Automating an online store and marketplace sales

Orders come from the store, Allegro, Amazon and other marketplaces. Stock levels are updated by hand or with a delay, so items that are no longer available sometimes get sold.

Orders from all channels go into one workflow. Stock levels update across all channels after every sale, and orders with unusual notes or shortages go straight to customer service.

Some customers give a tax ID and need an invoice; most buy as consumers. Staff check every order and issue invoices by hand, including for earlier receipts.

When an order has a tax ID, the system issues an invoice in the FA(3) structure, sends it to KSeF and passes it to the customer with the KSeF number. Consumer orders get a document that matches the store settings, and requests for an invoice for a receipt are handled by a form.

The customer emails, staff send instructions, a label and a bank account number for the refund. After the parcel arrives, someone has to remember the transfer and the invoice correction.

The customer reports a return in a form and gets a label straight away. After the parcel is received, the warehouse confirms the condition of the goods, the system triggers the refund and, for business purchases, issues a corrective invoice in KSeF.

Purchase invoices arrive separately from the goods. Quantities and prices are compared with the supplier order and warehouse receipt only occasionally.

Purchase invoices are downloaded from KSeF and linked to the supplier order and the goods receipt document. Matching lines go to payment, and differences in quantities and prices go to the person responsible for purchasing.

New products wait weeks for descriptions. Supplier data is incomplete, and every marketplace and language needs a separate version.

AI writes descriptions based on supplier data and your style, prepares marketplace versions and translations, and flags missing required product information, such as manufacturer details and warnings. A description is published after approval.

Every promotion must show the lowest price from the 30 days before the discount. Competitor prices are checked by hand, and the margin after a price change is calculated in a spreadsheet.

The system records the price history of every product and automatically shows the lowest 30-day price during promotions. It monitors selected competitors' prices and suggests changes, but never goes below the set margin.

You find out about shortages when a product has already sold out. Supplier orders are placed by gut feeling, so some products pile up while others run out.

A workflow tracks sales and delivery times from each supplier, warns about products that will run out in the coming weeks and prepares a draft order for approval.

Staff answer the same questions: where is my parcel, how do I return it, which size should I choose. Weekend messages wait until Monday.

A chatbot on the website, Messenger and WhatsApp checks order status, guides customers through returns and gives advice based on product descriptions and store terms. Unusual cases are passed to a person with a summary of the conversation.

Customers abandon carts, and after a purchase they get nothing but a confirmation. Campaigns are prepared by hand and rarely.

Customers who have given consent receive a cart reminder, usage instructions after delivery and a restocking suggestion at the right moment. The content matches what they bought.

No one asks for reviews, and marketplace reviews get a late reply or none at all. A negative review sits unanswered.

After delivery, the customer receives a review request. AI drafts responses to new reviews, and negative ones go straight to customer service together with the order history.

Labels are generated in several carrier portals, and tracking numbers are copied into orders. Customers ask where their parcel is.

Once an order is paid, the system chooses a carrier based on weight, size and the customer's choice, generates the label and adds the tracking number to the order. The customer gets a tracking link, and delivery problems go to customer service.

Payouts from payment gateways and marketplaces arrive in bulk, minus commissions. Matching them to orders takes accounting a lot of time.

Payout reports are automatically broken down into orders, commissions and refunds, and then matched with the bank statement. Accounting receives a ready summary, and the store sees the real margin in each channel.

When selling to consumers in other EU countries, you have to track the EUR 10,000 threshold, VAT rates in each country and the quarterly OSS return.

The system totals sales to EU consumers, warns before the threshold is exceeded, applies the VAT rate of the customer's country and prepares data for the OSS return for accounting every quarter.