In short: in an online store, every order triggers the same 5–6 actions — issuing the invoice, e-Factura reporting, generating the waybill, updating stock, notifying the customer and recording it in accounting. Properly automated, they run in under a minute, with no human involvement. Manually, they take 5–10 minutes per order — 8 to 17 hours a month at 100 orders.
That is the calculation that decides whether automation is worth it: not "it's modern", but how many hours per month you get back. We build these flows for Romanian stores and, in this guide, we show exactly what gets automated, in what order, and where the problems appear. For the budgeting context of a store, see how much an online store costs; if you are still choosing a platform, compare the options in WooCommerce vs Shopify in Romania.
What actually happens after a customer clicks "Order"
In short: a placed order is the start of a chain of steps, not the end. Every step is a place where human error can creep in — and every one of them can be automated.
| Step | Manual | Automated |
|---|---|---|
| Payment confirmation | You check the processor account | Webhook from the processor, instantly |
| Issuing the invoice | You type data into the invoicing app | Invoice generated via API from order data |
| e-Factura reporting | You upload the XML to SPV | Automatic submission, with confirmation |
| Waybill generation | You fill in the courier's form | Waybill generated from the delivery address |
| Stock update | You change quantities by hand | Automatic decrement on payment confirmation |
| Customer notification | You write the email | Email/SMS with invoice and tracking number |
The difference is not only time. An automated flow also removes the errors that cost money: an invoice issued with wrong data, a waybill with an incomplete address, a product sold when it was already out of stock.
The core flow: from paid order to handed-over parcel
In short: correct automation starts from payment confirmation, not from order placement. That is the difference between invoicing real sales and invoicing intentions.
- 1. Trigger: payment is confirmed by the processor (webhook) and the order moves to "processing".
- 2. Invoice: issued automatically through the invoicing service's API, using the customer data from the order. Practical detail in SmartBill API integration.
- 3. e-Factura: the invoice is submitted to SPV and the confirmation is awaited. Full context in e-Factura automation.
- 4. Waybill: generated at the courier from the order's address and weight, with the tracking number attached to the order.
- 5. Stock: quantities decrease, and if a product drops below its minimum threshold you get an alert.
- 6. Customer: receives a single email with the invoice attached and the tracking link — not three separate messages.
Order matters. Generate the waybill before confirming payment and you end up cancelling shipments. Decrement stock at order placement rather than payment confirmation and abandoned orders will block products that could have sold.
What to build it with: plugin, n8n or custom code
In short: for standard flows a plugin is enough and cheapest. For flows with your own conditions, a visual orchestrator such as n8n fits better. Custom code is justified only when volume or business logic demands it.
| Approach | Initial cost | Suited to | Limitations |
|---|---|---|---|
| Dedicated plugin | 50–300 EUR/year | Standard flow: invoice + waybill | Little flexibility; you depend on the plugin's developer |
| n8n (self-hosted) | 800–3,000 EUR setup | Own conditions, several connected systems | Needs hosting and someone who understands the flows |
| Custom code | 3,000–8,000 EUR | High volume, specific logic, ERP | Higher upfront cost; only sensible above a certain threshold |
Our honest recommendation: start with the plugin. If after three months you find you are still doing manual steps because the plugin does not cover a case, move to n8n — we wrote about what it can and cannot do in automations with n8n, and the comparison with the main alternative is in n8n vs Make. Do not start from custom code just because it sounds more serious.
The payback calculation: when it is worth it
In short: automation pays for itself when the hours saved exceed the cost, usually somewhere above 50–80 orders per month. Below that threshold you do it for correctness and peace of mind, not for savings.
| Orders/month | Manual time (at ~7 min/order) | Conclusion |
|---|---|---|
| 20 | ~2.5 hours/month | Manual is acceptable; automate invoicing only |
| 80 | ~9 hours/month | The threshold where automation pays for itself |
| 200 | ~23 hours/month | Manual becomes a real brake on growth |
| 500 | ~58 hours/month | The equivalent of a part-time employee |
The 7-minutes-per-order estimate is an average reference for a complete flow: invoicing, e-Factura, waybill, stock and notification. Measure your own time across 10 orders before deciding — your real figure is the only one that counts.
The pitfalls we see most often
- Automation without error handling. What happens if the invoicing API does not respond? With no retry and no alert, the order stays without an invoice and nobody finds out until the customer complains.
- Invoicing at order placement rather than payment. It generates invoices for orders that are never paid, which someone then has to credit back.
- No log. If the flow does not record what it did and when, the first problem becomes impossible to debug.
- Automating a bad process. If the manual flow is confusing, automation makes it confusing and fast. Clarify the process first.
- Zero human control over special cases. Returns, B2B orders and partial deliveries need an exception path, not to be forced through the standard flow.
What is not worth automating
In short: not everything that can be automated should be. Rule of thumb: automate what is repetitive and predictable; keep a human where judgement is required.
- Returns and complaints. Every case has its own context, and a badly worded automatic reply loses the customer for good.
- B2B negotiations. Volume pricing and payment terms are discussed, not computed.
- Product descriptions. Bulk-generated, unreviewed copy is exactly the kind of content Google treats as low quality.
Frequently asked questions
How much does order automation cost for a small store?
For the standard flow — invoice, e-Factura, waybill — a dedicated plugin costs between 50 and 300 EUR per year and covers most small stores. A custom n8n flow starts at roughly 800 EUR for initial setup, plus hosting.
Does automation work on platforms other than WooCommerce?
Yes. The principle is identical on any platform offering webhooks and an order API. What differs is the availability of local integrations: the WooCommerce ecosystem is richer in Romania for invoicing and couriers.
What happens if the automation gets stuck?
A properly built flow retries the operation a few times, then sends an alert and leaves the order in a clearly marked state so it can be resolved manually. A flow without alerts is more dangerous than manual work, because errors pass unnoticed.
Do I need a developer to maintain the automations?
For plugins, no — they are configured from the admin panel. For n8n, you need someone who understands the flows when an API changes or a new case appears. We hand over documented flows so you do not depend on us for small changes.
Does automation also handle e-Factura reporting?
Yes, if the invoicing service you use offers SPV submission through its API. The flow sends the invoice, waits for confirmation, then marks the order. Details in the e-Factura automation guide.
Want to know what can concretely be automated in your store? Tell us which platform you use, how many orders you get per month and which steps you do by hand today, through our request a quote page. We reply within 24 hours with a proposed flow and an estimate — including when the conclusion is that a 100 EUR plugin solves your problem. More on our approach on the AI automation page.