Table of Contents
- Key takeaways
- What is customer chat support?
- What are the three models of chat support?
- Are live chat support agents real people?
- How does chat support work?
- What are the benefits of live chat for customer service?
- 1. Customers get a faster first reply
- 2. One agent can handle several conversations
- 3. Every chat leaves a written record
- 4. Help arrives on the page where the question came up
- 5. Coverage can run past office hours
- 6. Chat is easy to measure
- What are the drawbacks of live chat?
- Can chat support software reduce customer wait times?
- Worked example: taking routine questions out of the queue
- Chat support vs phone and email support
- In-house, BPO, or chatbot: who should answer your chats?
- What is BPO chat support?
- How to choose between in-house, BPO, and a chatbot
- What to ask a call center chat provider
- Best features of chat support software
- The stack: which tool does which job
- Common chat support mistakes to avoid
- Frequently asked questions
- Basics
- What is chat support?
- What is the difference between live chat and chat support?
- Are live chat support agents real people?
- Benefits and wait times
- What are the benefits of live chat support?
- Can chat support software reduce customer wait times?
- Is live chat better than phone support?
- Staffing and BPO
- What is BPO chat support?
- How does call center live chat work?
- How can a small team offer 24/7 live chat support?
- Chatbots and tools
- What is the best chatbot for customer service?
- Is SiteGPT a live chat tool?
- Does B2B customer service need chat?
- Sources
Created by
Do not index
Created time
Oct 25, 2023 11:33 PM
Customer chat support is a customer service channel where people type a question into a chat window on a website, in an app, or in a messaging app and get a reply in real time from a human agent, a chatbot, or both.
TL;DR
- Chat support comes in three models: human live chat, chatbot only, and hybrid, where a chatbot answers first and hands off to a person.
- The main benefit of live chat for customer service is speed. Agents send the first reply in 35 seconds on average, according to LiveChat's report on 2 billion chats.
- The queue is the weak point. The average queue wait is 4 minutes 18 seconds, and 27.4% of customers leave before an agent answers.
- Chat support software cuts wait times by routing chats, saving replies, and letting a chatbot answer routine questions with no queue.
- A chat can be staffed in-house, by an outsourced BPO team, or by a chatbot with people behind it. In a 2026 Gartner survey, 87% of customers said access to a human agent is essential.
Key takeaways
Question | Short answer | Number to know |
What is chat support? | Real-time text help from a person, a chatbot, or both | About 90% of chats in LiveChat's data were handled by people only |
What is the main benefit of live chat? | A fast first reply on the page where the question came up | 35 seconds average first response time |
What is the main risk? | A queue that customers abandon | 27.4% queue dropout rate |
Can software reduce wait times? | Yes, mostly by keeping routine questions out of the queue | 4 minutes 18 seconds average queue wait |
Who should answer the chats? | In-house agents, a BPO team, or a chatbot with people behind it | 87% of customers call human access essential |
What is customer chat support?
Customer chat support is help delivered as typed messages in real time. The customer opens a chat window, types a question, and reads the reply in the same window a few seconds or minutes later.
The chat window usually sits in the corner of a website. It can also live inside a mobile app or run through a messaging app such as WhatsApp or Messenger.
"Chat support" and "live chat" are often used as if they mean the same thing. Live chat means a person is typing the replies. Chat support is the wider term and covers people, chatbots, and a mix of the two. If you need plain definitions of live chat and a chatbot with sample conversations, see live chat vs chatbot: what is the difference.
What are the three models of chat support?
The three models of chat support are human live chat, chatbot only, and hybrid. They differ in who writes the reply.
Model | Who answers | Strong at | Weak at |
Human live chat | An agent types every reply | Judgment calls, upset customers, account problems | Queues at busy times, cost of night and weekend shifts |
Chatbot only | Software answers every message | Routine questions at any hour with no queue | Anything outside its content, and customers who want a person |
Hybrid | A chatbot answers first and hands off to an agent | Short waits for routine questions, people for the rest | Needs clear handoff rules and someone to pick up the handoffs |
The longer comparison of the first two models is in chatbots vs live chat. If the chatbot side is new to you, start with what a customer support chatbot is, which explains how one finds its answers and when it should hand off.
Are live chat support agents real people?
Live chat support agents are real people in most chats, but some chat windows are answered by a chatbot. LiveChat's Customer Service Report counted 1,676,529,825 chats across the businesses using its product. A chatbot took part in 163,480,050 of them.
That is 9.8% of all chats. The other 90.2%, or 1,513,049,775 chats, were handled only by human agents. The report's data was last updated in 2024 and covers LiveChat customers only, so treat the split as one vendor's view.
A good chat window says which one is answering. The chatbot best practices guide covers how and when to tell visitors they are talking to a bot.
How does chat support work?
Chat support works in five steps, from the first message to the saved transcript.
- The customer opens the chat window and types a question. Some sites ask for a name and email first.
- The software routes the chat. It goes to a chatbot, to the next free agent, or into a queue if every agent is busy.
- The first reply arrives. A chatbot replies at once. An agent replies when the chat reaches them.
- The chat is handed off if needed. A chatbot passes the conversation to a person, or an agent passes it to a specialist, with the messages so far.
- The chat closes. The customer can rate it, and the transcript is saved for the team to review.
Each step produces a number a team can track. LiveChat's report gives these global averages.
Metric | What it measures | Global average |
First response time | Time an agent takes to send the first reply | 35 seconds |
Queue waiting time | Time a customer spends in a queue before reaching an agent | 4 minutes 18 seconds |
Queue dropout rate | Share of customers who leave the queue before being served | 27.4% |
Chat duration | Length of one chat conversation | 8 minutes 25 seconds |
Customer satisfaction | Satisfaction score across rated chats | 64.2% |
Chat availability | Hours per day a business is available on chat | 17 hours 58 minutes |
What are the benefits of live chat for customer service?
The benefits of live chat for customer service are a faster first reply, more conversations per agent, a written record, help on the page where the question came up, longer coverage hours, and numbers a team can manage by.
Benefit | What it changes | Number to watch |
Faster first reply | The customer gets an answer while still on the site | First response time |
More conversations per agent | One agent can hold several chats at once | Chats per agent per day |
A written record | Links and steps stay in the chat, and the transcript can be reviewed | Repeat contacts on the same issue |
Help on the page | Questions get answered at the pricing page or checkout | Chats started per page |
Longer coverage hours | Chat can stay open after the phone line closes | Hours available per day |
Easy to measure | Every chat has timestamps, a transcript, and a rating option | Satisfaction score |
1. Customers get a faster first reply
Live chat gives customers a first reply in seconds. The global average first response time in LiveChat's data is 35 seconds. An email reply is measured in hours, and a phone call starts with a hold.
2. One agent can handle several conversations
A chat agent can keep more than one conversation open, because customers take time to read and type. A phone agent can only take one call at a time. The same team can answer more customers on chat than on the phone.
3. Every chat leaves a written record
A chat leaves a transcript the customer and the team can both read later. Agents can paste a link, a tracking number, or numbered steps, and the customer does not have to write anything down.
4. Help arrives on the page where the question came up
Live chat answers the question on the page that raised it. A visitor who has a shipping question at checkout can ask it there and finish the order. With email or phone, that visitor has to leave the page.
5. Coverage can run past office hours
Businesses in LiveChat's data were available on chat for 17 hours 58 minutes a day on average. That is longer than a standard office day, and a chatbot can cover the remaining hours for routine questions.
6. Chat is easy to measure
Chat software records when each message was sent, how long the chat took, and how the customer rated it. A manager can see slow replies and long queues the same day.
What are the drawbacks of live chat?
The drawbacks of live chat are queues, staffing cost, and uneven satisfaction. More than one in four customers (27.4%) left the queue before an agent replied in LiveChat's data. The average satisfaction score across rated chats was 64.2%.
Chat also suits some problems badly. A billing dispute or an upset customer often goes better on a call. The guide on how to respond to an angry customer covers those conversations.
Can chat support software reduce customer wait times?
Yes. Chat support software reduces customer wait times in three ways: it routes each chat to a free agent, it gives agents saved replies for common questions, and it lets a chatbot answer routine questions with no queue.
The wait in live chat comes from the queue. Once a chat reaches an agent, the first reply takes 35 seconds on average. The average time in the queue before that is 4 minutes 18 seconds, or 258 seconds. The queue is 7.4 times longer than the reply.
Worked example: taking routine questions out of the queue
This example uses a hypothetical business with 300 chats a day that all wait for an agent. The dropout rate is LiveChat's global average of 27.4%. The 40% chatbot share is an assumption for illustration.
Setup | Chats answered by the chatbot | Chats in the agent queue | Customers lost at a 27.4% dropout rate |
Agents only | 0 | 300 | 82 |
Chatbot answers 40% first | 120 | 180 | 49 |
The queue falls from 300 chats to 180, which is 40% fewer. At the same dropout rate, 33 fewer customers leave without an answer each day. A shorter queue can also lower the dropout rate itself.
The example only holds if the chatbot's answers are right. A chatbot that guesses moves the wait from the queue to a second contact.
Chat support vs phone and email support
Chat support is faster than email and less personal than a phone call. Most teams run all three and send each kind of question to the channel that suits it. To write down which touchpoint goes to which channel, with an owner and a metric for each, use the customer engagement plan template.
Channel | How fast customers expect a reply | Best for | Weak at |
Chat | Seconds to a few minutes | Quick questions, order status, how-to steps, pre-sale questions | Long investigations and emotional conversations |
Phone | Right away, after any hold | Urgent, complex, or upsetting problems | Cost, since each agent takes one call at a time |
Email | Hours | Detailed cases, documents, anything that needs a paper trail | Speed and back-and-forth questions |
In-house, BPO, or chatbot: who should answer your chats?
A chat can be answered by an in-house team, by an outsourced BPO team, or by a chatbot with people behind it. The right choice depends on how routine the questions are, which hours need cover, and how much product knowledge a reply needs.
What is BPO chat support?
BPO chat support is chat support staffed by an outside company. BPO stands for business process outsourcing. The provider's agents log in to the business's chat software and answer customers in the business's name.
Call centers that once handled only phone calls now sell this as call center chat support or call center live chat services. The agents work from scripts and the knowledge base the business gives them.
How to choose between in-house, BPO, and a chatbot
Use this table to match the staffing model to the need.
If you need... | Pick... | Why |
Replies that depend on deep product knowledge or account access | In-house agents | The people who know the product answer the question |
People on chat at night and on weekends | A BPO provider | The provider already runs those shifts |
Extra hands for a seasonal peak | A BPO provider | Capacity can be added and removed without hiring |
Instant answers to routine questions at any hour | A knowledge-trained chatbot | It answers from existing content with no queue |
Short waits and a person for hard cases | Hybrid: chatbot first, in-house or BPO agents behind it | Routine questions skip the queue and agents take the rest |
B2B chats about contracts, pricing, or technical setup | In-house agents, with a chatbot for documentation questions | These chats affect revenue and need context |
What to ask a call center chat provider
Ask a call center chat provider these questions before signing.
- Which hours and languages are covered, and by how many agents?
- How is the service priced: per agent hour, per chat, or per month?
- What first response time and satisfaction score does the contract commit to?
- How are agents trained on the product, and who updates their scripts?
- Which chats come back to the in-house team, and how?
- Who can read chat transcripts and customer data, and where is that data stored?
Best features of chat support software
The best features of chat support software are the ones that shorten the wait and keep context when a chat changes hands. Eight matter most.
Feature | Why it matters |
Chat routing | Sends each chat to a free agent or the right team |
Saved replies | Agents answer common questions with one click |
Answers from a knowledge base | A chatbot or agent pulls the answer from existing content |
Handoff with the transcript | The customer does not repeat the question to the next person |
Queue position or wait estimate | Customers can decide whether to wait or come back later |
Satisfaction survey | Shows which chats went badly |
Reporting | Tracks first response time, queue wait, and dropout |
Multiple languages | Customers get replies in the language they write in |
Teams that serve several countries can read how AI handles the language feature in multilingual customer service.
The stack: which tool does which job
A working chat setup has three layers, and one tool rarely covers all of them. The table shows which tool does which job.
Layer | Job | Tool |
1. Live chat inbox | Agents read and reply to chats, with routing, queues, and saved replies | |
2. Knowledge-trained answer layer | Answers routine questions from the website, help center, and documents, then escalates to a person | |
3. Staffing | People who pick up escalated chats | An in-house team or a BPO provider |
SiteGPT is the answer layer. It is a chatbot trained on a business's own content. It is not a live chat suite and not a BPO service. It has no agent queue or shift planning. Teams whose live chat inbox is the legacy Zendesk Chat product can see what replaces it before the 2028 end date in the Zendesk Chat alternatives guide.
Escalation to a person is on every SiteGPT plan. By default the chatbot escalates when the visitor asks for a person, when the situation is urgent, or when the visitor is frustrated after an unhelpful answer. The team then replies from SiteGPT's chat history, and the iOS app on Growth and above sends a push notification.
The limits matter for this setup. Starter has no chat integrations. Slack, Crisp, Messenger, Google Chat, Freshdesk Messaging, and Zoho SalesIQ start on Growth, and Zendesk needs Scale. There is no Intercom or HubSpot integration. Teams whose inbox is Intercom can read the Intercom Fin AI Agent pricing review to see what its native AI agent costs per outcome.
SiteGPT plan | Billed monthly | Billed yearly | Messages per month | Chat integrations |
Starter | $59 per month | $39 per month ($468 a year) | Up to 4,000 | None |
Growth | $129 per month | $79 per month ($948 a year) | Up to 10,000 | Slack, Crisp, Messenger, Google Chat, Freshdesk Messaging, Zoho SalesIQ |
Scale | $429 per month | $259 per month ($3,108 a year) | Up to 40,000 | Growth integrations plus Zendesk |
The message figures are for GPT-4.1-mini. The same plans give one tenth as many GPT-4.1 messages. Starter on yearly billing costs $468 a year against $708 on monthly billing, a saving of $240 or 34%. Every self-serve plan has a 7-day free trial that requires a credit card, and there is no free plan.
For a side-by-side look at the inbox layer, see Crisp vs HubSpot vs Freshchat. For chatbots that pass conversations to agents, see AI chatbots with human handoff.
Common chat support mistakes to avoid
The most common chat support mistakes are about queues and handoffs. Four show up most often.
- Opening chat with nobody to answer it. A chat window with a long queue loses customers. The 27.4% dropout rate is the cost.
- Hiding the route to a person. Gartner's 2026 survey of 3,566 customers found 87% consider access to a human agent essential when a company uses GenAI in service.
- Letting a chatbot answer everything on day one. In the same survey, only 27% of customers said they would try a chatbot again after a bad experience.
- Tracking volume only. Chat counts say nothing about waits or dropouts. Track first response time, queue wait, dropout rate, and satisfaction.
Frequently asked questions
Basics
What is chat support?
Chat support is customer service delivered as typed messages in a chat window. The replies come in real time from a human agent, a chatbot, or both.
What is the difference between live chat and chat support?
Live chat means a person types the replies. Chat support is the wider term. It includes live chat, chatbots, and hybrid setups where a chatbot answers first and hands off to a person.
Are live chat support agents real people?
Usually, yes. In LiveChat's data, 90.2% of chats were handled only by human agents and 9.8% involved a chatbot. A well-run chat window tells the customer which one is answering.
Benefits and wait times
What are the benefits of live chat support?
The benefits of live chat support are a first reply in seconds, more conversations per agent than phone, a written transcript, help on the page where the question came up, longer coverage hours, and clear metrics.
Can chat support software reduce customer wait times?
Yes. Chat support software routes chats to free agents, gives agents saved replies, and lets a chatbot answer routine questions with no queue. The queue is where the wait happens. It averages 4 minutes 18 seconds, against 35 seconds for an agent's first reply.
Is live chat better than phone support?
Live chat is better for quick questions, order status, and how-to steps. Phone is better for urgent, complex, or upsetting problems. Most teams offer both.
Staffing and BPO
What is BPO chat support?
BPO chat support is chat support staffed by an outsourced provider. The provider's agents answer customers in the business's chat software, using the business's scripts and knowledge base.
How does call center live chat work?
Call center live chat works the same way as in-house live chat, with a different employer. The call center's agents log in to the chat tool, take chats from the queue, and pass cases they cannot solve back to the business.
How can a small team offer 24/7 live chat support?
A small team can offer 24/7 chat in two ways. A BPO provider can staff the night and weekend hours with people. A chatbot can answer routine questions at any hour and leave the rest for the team's next working day.
Chatbots and tools
What is the best chatbot for customer service?
The best chatbot for customer service depends on where it has to answer and what it has to know. The guide to a chatbot for customer service compares the options by use case.
Is SiteGPT a live chat tool?
No. SiteGPT is a chatbot trained on a business's own content. It answers routine questions and escalates to a person. Teams that need agent queues and routing pair it with a live chat or helpdesk tool. Small teams that want that inbox with a free plan can check plans and limits in the Tidio review.
Does B2B customer service need chat?
B2B customer service benefits from chat when buyers and users have technical questions during setup. Routine documentation questions suit a chatbot. Contract and pricing chats belong with an in-house person.
Sources
All sources were opened on October 4, 2026.
Written by
