Go to almost any business website and there is a little chat bubble in the corner. One site calls it live chat. Another calls it a chatbot. A third says it is an AI assistant. They can look almost identical, but what happens after you type a question can be completely different.
That is where the confusion starts. Live chat describes the channel: a conversation taking place through a chat window. It does not tell you whether the reply comes from a person, a script or artificial intelligence.
The short answer: A chat window is simply the way a customer gets in touch. Behind it might be a basic scripted chatbot, an AI customer service assistant, a human agent, or a combination of AI and human support. The important question is not “does the website have live chat?” It is who or what is answering, what can it actually do, and what happens when it reaches its limit?
First: “Live Chat” Does Not Always Mean a Live Person
Traditionally, live chat meant a staff member sitting at a computer and replying in real time. Plenty of businesses still use it that way. But the label is now used much more loosely.
A window marked “live chat” may start with AI, pass certain questions to a human, or be entirely automated. Some services even present an automated assistant as live chat because the response itself is immediate. From the customer's point of view, all they see is the same small box.
It is more useful to separate the chat window from the service behind it.
| Type | Who answers? | Best at | Main limitation |
|---|---|---|---|
| Basic chatbot | Rules and preset replies | Simple menus and FAQs | Struggles outside its script |
| AI customer service | AI trained on the business | Natural questions, enquiries and lead capture | Needs clear boundaries and maintenance |
| Human live chat | A person | Judgement, reassurance and unusual cases | Only works when someone is available |
| Hybrid service | AI first, human when needed | Fast coverage with human escalation | Requires a well-designed handover |
What Is a Basic Chatbot?
A basic chatbot follows a decision tree. It might open with three buttons:
- Book an appointment
- Check our opening hours
- Leave a message
Choose a button and it gives the answer attached to that option. Type something unexpected and it may repeat the menu, search a small list of keywords or tell you it does not understand.
There is nothing inherently wrong with this. If a business only needs to direct visitors to the right page, a scripted bot may be enough. It is predictable, inexpensive and tightly controlled.
The problem is that customers rarely describe their needs in neat menu options. They ask, “Could you fit me in after work next Thursday?” or “Do you cover Nantwich, and is there a call-out charge?” A fixed script cannot reliably deal with all the ways a real person might phrase the same question.
What Do We Mean by AI Customer Service?
An AI customer service assistant is designed to understand ordinary language rather than wait for an exact keyword or button press. It is given information about a specific business: its services, prices, locations, opening hours, policies and the questions customers regularly ask.
That means a visitor can ask a question in their own words. The assistant can interpret what they mean, use the relevant business information and respond conversationally.
But answering questions is only part of the job. A properly set-up assistant should also be able to:
- Ask sensible follow-up questions
- Collect the details needed for an enquiry
- Qualify a lead without making the conversation feel like a form
- Explain what will happen next
- Recognise subjects it should not answer
- Pass the conversation to a person or create a clear follow-up
This is why we prefer the phrase AI customer service to “chatbot”. Chatbot describes the thing a visitor sees. AI customer service describes the job it is there to do.
There is also an important difference between installing AI software and running an AI customer service function. The information changes. Customers find new ways to ask questions. Some answers need tightening, and some conversations reveal gaps nobody anticipated at launch. Someone has to review, train and maintain the assistant. Without that work, even good technology becomes unreliable.
What Is Human Live Chat?
Human live chat connects the visitor to a member of staff or an outsourced support agent. The person reads the message and replies through the same kind of chat window.
A capable person is still better at conversations involving emotion, judgement or ambiguity. They can reassure an unhappy customer, spot when the stated question is not the real concern, negotiate, make an exception or deal with a situation the business has never seen before.
The limitation is availability. A person can only answer when they are working, and can only handle so many conversations at once. If nobody is available, the “live” chat usually becomes another contact form. That is often exactly when customers are browsing: during the evening, at weekends or while the team is busy doing the actual work.
A Realistic Example
Imagine someone visits an estate agent's website at 10.30pm and asks:
“Is the house on Market Street still available, and could I view it on Saturday morning?”
A basic chatbot may offer a link to the property page or ask the visitor to fill in a form. Human live chat can deal with it well, but only if somebody is on duty. An AI customer service assistant can answer from the information it has been given, collect the buyer's details and preferred time, then send the agency a structured viewing enquiry to confirm in the morning.
Now imagine the same visitor says:
“My purchase may collapse because the survey found subsidence. What should I do?”
That is not a routine enquiry. It calls for context, judgement and careful handling. The assistant should recognise the boundary, avoid pretending to advise and make sure the right person follows up.
Good AI customer service is not defined by answering everything. It is defined just as much by knowing what it should not answer.
So, Which Is Better?
There is no universal winner. Each option solves a different problem.
Use a basic chatbot when:
- You only need a small number of fixed journeys
- The answers rarely change
- You want visitors to choose from a menu rather than hold a conversation
Use human live chat when:
- Most conversations are sensitive, complex or high-value
- You have staff available when customers are browsing
- Personal judgement is central to the service
Use AI customer service when:
- You receive repeat questions and enquiries throughout the day
- Customers browse when your staff are unavailable
- A fast first response makes a commercial difference
- You want to capture useful details before a person follows up
For many small businesses, the best answer is a hybrid. AI handles the immediate, repeatable first response. People handle the conversations where experience, empathy and authority matter.
Where Delegait Fits
Delegait is not a scripted menu dropped onto your website and left there. We provide managed AI customer service: an assistant trained around your business, your customers and the way you want enquiries handled.
It can answer routine questions, capture and organise leads, and be available when your team is not. Just as importantly, it is given boundaries. When a conversation needs a person, it should say so clearly and create a useful handover rather than bluff its way through.
The word “chatbot” is still useful because it is what many people search for and what the customer sees on the page. But it does not fully describe the service behind it. The better description is:
Immediate when it can help. Human when it should be.
The Bottom Line
Do not judge a website chat service by the label on the button. “Live chat” may be human or automated. “Chatbot” may mean a rigid menu or a capable AI assistant. The useful questions are what it knows, what it can do, how it is maintained and what happens when the conversation needs a person.
If you are comparing options for your own business, start with the enquiries you receive repeatedly and the times they arrive. That will tell you whether you need a simple signpost, a staffed conversation, or managed AI customer service that can cover the gap between the two.
You can see how Delegait's managed AI customer service works, or read our guide to what an AI chatbot costs for a UK small business.