Is live chat worth implementing? Five concerns and honest answers

The question ‘is live chat worth implementing?’ is almost always asked by someone who already has a specific concern. The issue is not whether website conversations can be effective, but whether they will work for this business, with this team and this volume of enquiries. Here are the five concerns we hear most often.

For each one, we also explain when it is justified and when live chat is better left disabled. This is not an article about choosing a tool or about implementation itself—if you have not yet chosen a channel, start with our comparison of live chat, chatbots and callbacks. Here, we focus solely on the reasons businesses hesitate.

1. ‘We will not be able to reply in time’

This is the most common concern, and it is partly justified. Usually, however, it is about something other than what people assume. The problem is rarely typing speed itself—the problem is a promise that cannot be kept. A widget promising an instant response at 10 p.m. when the business closed at 4 p.m. does more harm than having no chat at all.

The answer is not to insist that the team will cope. It is to narrow the promise to what the team can actually deliver. Availability schedules are configured separately for each website, so chat can appear only during hours when someone is genuinely watching the panel. Outside those hours, offline mode takes over: the customer fills in a short form, and their message becomes a ticket with a number and an event timeline. Nobody waits for an answer that will never come, and the message does not disappear into someone’s private inbox.

The second part of the answer is measurement. Instead of debating whether the team is keeping up, open the first-response-time report and inspect the longest cases rather than the average. We explain how to calculate this metric within business hours and what actually reduces it in our article on first response time in live chat.

When this concern is justified: if nobody in the business can commit even two fixed hours a day and nobody agrees to own incoming tickets, live chat should not be enabled. An unattended widget and an offline mode that nobody checks produce the same result as having no channel—only with added customer disappointment.

2. ‘Our customers prefer the phone’

Sometimes that is true. More often, it is a conclusion drawn from the only data the business has—and the business has data only about people who called. Someone who visited the website, failed to find a quick answer and left will not appear in any report. With evidence like that, the conclusion will always be that customers prefer the phone. The same goes for the belief that customers prefer WhatsApp; our guide to WhatsApp Business vs website chat shows how to check it against where your enquiries actually come from.

Instead of guessing, you can check this with your own data and very little effort.

1. Count what comes in today.Break down last month’s contacts into phone calls, emails and forms. This is a baseline, not an answer.
2. Check when they arrive.If many enquiries arrive in the evening or at weekends, the phone was not the customer’s choice. It was the only option during hours when nobody would answer anyway.
3. Ask directly during calls.For two weeks, end phone calls with one question: would it have been more convenient to write? The answers usually form a clear pattern.
4. Enable chat on one page.Choose the highest-traffic page and use a visibility rule to limit the widget to that page alone. A month is enough to see whether anyone writes.
5. Compare cases, not counts.One written request for a quote may be worth more than ten calls about opening hours. Look at the type of case, not just the counter.

When this concern is justified: if you sell services where decisions are made by talking—negotiating scope, arranging an on-site appointment or handling matters that require back-and-forth questions—chat will not replace a conversation. A callback then makes more sense: the customer leaves a number and chooses from server-generated time slots, while availability outside business hours is configured separately. It addresses the same need through the channel best suited to it.

3. ‘We have nobody to handle it’

This concern is usually based on an image of a call centre: someone wearing a headset and staring at a screen for eight hours. That is not how it works in a small business. Live chat is not a separate full-time job—it is another notification alongside the email inbox, only with a shorter expected response time.

↔ Scroll the table sideways to see all columns.

The real question is not ‘do we have someone to sit there?’ but ‘when do enquiries actually arrive?’ Their hourly distribution rarely matches the rhythm of the office. We explain how to build that distribution and use it to calculate staffing in our article on customer-service staffing.

When this concern is justified: if the only person who could reply works in the field and has no access to a computer for most of the day, live chat would be a fiction. The same applies when that person already handles sales, fulfilment and invoicing, and every additional item on the list means something else stops working. In that case, it is more honest to stick with the phone and a form.

4. ‘We are worried about difficult conversations and negative reviews’

This concern has two layers. The first is that someone may write something unpleasant. The second is that the reply will be recorded and could later be shown to someone else. Both are real, but they lead to opposite conclusions, so it is worth separating them.

• You have time to respondUnlike on the phone, nobody expects a reaction in the same second. You can check the facts, ask a colleague and only then reply.
• There is a recordThe record works both ways. It protects you from disputes about what was promised, but it will not forgive a sentence written in anger.
• You can standardise the toneDifficult situations recur in a handful of forms. Prepared response templates keep the standard consistent regardless of who is on duty.
• Escalation is visibleA conversation that exceeds an operator’s remit becomes a numbered ticket with an event timeline, instead of ending in silence and a promise to call back.

It is worth separating the emotions in the conversation from the process behind it. Difficult cases are usually lost not because of tone, but because they have no owner or deadline. We cover the operational side in our article on handling complaints through written channels.

A negative review rarely stems from the problem alone. Much more often, it comes from the customer not knowing what is happening with their case. A ticket number and a promised response time do more here than the most carefully chosen words.

When this concern is justified: if the business currently has an unresolved problem—delayed fulfilment, a defective batch or a backlog of refunds—chat will become a queue of people the operator cannot help. Opening a new channel at that point will not calm the situation; it will move it somewhere every visitor can see. Fix the source first, then open the conversation.

5. ‘We already have a contact form’

A form and live chat may look like two versions of the same thing, but they serve different moments. A form assumes that the customer knows what they want to ask, can express it in one paragraph and is willing to wait. Live chat serves situations where the question is short and the answer determines whether someone stays on the website for another five minutes.

A form is entirely sufficient in several fairly common situations.

• Enquiries are long by nature: a brief, a specification or a case description with an attachment.
• The customer must provide a complete set of details before anyone can start preparing a quote.
• The answer requires preparation and nobody can give it immediately, regardless of the channel.
• The volume of enquiries is low enough that nobody waits more than a few hours for a reply.
• The form has an owner, and every submission ends with a confirmation and a response deadline.

If all these conditions are met, adding chat will change little except creating another place that needs attention. It makes more sense to refine what you already have: the number of fields, the wording of error messages and what happens after the button is clicked. We cover these points in our guide to website contact forms.

A form stops being sufficient in a different scenario: when questions are short and arise while people compare offers. Few people will fill in six fields just to find out whether the price includes installation. They will either find the answer elsewhere or leave without a trace—and contribute nothing to the statistics beyond a single page view.

When this concern is justified: in project-based sales, where every enquiry begins with a brief anyway and the first conversation happens only after the materials have been reviewed. There, chat would be little more than a more convenient gateway to the same form.

Before enabling chat: conditions you need to meet

After these five concerns, one question remains: are the conditions met in your case? The list is short and deliberately says nothing about the tool—all the points concern the business, not the software.

• It is clear who answers messages, and that person has been explicitly told.
• You know the hours when someone is genuinely at a computer, and those are the hours entered in the availability schedule.
• Outside those hours, chat does not promise a reply; it accepts a message and turns it into a ticket.
• Every ticket has a number, an owner and a visible event timeline, so it does not depend on anyone’s memory.
• You can check first response time without asking anyone on the team.
• You have agreement that, for the first month, chat will serve one page rather than the entire website.

If three of these points are not met today, implementation can safely wait. Not because chat is a bad tool, but because website conversations expose gaps in customer service faster than any other channel. A form forgives a day’s delay. Chat does not forgive fifteen minutes of silence while the status says someone is available.

The worst possible implementation result is not a lack of conversations. It is conversations that begin and are abandoned—the customer writes, sees that nobody is responding and leaves with a worse opinion of the business than when they arrived.

If the conditions are met, the decision stops being a matter of faith in the channel. Enable chat on one page, during hours the team can cover, and give it a month. After a month, you will have your own data instead of someone else’s arguments—and that data will answer the question in the title.

Once you launch, do not judge the outcome after the first quiet week. We explain what to watch at the beginning in our article on the first month after implementation.