Call Queues and
Ring Groups for a Brokerage: Routing That Matches How You Work

Where a call lands is an operational decision with regulatory weight attached to it. A client who rings to say their claim has gone quiet for three weeks has, in ASIC's terms, expressed dissatisfaction. Under ASIC's regulatory guidance on internal dispute resolution, that expression triggers the obligation, and it triggers it at the moment the client says it. The client does not have to use the word "complaint". They do not have to put anything in writing. And the obligation is not deferred until someone forwards the matter to whoever handles complaints in your firm. The relevant timeframes begin when the firm receives the complaint, not when the matter is later referred internally to whoever handles complaints.

That single point changes how a phone tree should be read. A routing design is not only a convenience layer. It shapes who is likely to hear that sentence in the first place, whether they recognise it for what it is, and whether the acknowledgement ASIC expects within one business day is set in motion.

Claims sit in a similar position. Since 1 January 2022, claims handling and settling has been within the AFS licensing regime. Whether a particular person handling a particular call is providing a regulated claims handling and settling service depends on the nature of what they do and the licensing arrangements in place, and ASIC's guidance sets out qualifications and exceptions. The operational point for a routing design stands regardless: claims traffic can carry work that needs suitably authorised and competent people, so it is worth routing deliberately rather than treating it as overflow from the real work.

Then there is the service standard the profession sets for itself. NIBA's Code of Practice, monitored independently by the Insurance Brokers Code Compliance Committee, commits subscribers to handling complaints promptly and fairly and to putting client interests first. None of that is delivered by a policy document. It is delivered by the person who answers, and whether that person was the right person to answer.

The practical position for most brokerages is that all four of those call types arrive on one published number. New business, renewals, claims and general service enquiries land in the same place, ring the same handsets, and get triaged by whoever is free. That approach has a real advantage worth naming: it is simple, it needs no maintenance, and nothing falls through a badly built menu. For a small team in one room it can work well for years.

What changes is scale. As headcount grows and the team stops sitting within earshot of each other, the informal triage that made a shared line work stops being available. Nobody can hear who has picked up what. The person best placed to handle a premium dispute is on another call, and the call goes to whoever was nearest instead. At that point the routing design starts doing work that used to be done by proximity, and it is worth designing deliberately rather than leaving as it was set up on day one.

Ring groups and call queues are often treated as the same feature, and they behave quite differently under load.

A ring group rings multiple extensions at once on a single incoming call, and anyone in the group can answer it. On MaxoTel's hosted system the groups are configurable, so who is in a group and how the group behaves is set by the firm rather than fixed. The value is speed. One call, several phones, first available person takes it.

A call queue does something else. It holds callers in order when everyone who could answer is busy, and it manages the waiting. MaxoTel's queuing supports custom hold audio and playlists, interactive menu breakouts, and an automatic callback option so a caller can request a call back from the next available person instead of holding. Multiple queues can be configured with different audio and different menus, and calls can be routed to different agents or departments based on date and time rules.

The distinction matters at the point where a brokerage is busiest. A ring group with nobody free does not queue anyone. It rings out and falls through to whatever the timeout is set to, which for many firms is a shared voicemail. A queue with nobody free holds the caller, tells them something, and offers them a way out that keeps the contact alive. Both are useful. They solve different problems.

What a single shared ring group does to a mixed call load is worth being precise about. Missed calls are usually not the problem. What tends to happen instead is that every call gets treated as the same kind of call.

A renewal conversation with a premium increase in it needs the person who wrote the original cover and knows the claims history. A new business enquiry needs someone who can qualify and quote, and needs them quickly, because that caller is very likely working down a list. A claims call may need someone suitably authorised to act on it. A certificate of currency request needs almost nobody at all, and can be handled by whoever is free in under two minutes.

Route those four through one group and the allocation is effectively random. The new business enquiry may reach the person deep in a claim file. The claims call may reach the newest member of the team. The certificate request may interrupt a broker mid-renewal. Nothing is dropped. The work simply lands in the wrong hands often enough that the cost shows up as rework, callbacks and slower resolution rather than as missed calls.

There is a second cost that is easier to miss. When every call comes through one path, every call looks the same in the reporting afterwards. Call volume can be counted. Call type cannot. A firm in that position can tell you Tuesday was busy without being able to tell you what Tuesday was busy with.

What the reporting can tell you later is largely decided when the phone system is set up, rather than at the point you go looking for an answer.

Routing problems tend to come from a small number of situational causes, and most of them are the natural result of a system that grew rather than one that was designed.

None of these are failures of management. They are what happens when a phone system is treated as infrastructure that either works or does not, rather than as a design that expresses how the firm wants to work.

A routing map starts by naming what actually arrives. The four call types below are offered as a working model rather than a measured distribution, and they cover the traffic many brokerages will recognise. Each one has a different answer to the questions of who should take it, how fast, and what happens if they cannot. Your own tally may divide differently, which is the point of step 1 in the next section.

The pattern across all four is the same. Match the routing behaviour to what the call needs, then decide the fallback before you need it. A routing map is really a set of answers to one question asked four times: if the first person does not pick up, then what?

A well-built routing map generates two kinds of data, and they answer different questions.

The phone system produces the traffic picture. MaxoTel's queue statistics report on wait times, staff performance and abandoned calls, and show which times of day and which days of the week are busiest, which is what you need to adjust staffing. That is genuinely useful, and it is the right tool for the question of how many calls arrived and how quickly they were answered.

Where queue reporting is strong is traffic. What it does not reach is content. Queue statistics can tell you that 40 calls came through the claims path on Thursday and that the average wait was 90 seconds. They will not tell you that nine of those callers were ringing for the second time about the same claim, or that four of them expressed dissatisfaction in terms that meet ASIC's definition of a complaint.

Callyx.ai works on the recordings those calls produce. Every recorded call is transcribed and analysed automatically, without anyone listening to a recording, and the subject of each conversation is identified from what was said on it. That gives you call type derived from the content of the call itself, which is a separate signal from the path the call arrived on.

Call routing is one of the few operational levers in a brokerage that can be adjusted in an afternoon and felt the next morning. Ring groups get someone on the line fast. Queues hold the line when everyone is busy and give the caller a way out that does not end the contact. Time switches cover the hours the map assumes. Used together and mapped to the four call types a brokerage actually handles, they turn a phone system from a shared front door into a design that reflects how the firm works.

The reporting benefit follows the routing rather than leading it. Once calls arrive on separate paths, the traffic data they generate can be read by path, and the recordings they produce can be analysed for what was actually said on them. That is the point at which Callyx.ai adds the layer the phone system cannot: what was said on those calls, by call type, across every conversation rather than a sample. Queue reporting already covers how many arrived and how long they waited. The routing work is worth doing on its own. It also happens to be what makes everything downstream of it legible, and it is a direct contributor to the customer experience clients actually receive.