Transfer Rate: Why Calls Bounce Around and How to Stop It
How to measure transfer rate, tell good transfers from bad ones, find the causes of bouncing calls and fix routing, training and handover habits.
Few things annoy a caller more than being passed around. Each transfer adds waiting, risks a dropped call and forces the customer to explain the issue again. The transfer rate shows how often it happens. Like most metrics it is neither good nor bad by itself, but an unexpected change tells you something has gone wrong in routing, training or process.

This guide covers how to measure transfers, separate necessary ones from avoidable ones, find the causes and improve. It sits next to First-Call Resolution and Repeat Call Rate, and is influenced by menu design (see IVR Menu Design).
What to measure
At its simplest, the rate is calls transferred divided by calls handled. To make it useful, record three more things about every transfer: where it came from, where it went and why.
| Field | Why it matters |
|---|---|
| Origin (queue or agent) | Shows which team receives calls it cannot handle. |
| Destination | Shows which team is the usual answer. |
| Reason | Separates routing mistakes from legitimate escalations. |
| Number of hops | A call transferred twice or more is almost always a problem. |
| Type (warm or cold) | Shows whether the caller's context travelled with them. |
Do not stop at an overall figure. A moderate overall rate can hide one queue where most calls are bounced. Break it down by queue, by call reason and, carefully, by agent.
Good transfers and bad ones
A transfer is not a failure by default. Some calls need a specialist, a supervisor or someone with authority to approve a decision. The goal is to make sure that transfers are deliberate and few, not to ban them.
Necessary and avoidable transfers
| Necessary | Avoidable | |
|---|---|---|
| Reason | The issue needs a specialist or an approval | The caller reached the wrong team |
| Cause | Expected design of the service | Menu, routing or knowledge gap |
| Caller experience | Warm hand-off with context | Repeats details to a new person |
| Hops | One | Two or more |
| Fix | Make it smooth | Remove the cause |
Avoidable transfers are the target. They usually fall into a few groups, which are easier to attack as groups.
Typical causes of avoidable transfers
- Menu mismatchCallers pick the option closest to their issue, but it leads to the wrong team.
- Skills gapsAgents could solve the issue with a little more information or authority.
- Unclear ownershipTwo teams each believe the other handles a topic.
- Hours and capacityThe right team is closed or busy, so calls are moved to whoever is free.
- HabitAgents send hard calls on instead of owning them.
Finding the causes
Numbers narrow it down; reading calls explains it. Take the queue with the highest transfer rate and read or listen to a dozen transferred calls. For each, ask: should this have gone to the first team? What did the first agent need to solve it? Could the menu have sent the caller directly to the destination?
A transfer review in four moves
- Pick the pairChoose the origin and destination with the most transfers between them.
- Sample the callsRead the summaries or transcripts of a dozen of those calls.
- Name the reasonWrite a short cause for each, such as "billing question asked in the technical queue".
- Choose the fixChange the menu, route that topic directly, train the first team or document the answer.
If the same cause appears in most of the sample, you have found your first project. If every case is different, the rate is probably healthy and you should look elsewhere.
Fixing the causes
Fixes depend on the cause, but most fall into three groups.
Routing and menu design
If callers pick the wrong option, test the wording. People choose by their own words, not your department names. Rename options after the customer's problem, not your org chart, and put the most common reasons first. See Common IVR Mistakes for traps to avoid, and What Is ACD? for ways to send each call to the right group of agents.
Knowledge and authority
When the first team could solve the problem with more information, write the answer down where agents can find it. Give front-line staff clear rules about what they may decide, such as small refunds or rebookings. A decision tree in the knowledge base removes many "let me pass you to someone" moments.
Handover practice
Some transfers will always be needed. Make them painless with a short routine.
Warm transfer routine
- Tell the caller why you are transferring and who they will speak to
- Check that the destination is open and has someone free
- Brief the next agent in a sentence or two, including what has been tried
- Stay on the line until the caller and the next agent are connected, if your system allows
- Record the transfer reason in the call notes
- If the transfer fails, offer a callback instead of dropping the call
Setting a sensible target
Because some transfers are legitimate, avoid a flat target such as "under a fixed number". A better approach is to set a goal for avoidable transfers only, or to target a trend, such as a lower rate on the pair of teams that bounce calls most. For the logic behind choosing numbers, read Setting Realistic KPI Targets.
Beware of side effects. If agents are judged on transfers alone, they may keep calls they cannot solve, raising handle time and repeat calls. Look at transfers together with resolution, repeats and caller feedback.
Review the numbers monthly, with the two teams in the same room, and keep a short list of agreed changes. A transfer problem is rarely fixed in one place; it usually takes a small adjustment on each side of the hand-off.
Transfers and routing in Call
In Call, calls arrive in queues and the agent can pass a call to a colleague or another team, or call a colleague first before handing over. Queue routing, business hours and thresholds decide where a call goes before an agent ever sees it, so tuning them is the first place to look for avoidable transfers. When an AI voice agent hands a call to a person, the handoff is designed to give the person context, which reduces the questions the caller has to answer again; the how-to articles for the AI agent hand-off describe the settings. The call history keeps summaries and transcripts that make the sampling step above straightforward. See Setting Up Call Queues.
Frequently asked questions
What is a normal transfer rate?
It varies with the type of business. A team that handles simple questions will transfer rarely, while one that triages for specialists will transfer often by design. Compare against your own history and focus on avoidable transfers.
What is the difference between warm and cold transfers?
In a warm transfer the agent speaks to the next person first and passes on context. In a cold transfer the call is simply connected, and the caller may have to start again.
Can a menu change reduce transfers?
Often it can. If many transfers go from one queue to one other queue, a new menu option or clearer wording may send those callers directly to the right place.
Should I discourage transfers entirely?
No. A good transfer to a specialist is better than a wrong answer. The aim is to remove transfers that exist because of routing mistakes or missing information.
How do I find calls transferred more than once?
Record the number of hops in your call notes or reports and review any call with two or more. These are the clearest examples of a process that needs fixing.
Who should own the fix?
The team lead of the queue receiving the wrong calls, with help from whoever owns the menu. Share the findings with both teams so each understands the other's constraints.