founder of Makani Tech, 14 years in the tech industry. Builds AI and automation systems for students, developers and business owners from Surat.