Repeat Contact Rate: Why Customers Chat Twice

Repeat contact rate is the share of customers who have to get in touch again about the same issue after you thought it was handled. It is one of the most honest support metrics, because it measures what the customer experienced rather than what the agent recorded. A chat marked “solved” that brings the same person back two days later was not really solved.

This guide explains how to define the metric so it means something, how to measure it in a small team without special software, what usually causes customers to chat twice, and the practical changes that bring the number down.

What Repeat Contact Rate Measures

The idea is simple: of all the issues you closed in a period, how many came back? The details are where teams go wrong, so fix three things before you count anything.

Same customer, same issue

A customer who asks about delivery today and about a different product next month is not a repeat contact. That is a loyal customer. Count only contacts about the same issue, or so closely related that the first answer should have covered it.

A time window

Pick a window after the first contact is closed, commonly 7 days for e-commerce and up to 14 or 30 days for technical products where problems take longer to resurface. Use the same window every month, or the trend means nothing.

Any channel

A customer who chatted on Monday and e-mailed on Wednesday about the same thing is a repeat contact, even though the channel changed. Measuring chat alone will understate the problem.

The formula, then:

Repeat contact rate = issues with at least one repeat contact in the window ÷ all issues closed in the period × 100

Its mirror image is first contact resolution, the share of issues resolved in a single contact. The two are related, but counting repeats is easier to measure honestly, because it does not depend on the agent deciding whether the issue was resolved.

Why Customers Chat Twice

When you read repeat contacts side by side, the causes cluster into a few groups. Each one points to a different fix.

The answer was incomplete

The customer asked how to return an item and got the address, but not the deadline or the label instructions. They come back for the missing piece. This is the most common cause in small teams, and the easiest to fix.

The answer was wrong

An agent or an assistant gave information that did not match the policy, or that was out of date. The customer acted on it, it did not work, and now they are back and less patient.

The promise was not kept

“I’ll check with the warehouse and get back to you tomorrow.” Tomorrow passes without a reply, and the customer writes again. Broken follow-up promises create repeat contacts that are also complaints.

The customer did not know what would happen next

The issue was handled correctly, but nobody said when the refund would arrive or that the replacement would ship separately. The customer comes back to ask, not because anything went wrong, but because nothing was said.

The root cause was never fixed

A confusing setting, a broken link or a product defect keeps generating the same question from different customers and sometimes from the same customer twice. Support answers it every time. Nobody fixes the cause.

How to Measure Repeat Contact Rate in a Small Team

Large contact centres use dedicated analytics for this. A team of one to ten people can measure it well enough by hand, in about an hour a month.

Take a sample. Pick 50 to 100 conversations closed in a given week.
Look up each customer’s history. In Talkmio, the Contacts list keeps everyone who ever wrote to you, with their conversation history, so you can open each customer and see whether they came back within your window.
Mark each issue. Repeat about the same issue: yes or no. Note the cause from the list above.
Calculate the rate and the cause split. The rate tells you how big the problem is. The causes tell you what to do.

Search helps here: the Inbox can be searched by name, e-mail, text or ticket number, which makes it quick to find a returning customer’s earlier conversation. On the €49 plan and above, reports can be exported as CSV if you prefer to work in a spreadsheet.

Do this monthly with the same window and the same sample size. A hand-counted metric measured consistently is more useful than an automated one nobody trusts.

A Worked Example

A small online shop samples 80 chats closed in one week and checks each customer’s history seven days forward. It finds 12 issues with a repeat contact about the same thing. The rate is 12 ÷ 80 = 15%.

The cause split is more useful than the headline:

Cause	Repeat contacts	Fix
Did not know what happens next	5	Add timescales to saved replies
Incomplete answer	3	Rewrite the return-policy reply
Follow-up promise not kept	2	Convert “I’ll get back to you” chats into tickets
Wrong or outdated answer	1	Update the shipping page the assistant reads
Root cause in the product	1	Pass to the product owner

These numbers are an illustration, not a benchmark. What matters is the shape: most repeat contacts come from two or three fixable causes, and the biggest one is often simply not telling the customer what happens next.

How to Bring Repeat Contacts Down

Each cause has a direct fix. Work on the largest one first.

Answer the next question too

Before closing a chat, ask yourself what the customer will need to know next. If they asked how to return something, the next questions are the deadline, the label, and when the money comes back. Put them in the same reply. Saved replies are the easiest place to build this habit: in the Talkmio Inbox, typing / inserts a quick reply, so a well-written reply that already includes the next steps is used every time. Our guide to canned responses that sound human has examples.

Always state what happens next and when

End every non-trivial chat with the next step and a timescale: “The refund goes out today and usually appears on your card within five working days.” A customer who knows when to expect something does not need to ask.

Turn promises into tickets

If you promise to come back, track it. Convert the chat into a numbered ticket, set its status to Pending, and give the customer the number. On Pro and higher plans, the e-mail channel lets the follow-up go by e-mail in the same thread. Our guide on tickets vs live chat explains when to switch.

Keep the knowledge base current

If an AI assistant gives an outdated answer, the fault is usually in the source page. Talkmio’s assistant answers only from your own content, so correcting the page and re-reading it corrects the answer. A quarterly check of the pages behind your most common questions prevents a lot of wrong answers. Our article on auditing AI chatbot accuracy describes a simple review routine.

Fix root causes, not just answers

When the same question keeps coming back, from the same customer or many, take it out of support. A clearer setting, a fixed checkout step or a better product description removes the contacts entirely. Keep a short list of these and review it with whoever owns the website or product.

Repeat Contacts and AI Assistants

An AI assistant changes the picture in two ways. It can lower repeat contacts by answering instantly and consistently from your policy, at any hour, so customers do not have to wait and ask again. It can also raise them if it answers too confidently from incomplete content, because a wrong answer almost guarantees a second contact.

The safeguard is grounding and handoff. Mio answers only from your own website, FAQ and documents, and it hands the conversation to your team when it cannot answer, when the visitor asks for a person, complains, or asks about a specific order. When reading repeat contacts, note whether the first contact was answered by the assistant alone or by a person. If the assistant’s answers bring more people back, look at the source content behind those answers first.

Talkmio’s reports show the share of conversations answered by Mio alone and the number of handoffs, which helps you see how much of your volume each side handles. The repeat contact check tells you how well each side handles it.

How It Fits With Other Support Metrics

No single number describes support quality. This metric is most useful next to a few others:

First-reply time tells you how fast you start. A low first-reply time with a high repeat rate means fast but shallow answers.
Resolution rate tells you how many issues you close. Our guide to chat resolution rate explains it. If resolution is high but repeat contact is also high, issues are being closed too early.
Customer ratings after a chat tell you how the customer felt in the moment, a narrow slice of overall customer satisfaction. Repeat contact tells you whether that feeling held up a few days later.

Watch for one trap: pushing agents to close chats quickly can lower handle time and raise repeat contacts. The customer pays the difference.

Common Mistakes When Tracking Repeat Contacts

Counting every return visit. Customers who come back with new questions are not a failure. Count only the same issue.
Changing the window. A 7-day window one month and 30 days the next makes the trend meaningless.
Measuring chat only. If the second contact arrives by e-mail or social media, it still counts.
Blaming agents. Most repeat contacts come from policies, saved replies, missing content and broken processes. Fix the system first.
Chasing zero. Some issues genuinely need two contacts, for example when the customer has to send a photo. Aim for a steady downward trend, not perfection.

Frequently Asked Questions

What is a good repeat contact rate?

There is no universal target, because it depends on your products, your window and how you count. Measure your own rate consistently each month and aim for a steady decline. The cause split is more useful than any benchmark, because it tells you which fix will remove the most repeat contacts.

What is the difference between repeat contact rate and first contact resolution?

First contact resolution is the share of issues resolved in one contact. Repeat contact rate is the share of closed issues where the customer had to come back. They are near mirror images, but counting repeats is easier to measure honestly because it is based on what the customer did, not on how the agent marked the chat.

What time window should I use?

Seven days suits most e-commerce and simple service questions. Technical products, where a problem may take longer to reappear, often use 14 or 30 days. Choose one window that fits your business and keep it fixed, so month-to-month changes reflect real improvement rather than a different measuring stick.

How can I measure repeat contacts without special software?

Sample 50 to 100 closed conversations, open each customer’s history and check whether they came back about the same issue within your window. Note the cause each time. In Talkmio, the Contacts list keeps everyone who wrote with their conversation history, which makes this check quick.

Do AI chatbots increase or decrease repeat contacts?

Both are possible. A chatbot that answers instantly and consistently from accurate content reduces repeat contacts. One that answers confidently from incomplete content increases them. Ground the assistant in your own pages, set clear handoff rules, and check repeat contacts separately for chats the assistant handled alone.

The Bottom Line

Repeat contact rate shows how often your answers fail to finish the job. Define it carefully, as the same customer, the same issue, a fixed window and any channel, then sample it by hand each month and record the cause behind each repeat. Most of the gains come from answering the next question in advance, always stating what happens next, tracking promises as tickets and keeping the content behind your answers current. To run support where AI answers from your own pages and every conversation history is in one place, start free at app.talkmio.com.