You answer "what's the delivery time?" twenty times a week, and you often do it in the evening, after the real workday is over. An AI customer service agent is a program that reads a customer's question, looks up the answer in your own information (FAQ, terms of sale, catalog, order tracking) and replies on its own, 24 hours a day. For a small business handling 200 to 500 messages a month, the technical bill usually lands between €5 and €40 per month. The real cost is somewhere else: setup time, maintaining the knowledge base, and the requests the machine absolutely has to hand off to a human. Below you'll find the numbers broken down line by line, three ways to set it up, and an honest list of what breaks in production.
What an AI customer service agent does, and what it doesn't
An AI agent handles questions whose answer is already written down somewhere very well, and handles anything requiring a business decision very badly.
What it handles properly from week one:
- delivery times, shipping costs, areas you ship to
- opening hours, your address, how to book an appointment
- the status of an order, as long as it's connected to your sales tool
- product questions: material, size, compatibility, warranty, care
- triaging requests: it qualifies the message and files it in the right queue before you even open it
What it doesn't handle, and what you need a human backup for:
- a refund outside your stated terms, a goodwill gesture, a price negotiation
- an angry customer threatening to leave a one-star review
- anything touching a person's health, legal situation, or safety
- a request whose answer doesn't exist anywhere in your documents: in that case it has to say it doesn't know and pass it on
The difference from a classic button-based chatbot is simple: the agent understands a question phrased any old way, with typos and loose vocabulary, and it can trigger actions (look up an order, create a ticket, send a confirmation email). If you want to see what building one step by step looks like, the guide on creating a chatbot for your business with Claude covers the technical side.
How much an AI customer service agent costs a small business
Budget €2 to €7 per month in model calls for 300 conversations, €0 to €55 for the tools around it, and 6 to 15 hours of work the first time.
AI models bill by the token. A token is roughly three quarters of a word. You pay for the tokens you send (the customer's question, your FAQ, the conversation history) and the ones the model produces. According to the public Claude API pricing, a lightweight model like Haiku sits around $1 per million input tokens and $5 per million output tokens; a top-tier model lands around $3 and $15.
Here's the concrete math for an average six-message conversation, with your FAQ injected into every call: roughly 4,000 input tokens in total and 600 output tokens. With the lightweight model that comes to $0.007, less than a cent. With the top-tier model, about 2 cents. Over 300 conversations a month, you're looking at €2 to €7.
| Line item | Ballpark for 300 conversations/month |
|---|---|
| AI model calls | €2 to €7 |
| Hosting the automation (n8n cloud or a small server) | €0 to €25 |
| Chat widget on your site | €0 to €30 |
| Upskilling | Monthly subscription with Skilzy |
| Your setup time | 6 to 15 h, one time only |
| Monthly upkeep | 1 to 2 h |
The line item that really blows up a budget is the per-resolution pricing some support platforms use: around €1 per conversation resolved by the AI. On 300 conversations with half resolved automatically, you end up at €150 per month, versus €5 calling the model directly. It's not a rip-off: you're paying for the interface, the hosting, the analytics, and the fact that you don't have to build anything. But know what you're buying.
One last number to keep in mind: your own time. If you spend 20 hours a month answering repetitive messages and you value your hour at €40, your customer service currently costs you €800 a month. That's the number to compare against, not zero.
Three ways to set up your AI agent
The most cost-effective option over twelve months is learning to build the agent yourself; the fastest is renting something ready-made.
Method 1: learn to build it and keep it in-house
You assemble three building blocks: an AI model called through an API, a knowledge base (your product pages, your FAQ, your terms of sale) and an automation tool like n8n that wires it all up to your inbox, your website, or your WhatsApp. No code to write: n8n works with blocks you connect together, and the official n8n documentation covers every piece.
The sticking point for a beginner is never technical, it's methodology: what to put in the system prompt, how to force the agent to answer "I'm passing this to a human" instead of making something up, how to test on 50 real questions before going live. The Skilzy AI freelance program covers exactly this chain, from writing reliable prompts to shipping an automation that holds up in production. It's designed for people who want to bill clients for this work, and it's the same skill set you need for your own business.
Two practical details. First, the included AI Lab lets you test the tools with credits provided, instead of stacking up three €20 subscriptions while you learn. Second, two state-recognized certifications (RS7439 in AI content marketing, RS6792 in AI and sales) can be funded if you go through a training scheme.
Realistic budget: €30 a month for training over two to three months, then €5 to €30 a month to run. You stay in control of the agent, and you can tweak it on a Sunday night without waiting on a vendor.
Method 2: an off-the-shelf support tool with built-in AI
Chat and ticketing software all now offer an AI layer that reads your help-article library and answers on your behalf. You install a script on your site, you import your FAQ, and it's live in two hours.
Pros: zero building, clean interface, analytics included, human escalation handled natively, compliance documented by the vendor.
Cons: pricing ranges from around twenty euros per month per user up to roughly €1 per AI resolution, which climbs fast with volume. You're also limited to what the vendor planned for: hooking up your homemade inventory spreadsheet or your local point-of-sale software is often impossible. Check current pricing tiers before you commit, they change several times a year.
Who it's for: you have a standard e-commerce site, stable volume, and zero interest in touching anything technical.
Method 3: have a custom agent built for you
You hand the project to a freelancer or an agency. Quotes in this range run from €1,500 for a FAQ agent wired into a brochure site to €8,000 and up for an agent connected to an ERP, with customer authentication and real-time order tracking.
Pros: it works, and quickly, without you having to learn anything.
Cons: every change goes through the vendor, and a customer service agent changes constantly (new product, new delivery time, new promo). Negotiate access to the automation editor into the quote from day one, plus a training session so you can update the knowledge base yourself. Without that, you'll pay €150 in billable work to change one sentence about shipping costs.
| Method | Year 1 cost | Time to launch | Autonomy |
|---|---|---|---|
| 1. Build it after training | €150 to €500 | 2 to 6 weeks | Full |
| 2. Off-the-shelf tool | €300 to €2,000 | 2 hours | Low |
| 3. Custom development | €1,500 to €8,000 | 3 to 8 weeks | Medium |
The real limits, the ones you discover after going live
An AI customer service agent mainly fails on three fronts: it makes things up when it doesn't know, its knowledge base goes stale, and it's subject to legal obligations nobody reads.
It makes things up. A language model produces the most likely continuation of a text, not the truth. Ask it whether your product is compatible with a model number it has never seen, and it will sometimes answer "yes, absolutely" with perfect confidence. The fix lives in the system prompt: an explicit ban on answering outside the documents provided, plus a mandatory fallback sentence. Test that instruction with twenty trick questions before you open it up to the public.
Its knowledge base rots. You switch carriers, you move shipping to €4.90, you discontinue a product. If nobody updates the documents, the agent spends six months quoting a price that no longer exists, and you find out from a customer review. Block one hour a month in your calendar for this review pass — it's the main hidden cost.
It's legally regulated. As of August 2026, the transparency obligations of the European AI regulation apply: the person has to know they're talking to a machine. One clear greeting sentence is enough. On the personal data side, your conversations contain names, addresses, sometimes order numbers: you need a legal basis, a retention period, a line in your privacy policy, and a check on where your vendor hosts the data. The CNIL's recommendations on AI lay out the framework that applies in France.
It can't break out of a loop. The worst scenario for your reputation isn't a bad answer, it's the stuck customer who can't find any way to reach a human. Put a visible handoff button on every message, and trigger escalation automatically after two unsatisfactory replies or as soon as a word like "refund", "lawyer" or "cancel" shows up.
Its tone goes off the rails. By default, an AI agent writes like a translated American legal department. Give it three examples of your own past replies in the prompt and the result changes completely. The same principle applies to your written replies in general, as covered in the guide on automating your emails with AI.
Measuring the return and starting out cleanly
Your agent pays for itself once it closes at least 40% of your conversations without human involvement, with a stable level of complaints.
Three metrics are enough, checked every month:
- Autonomous resolution rate: the share of conversations that end without you. Below 30%, your agent doesn't have enough information. Above 70%, check that it isn't blocking customers who wanted a human.
- Time saved: number of resolved conversations multiplied by your average response time. 135 conversations resolved at 4 minutes each works out to 9 hours a month.
- Complaints and reviews: if the number of negative reviews mentioning support goes up, your resolution rate is a false friend.
Your launch plan, in this order:
- Step 1. Export your last 100 customer messages and sort them. You'll find that 6 or 7 questions cover 80% of the volume.
- Step 2. Write the ideal answer to those 7 questions, in your own voice, with exact figures. That's your knowledge base, and it's 80% of the work.
- Step 3. Set the agent up on a single channel, the one with the highest volume (site contact form, or email).
- Step 4. Keep it in draft mode for two weeks: it prepares the reply, you approve before it goes out. That way you fix mistakes without showing them to a customer.
- Step 5. Switch to automatic on simple questions only, and keep human approval on everything else.
This logic of approved drafts before full automation applies to almost every process in a small business; the guide to automating your business with AI shows how to apply it to invoicing, prospecting and reporting.
Key takeaways
The technical cost of an AI customer service agent has become marginal for a small business: a few euros a month if you build it, a few dozen if you rent it. What you're really paying is about ten hours of prep and one hour of upkeep a month. In exchange, you often get 8 to 10 hours back. Start small, on one channel, with seven questions and human approval, and expand only once the numbers look good.
