Automating AliExpress Dropshipping: Orders and Tracking

AliExpress dropshipping rests on three things: cheap goods, a wide range and no warehouse at all. The price for that is mountains of manual work. Product cards are copied one by one, prices have to be recalculated after every exchange-rate jump, orders are placed by hand on the supplier's site, and delivery takes three weeks during which the customer keeps asking «where is my parcel». Almost everything here can be automated, except the delivery itself.

What actually eats the time in AliExpress dropshipping

Before automating anything it helps to count where the day goes. In most shops the picture is the same: filling the catalogue, checking prices and availability, placing orders with the supplier, collecting tracking numbers and answering customers about delivery status. Sales and advertising are what there is no time left for.

Crucially, all these tasks are mechanical. None of them requires judgement: copy a description, multiply a price by a coefficient, paste an address into a form, forward a tracking number. That is precisely why they automate most easily and pay off fastest.

The order matters too. Prices and availability come first, because a mistake there costs money and reputation. Then order placement, because that is the biggest volume of routine. And only then card imports - they look like the dullest job, but they happen rarely and in batches.

There is another hidden cost: constant switching between tabs. The shop catalogue, the supplier page, the order spreadsheet, the customer chat and a banking app are all open at once, and with every switch a person loses context. It shows up in no report, yet it is exactly why thirty orders take a full day instead of two hours.

Importing products and building cards without manual copying

Technically an import is parsing the product page: title, specifications, photo gallery, size and colour variants, the price of each variant. A script pulls this data and assembles a draft card in your shop. What remains for a human is to rewrite the title and the first two sentences of the description - and that is exactly the part not worth automating.

The reason is simple: a machine translation of a description from a Chinese marketplace reads like a machine translation, and search engines do not like such text. Thousands of shops import the same cards with the same phrases, and the winner is whoever writes a human description. So the import should produce a skeleton, not a finished page.

Photos are a separate story. The gallery from the marketplace often contains shots with someone else's watermarks, Chinese captions and collages of different sizes. The minimum processing on import is to crop to a single format, drop the worst shots and store them locally rather than hotlinking a foreign server that will serve nothing tomorrow.

Pricing: exchange rate, fees and markup

The price on the marketplace is in dollars, you sell in hryvnia, and the rate moves. If prices are fixed numbers, within a month part of the range is being sold at almost no margin. So the price has to be calculated by a formula rather than typed in by hand.

A workable formula combines the purchase price in currency, the exchange rate with a small buffer, the shipping cost, the payment processing fee and your markup. The markup is not the same for everything: on cheap goods it is high in percentage terms, otherwise the order does not cover the work, while on expensive ones it is lower because the absolute amount does the job.

This turns into a simple rule: once a day the system pulls the rate, recalculates prices by the formula and shows a list of changes. Items whose margin has dropped below a threshold are highlighted separately - they are either repriced or pulled out of advertising. That way the range does not quietly rot but stays inside a defined profitability band.

Placing orders with the supplier

This is the dullest part of the day: open the customer's order, find the item at the supplier, pick the same variant, paste the name, address and phone number, pay, then copy the order number back into your spreadsheet. Five minutes per order, thirty orders a day - and an hour and a half simply disappears.

Even without an integration, minimal automation helps a lot here. The system prepares a fully ready block of data to paste: the exact link to the right product variant, the address in the correct format, the transliterated name, a valid phone number. The person only confirms and pays, and the order number is pulled back into the card.

The second part is control. An order you have accepted but not yet passed to the supplier must be visible as a separate list with a timer. This is exactly where paid items get lost: the request was accepted, the money received, and nobody ever placed the order on the marketplace. One simple filter removes this problem entirely.

Slow delivery: tracking and talking to the customer

Delivery from China takes two to five weeks and you cannot shorten it. But almost all the frustration comes not from the wait but from the silence. A customer who received an honest «the parcel has left, roughly twenty days, here is the tracking» and then three updates along the way waits calmly. A customer with no information writes every day and leaves a bad review before the parcel even arrives.

Technically this is one script: once a day the system polls the tracking numbers, compares the new status with the stored one and sends a message in Telegram or by SMS when it changes. There are only a few key events - shipped, arrived in the country, handed to local delivery, ready for pickup. Four messages over three weeks remove almost every question.

You also need a «stuck» list: parcels with no movement for more than ten days. They are visible in advance, and you write to the customer first rather than hearing from them after a month of silence. That is the difference between a shop that looks reliable and a shop where «I ordered something and it vanished». Devlly builds systems like this for a specific store: card imports, a price formula matching your rate and markup, order preparation and automatic delivery statuses for the customer.