+39 348 9323674 +39 011 9222277 info@webcreation.it

Website·E-commerce·Customer service

AI chatbot
for business

Try our chatbot

Real case

Which kibble for my dog?
Now the chatbot answers, even at night.

Pet Food Web got the same questions every evening: which kibble for a puppy, for a neutered cat, where is my order.

We built an e-commerce chatbot that answers with the real catalog, recommends the product by size and age and shows the order status after verifying the customer. On health questions it hands over to the store.

  • 70%of pre-sales questions closed without the store stepping in
  • Evenings and weekendsanswers outside store hours too
  • 2 weeksfrom the first meeting to the chatbot going live

A business chatbot that answers
with your data, not at random.

An AI chatbot answers customers on your website and online store at any hour, with your company's real information, and hands the requests that matter to a person. The problem a business chatbot solves is easy to measure: requests arriving at any hour, always the same ones, and people spending the day answering instead of selling. Or worse, requests left unanswered and customers buying elsewhere.

Compared with the button-based website chatbots of a few years ago, conversational AI understands questions written in people's own words, even with mistakes, and answers with complete sentences. We build chatbots that answer starting from your real data: catalog, availability, prices, shipping terms, FAQ and documents. The language model is there to understand the question and write the answer well; prices and times come from your systems, not from the model's memory.

We put it on your website or online store; for customers who write on WhatsApp there is the page dedicated to the WhatsApp chatbot. When the conversation turns into a negotiation, a complaint or a delicate case, the chatbot collects name and number and passes the request to a person. From €1,000 to €2,000, live in 1-2 weeks. If besides answering you need data entered into your systems, take a look at AI agents and all our AI for business services.

Business chatbot connected to catalog and orders

By sector

Chatbot examples for e-commerce,
hotels and service businesses

The questions change from sector to sector, and with them the sources the chatbot has to answer from. These chatbot examples are the cases we handle most often.

Ecommerce chatbot connected to an online store's catalog

Ecommerce chatbot on WooCommerce and Shopify

Product recommendations from the real catalog, availability, shipping and order status after verifying the customer. On WooCommerce online stores we work with the official APIs; for a Shopify chatbot we use the platform's APIs, and if the store still has to be built there is the Shopify agency page.

Chatbot for hotels and accommodation

Rooms, services, check-in times, parking, how to get there: questions that arrive at every hour and in several languages. The chatbot answers from the property's information, collects booking requests and passes them to reception; if the property management system has APIs it can read availability too.

Service businesses, B2B and professional practices

Data sheets, compatibility, delivery times, quote requests. The chatbot asks the right questions and sends you a contact ready to call back. For professional practices it collects requests without giving opinions, as we explain on the pages about AI for lawyers and AI for accountants.

How we build it

How we build
your chatbot

Building a chatbot that holds up in front of customers takes more work on data and rules than on the model. These are the steps we always take.

  • Knowledge base from your sources

    We gather FAQ, terms of sale, product sheets, support procedures and the real questions that came in over recent months by email and phone. The quality of the answers depends on this.

  • Connection to catalog and orders

    On WooCommerce we read products and orders with the official APIs and read-only permissions. With other online store platforms or ERP systems we use their APIs or a custom connector.

  • Answer and handoff rules

    What it can say, what it must not say, when to ask for contact details and when to hand over to a person. Your brand's tone of voice, languages, hours when an operator is available.

  • Privacy and transparency

    Privacy notice and consent in the widget, minimized personal data, conversations kept for as long as you decide. The chatbot introduces itself as an AI system, as the AI Act requires from 2 August 2026.

  • Logging and continuous improvement

    Every conversation can be consulted. In the first weeks we read the wrong or incomplete answers and correct sources and rules: that is where a chatbot goes from decent to useful.

Before you start

When you really
need a chatbot.

A chatbot is a sales tool: it makes sense when there is a volume of requests that today costs time or loses customers.

You need it if

  • Every day you receive dozens of similar questions through your website, email or WhatsApp.
  • Requests arrive after hours and in the morning you find leads that have gone cold.
  • You have a large catalog and customers struggle to find the right product.
  • You want to handle after-sales without hiring more people.

Something simpler is enough if

  • You receive a few requests a week and a good FAQ page costs less.
  • Every question is different and requires a site visit or a technical quote.
  • Almost all questions arrive on WhatsApp. In that case look at the WhatsApp chatbot.
  • The information is not written down anywhere and has to be put in order first.

Pricing

How much does
an AI chatbot cost

The price depends on channels and connections. We set it at a fixed price in the quote, together with the estimate of the monthly model cost.

Chatbot

€1,000 to €2,000

Website · WhatsApp · Telegram


Timeline 1-2 weeks

  • Chatbot for your customers
  • Connection to catalog and FAQ
  • Integration on one channel
  • Your brand's tone of voice
  • Conversation log
  • GDPR privacy notice and consent
Custom

AI programming

€5,000 to €15,000

Orders · ERP · CRM


Timeline 6-8 weeks

  • Order status and returns
  • Connection to ERP or CRM
  • Custom APIs and connectors
  • Handoff to an operator with context
  • Custom logic
  • Your code, documented
Most requested

Virtual assistant

€3,000 to €7,000

ChatGPT for business · RAG


Timeline 3-6 weeks

  • Assistant for your employees
  • Search across company documents
  • Two integrations (CRM or ERP)
  • Dedicated web interface
  • User permissions and log
  • Training for the people who use it

Small jobs

from €350

Hour packages


5 hours €350 · 10 hours €600 · 20 hours €1,100

  • New knowledge sources
  • New answer rules
  • One more channel
  • Conversation review
  • Model update
  • Night work at a higher rate

Clear monthly costs before you sign.

Model usage is paid to the provider based on conversations. We estimate it on your volumes and set spending limits, so there are no surprises.

Tell us about the project →

On request, payment can be spread over installments with no interest. 10% discount for payment in advance on amounts over €3,000.

Let's talk

Ask the chatbot the questions
your customers ask

Put it to the test and then tell us about your case: channels, volumes, ERP or online store. If you would rather talk it over, leave your number with "I want to be called back".

AI

AI chat · Business chatbot

online · instant replies
Hi, I'm Webcreation's chatbot. Ask me a question the way one of your customers would, or tell me where you want to use a chatbot.

Frequently asked questions

Frequently asked questions
about AI chatbots

How much does a website chatbot cost?

Development ranges from €1,000 to €2,000 for a chatbot on one channel connected to catalog and FAQ, with a timeline of 1-2 weeks. If it has to read orders, the ERP or the CRM, it becomes custom programming, from €5,000. Model usage is separate, and we estimate it on your volumes.

What are some examples of business chatbots?

An ecommerce chatbot that recommends products from the catalog and gives the order status, like the one for Pet Food Web; one for hotels that answers about rooms, services and times and collects booking requests; one for B2B companies that answers about data sheets and compatibility and prepares the quote request. The chat on this page is an example too: it answers with our price list and passes the contacts on to us.

Which chatbots are the most widely used?

Among general-purpose assistants the best known are ChatGPT by OpenAI, Gemini by Google, Claude by Anthropic and Copilot by Microsoft. On business websites, chatbots are built on these models and connected to the company's data, because a generic assistant does not know your store's prices, availability and terms.

Does the chatbot answer about my products?

Yes, if it is connected to the catalog. On WooCommerce we read products and orders with the official APIs and read-only permissions, on Shopify we use the platform's APIs; prices and availability come from the store at the moment of the answer. It shows an order's status only after verifying the order number and email.

When does it hand over to a person?

When it cannot find the answer in the connected sources, when the customer asks for it, and in the cases you decide, such as complaints, negotiations, quote requests or sensitive technical questions. It collects name and number and sends you the full conversation, so whoever calls back already knows everything.

How long does it take?

A chatbot connected to catalog and FAQ is live in 1-2 weeks. If it has to read orders, the ERP or the CRM, it becomes custom programming, with a timeline of 6-8 weeks. In the first weeks after launch we read the conversations and correct sources and rules.

Does the chatbot have to say it is an AI?

Yes. From 2 August 2026 the AI Act requires that people talking to an AI system are informed of it, except where it is obvious. The chatbot introduces itself clearly from the first message.

Can I build a chatbot myself?

With ready-made platforms you can build a basic chatbot quickly. The real work is connecting it to catalog and orders, writing the rules for handing over to a person, handling privacy and transparency, and correcting it by reading real conversations: that is the part we do.