Repeat Call Rate: Finding the Problems Behind Second Calls
Learn how to measure repeat call rate, separate healthy follow-ups from failures, and trace second calls back to causes you can fix at the source.
Every call a customer should not have needed to make is waste for them and for you. The repeat call rate is a simple, reliable signal that waste exists. It counts how often a customer calls again about something you believed was handled, and it quietly hides inside your call volume where nobody notices.

This guide shows how to define a repeat call, measure it from ordinary call records, sort the causes and act on them. It pairs naturally with First-Call Resolution, which is the same question seen from the first call's side.
Defining a repeat call
A repeat call is a contact from the same customer about the same topic within a window you choose. Both parts need a decision.
Deciding what counts as a repeat
- Identify the customerMatch calls by caller number, or by contact record when your system links them. Anonymous calls will always be undercounted.
- Choose the windowSeven days is a common starting point for simple issues. Longer processes such as deliveries or repairs may need a longer window.
- Match the topicCount only calls about the same issue. A customer who calls about billing and then about a new order is not a repeat.
- Exclude planned follow-upsA callback you scheduled or a visit that needs a status check is expected, not a failure.
- Record the resultKeep the count and a reason for every repeat you review.
The rate itself is repeat contacts divided by total contacts in the same period. Report it by topic as well as overall: an overall figure can look steady while one topic is badly broken.
Healthy and unhealthy repeats
Not all second calls are bad. Some are normal, some are signs of a broken process, and the response is different for each.
Types of repeat calls
| Usually healthy | Usually a problem | |
|---|---|---|
| Planned follow-up | Customer calls back on the date you agreed | Customer chases because nobody called as promised |
| New information | A new question after the first was answered | The same question, because the answer was unclear |
| Complex issue | A multi-step case that needs several contacts | A simple issue that needed several contacts |
| Outcome | Customer has what they need after the second call | Customer is more frustrated each time |
Be careful here, because the label depends on context. A customer who calls three times about an incomplete delivery is a different story from one who calls twice to book an appointment and then to change it.
Common causes behind repeat calls
When teams review repeat calls, the same causes surface again and again. Grouping your own findings into these buckets makes the pattern easier to see.
Where repeat calls come from
- Unclear answersThe agent answered, but the customer left unsure what happens next.
- Promises not keptA callback, email or document that did not arrive on time.
- Incomplete fixesThe immediate symptom was treated, but the cause remains.
- Missing informationThe website, a confirmation message or a letter leaves out a key detail.
- Handover gapsThe case moved between people or teams, and context was lost.
- Process delaysThe issue is waiting on a step the customer cannot see.
Tagging each repeat with one of these makes a simple ranking. The goal is a short list of reasons that account for most of the repeats, and often two or three causes dominate. Fixing a confusing confirmation message may remove more calls than any coaching programme.
Choose the window with the customer's patience in mind. A caller who has been told "we will update you in three days" is not a failure on day four. A caller who has been told nothing may reasonably call after a day. Setting the window per topic, and writing it down, keeps arguments out of the weekly review.
Measuring from the records you already have
You do not need a special tool to begin. A call log with the caller number, date, queue and a short outcome note is enough. If your calls are summarised or transcribed, you can also search for phrases that signal a repeat, such as "I called yesterday", "I was told someone would call" or "still not working".
A reasonable routine for a small team:
- Once a week, export or list calls from the last seven days.
- Group by caller number and mark numbers that appear more than once.
- Remove planned follow-ups and calls on different topics.
- Read the remaining repeats and assign each a cause from your list.
- Count the causes and bring the top three to the weekly team meeting.
- Choose one fix, note the date and compare the next few weeks.
If your volume is large, sample rather than reading everything. A random set of fifty repeats is more useful than a thorough review of ten.
A hypothetical week of repeats
Imagine a small repair business that reviews thirty repeat calls from one week. On reading them, twelve customers wanted a status update that nobody had promised to give, eight were confused by the wording of the booking confirmation, six called because a promised callback never came, and four were genuinely new questions. The team does not need to coach anyone. It needs to send a status message at the halfway point, rewrite the confirmation, and assign an owner to the callback list.
That is the real value of the metric: it converts a vague feeling that "customers keep ringing" into three concrete changes with a start date. Re-run the same review a month later and compare the reasons, not only the rate.
Acting on what you find
Match the fix to the cause. Coaching is the right answer when agents explain things badly; it is the wrong answer when a form is confusing.
- Unclear answers: add a closing step to every call, repeating what will happen next and when. See Call Openings and Closings.
- Promises not kept: review how follow-ups are assigned; the habits in Callback Completion Rate apply.
- Missing information: rewrite the message or web page that causes the question, then check whether the calls drop.
- Handover gaps: use warm transfers with a short summary, and look at Transfer Rate alongside repeats.
- Knowledge gaps: feed recurring questions into your knowledge base; Finding Knowledge Gaps from Calls explains how.
Always measure the effect. If the repeat rate on one topic does not move after a fix, the cause was probably different from what you assumed.
Share the results with the people who own the causes. The website owner, the person who writes confirmation messages and the team that runs deliveries all hold part of the answer, and they rarely see call data unless someone brings it to them. A short monthly note listing the top three reasons for repeats, with two or three real quotes, usually gets faster action than a chart.
Tracing repeats in Call
In Call, every call lands in the call history with a summary and transcript, and the contact record collects the calls, notes and appointments for one person. That is enough to see whether someone has called before. Transcript search lets you look for phrases that mark a repeat, and the Reports area gives intervals and patterns that show when repeat contacts cluster. Duplicate contact records can split one person's calls across two entries, so merge duplicates in CRM setup before you count. For a wider picture of how phone data fits together, read Call Analytics Dashboard.
Frequently asked questions
What is a normal repeat call rate?
There is no universal figure, because it depends on what you sell and how complex the issues are. Compare against your own history and track the direction after each fix.
How long should the repeat window be?
Seven days is a common start for simple issues. Use a longer window for processes that take weeks, and keep the same window so results stay comparable.
Is a low repeat rate always good?
Not always. If customers give up and go elsewhere instead of calling again, the rate looks good while the experience is poor. Read it with satisfaction and effort measures.
Can I measure repeats without a CRM?
Yes. A spreadsheet of call dates, numbers and short notes will do for a small team. A CRM or call history makes it faster and more accurate.
Should agents be judged on repeat calls?
Use it to find causes, not to rank people. Many repeats come from process and information problems that no agent controls.
How do I handle callers who withhold their number?
Anonymous calls cannot be matched, so your measured rate is a lower bound. Ask for a reference or contact detail when it is appropriate and useful to the customer.