# AI and automation for e-commerce

Most shops spend their support budget answering "where is my order" and lose sales to pre-purchase questions nobody answers. Connecting the assistant to order data solves both; better canned replies don't. The integration is the real work, and the chat window is the easy part.

Data checked: 12 October 2026

## TL;DR verdict

An assistant that can't see the order is a FAQ page with a cursor. It becomes useful once it can look up the shipment, and that takes integration work, not a better prompt.

*AI and automation for e-commerce*

| Bottleneck | What it costs today | What we build |
| --- | --- | --- |
| "Where is my order?" | The largest single share of support volume | Order lookup by email or number, answered from the carrier status |
| Pre-purchase questions | An abandoned basket you already paid to fill | An assistant that answers from the product data, with sizing and stock |
| Returns | Manual approval and a label sent by hand | Policy check, label generation and a status the customer can see |
| Stock across channels | Overselling, then an apology | Sync between shop, marketplace and warehouse on every change |
| Order to invoice | Someone re-typing paid orders each morning | Invoice created automatically when payment clears |
| Abandoned baskets | One generic email | A sequence keyed to what was in the basket and whether it is still in stock |

## Which we'd pick, scenario by scenario

- ### The order status question
  It is the highest-volume ticket in almost every shop, and it needs no judgment: match the customer to the order, read the carrier status, answer. Taking that volume off the team leaves them free for the tickets that need a person.
- ### The question asked at 22:00
  Will this fit, is it in stock, when will it arrive? Left unanswered, the basket is abandoned and the ad spend that filled it is wasted. Answered from live product data, it converts.
- ### The return that took four emails
  Checking the policy, approving the return, issuing a label and telling the customer where it is are four mechanical steps. Automating them cuts the handling cost and, more usefully, removes the follow-up tickets each step used to generate.
- ### Marketplace stock
  Selling the same stock in your own shop and on a marketplace without a sync means overselling. This is a plain integration with no AI in it, and it is often the first thing we build because it stops losses that are happening right now.

The shops that get value here are the ones that let the assistant read the order database. The rest end up with a nicer FAQ and the same ticket volume, because people ask about their own order, not about your policies in the abstract. So the build is mostly plumbing: a clean read path to orders, shipments and stock, then a conversational layer on top that is strictly limited in what it may assert.

That limit matters more in retail than anywhere else we work. An assistant that guesses a delivery date makes a promise the warehouse never made, and the customer holds you to it. Ours answer from a retrieved record or say they will check; they never infer. In Poland the integration list is specific: Baselinker in the middle, Allegro and WooCommerce on the selling side and one of the invoicing systems behind them. Getting those to talk to each other usually removes more manual work than the chatbot does. The messaging side is the same build described in [the WhatsApp bot tutorial](/en/blog/whatsapp-bot-that-qualifies-leads-in-60-seconds/), and the back-office half is ordinary [process automation](/en/solutions/automations/).

## Sources

- [Baselinker: API documentation](https://api.baselinker.com/)
- [WooCommerce: REST API documentation](https://woocommerce.github.io/woocommerce-rest-api-docs/)

## Frequently asked

### What should an online shop automate first?

Order status, because it is the biggest share of support volume and needs no judgment at all. If you sell on more than one channel, stock sync comes next, since overselling costs more than a slow reply.

### Can the chatbot see my customers' orders?

Yes, and that is what makes it useful. It identifies the customer by order number and email, then reads the order through your shop's API. It reads only what it needs to answer, and every lookup is logged.

### Will an AI assistant invent delivery dates?

Not the way we build them. The assistant may state a date only if it came from the carrier or your own records, and says it will check when it has nothing. Retail is the worst possible place for a confident guess, because the customer will hold you to it.

### Does this work with Baselinker, Allegro or WooCommerce?

Yes. Those three plus an invoicing system are the usual Polish stack, and they all have APIs we integrate with. The work is mapping your specific flow between them, not building connectors from scratch.

### How much support volume can this actually take?

It depends on your mix, but order status alone is usually the largest single category in a shop's inbox. We measure your real ticket mix in the audit before promising a number, because quoting an industry average at you would be guessing.

## Related pages

- [How to build a WhatsApp bot that qualifies leads in 60 seconds](/en/blog/whatsapp-bot-that-qualifies-leads-in-60-seconds/)
- [AI chatbots](/en/solutions/ai-chatbots/)
- [Process automation](/en/solutions/automations/)
- [AI and automation for law firms](/en/industries/ai-for-law-firms/)
- [AI and automation for medical clinics](/en/industries/ai-for-medical-clinics/)

## Which process costs you the most?

Thirty minutes, free. You leave with a concrete fix plan and what the process costs you. The plan is yours, even if you build it without us.

[Book a free audit](/en/free-audit/)

---

Where an online shop leaks margin: pre-purchase questions nobody answers, order status tickets and manual returns. What to automate first, and why.

Source: https://dobify.dev/en/industries/ai-for-ecommerce/