Renewal Season Call Load: Planning Capacity From Last Year's Data

Where a brokerage has a stretch of the year when the phones run heavier, the fact of it is not news to anyone who has worked through one. What is harder to pin down is its shape: which weeks run heaviest, which days inside them carry the most, and which hours are where callers stack up. That detail is the difference between a plan and a brace.

The record that describes it already exists, produced as a by-product of running the phone. On MaxoTel the call history export holds every inbound and outbound call on the service with its type, time, origin, destination, status, duration and cost, downloadable as a CSV file over any date range. Where the account holds 12 months of it, which depends on the retention the plan provides, that describes last year's peak week by week in a file that takes a few minutes to pull. A shorter history still supports the day and hour cuts below. It is the year-on-year comparison that needs the full span, so it is worth checking what the account actually retains before planning the work around it.

Cover decisions for a peak period can end up being made close to the event, which is a reasonable response to the information usually to hand: a general sense that the next six weeks will be heavy is actionable, and for many firms it is what is available. What deciding late can cost is range. Weeks out, routing changes, extended hours and a reallocation of who handles what are generally all in play. Days out, the practical choices often narrow, and how far they narrow depends on how quickly a firm can change a roster and a routing configuration.

What makes the timing commercially interesting is where renewal calls sit in a brokerage's year. The 2026 Vero SME Insurance Index, which surveyed 1,500 Australian businesses in September 2025, found that roughly four in five use a broker, that 42 per cent of those broker users have been with the same broker for more than three years, and that broker satisfaction stabilised at 69 per cent, down from a high of 87 per cent in 2024. The report measures the relationships. What follows from them operationally is that a relationship of that length comes back round every renewal, and a renewal is usually where a premium movement gets explained. A call that goes unanswered in an ordinary week is an inconvenience.

The same call during renewal weeks behaves more like a retention event, which is the argument set out in the piece on missed calls in a brokerage.

A forecast does not make the peak smaller. It buys the lead time to choose a cheap lever over an expensive one, and moves the conversation from how busy everyone feels to how many hours the busiest week is short.

The forecast is three cuts of one export, and none of them needs software the brokerage does not already have.

Volume by week across 12 months. Pull a full year of call history, filter to inbound, and count calls per calendar week. Index each week against the median week rather than reading raw counts, because an index travels: a week at 1.6 means the same thing next year even when the portfolio has grown. Carrying that index forward on a portfolio growth rate assumes call demand scales with the portfolio, which is a starting assumption rather than a known relationship, and client mix, policy count and renewal timing can all break it. This cut identifies the peak window and, just as usefully, its edges. The ramp often starts earlier and tails off later than the internal shorthand suggests.

Volume by day of week inside the peak window. Filter to the peak weeks only and count by weekday. The shape of a peak week is not always the shape of an ordinary one, and it may concentrate on the days following a renewal notice going out, which is a pattern the brokerage controls rather than inherits.

Volume by hour inside the peak days. The same filter, counted by hour of day. This is where a roster is actually set. The view is described in more detail in the piece on what your call history data says about your brokerage, which treats volume by hour and weekday as the first view worth building from any export. Run across a renewal period rather than a flat month, it becomes a capacity plan.

The step that turns three charts into a decision is converting the gap into hours, because call counts do not compare to anything a roster is written in.

Illustrative example: converting a forecast into hours. Every figure below is an assumption rather than measured data, included to show the arithmetic. Assume a median week of 400 inbound calls and a heaviest week last year of 640, an index of 1.6. Assume the portfolio has grown 10 per cent since, so the baseline is now 440 and the equivalent peak week forecasts at about 704 calls, roughly 264 above baseline. Assume an average call duration of nine minutes and six minutes of work created by each call, giving a full cycle of 15 minutes. On those assumptions the gap is about 3,960 minutes, or 66 hours, in one week. Across a team of six that averages roughly 11 hours each, which is a team-wide average rather than an individual rostering requirement: the load is concentrated in whichever days and hours the second and third cuts identified, so how it can actually be spread is a separate question. The arithmetic also assumes the extra calls carry the same 15-minute cycle as the rest, which is worth checking against call type before the result goes near a roster. Every assumption here moves on a firm's own numbers, which is the reason to run it on the real export rather than carry this example into a plan.

Sixty-six hours in a week is a useful number precisely because it is awkward: too large to absorb quietly, too small to hire against. That is the size of problem routing, hours and task allocation are built for. Two sense checks before the forecast goes anywhere: compare last year's peak index against the year before where the history reaches that far, and check it against what the business already knows is changing, such as a book acquired or a large account lost. The export describes what happened, not what is scheduled to.

A call history export answers volume, timing and reachability questions well. Several of the questions a capacity plan asks sit outside it, and knowing which saves rebuilding the forecast later.

A forecast that lands weeks ahead opens three levers a forecast landing on the day does not. All three are configuration rather than payroll, and each changes the size of the gap the next has to absorb.

Queue settings also carry a queue weight, from one to ten, so renewal calls can be presented to available staff ahead of general enquiries for the duration of the peak and set back afterwards.

The sequence matters, and it runs roughly from most reversible to least. Routing changes are usually configuration, so they can be undone, though some may involve setup time or vendor assistance worth checking first. Hours changes touch rosters and can carry a cost in money or goodwill depending on how they are arranged. Reallocating who handles what costs attention and needs a conversation with the team. Hiring is generally the slowest of the four and the hardest to reverse, which is why it is better reached by elimination than by default. Where a gap survives all three levers and shows up outside the peak window as well, that is a structural capacity question rather than a seasonal one.

The peak is also the best measurement opportunity of the year. MaxoTel's queue statistics are built to show how quickly calls are being answered and which times and days are busiest, so staffing can be adjusted against real demand. Reading them weekly during the peak rather than after it allows a routing change mid-season, and leaves next year's forecast with a measured peak rather than a remembered one.

Every lever above is a decision about load, and load is only half of what a capacity plan is protecting. A phone system reports a peak week accurately in its own terms: how many calls came in, how long people waited, how many were abandoned, how long each ran. None of that describes whether the renewal conversation that ran for 14 minutes on the Tuesday of the busiest week did its job. Suppose a routing change halves the wait and a reallocation puts a less experienced staff member on renewal calls. The queue report can show what both did to wait time, abandonment and answer rates, and that is a real measure of operational performance. What it does not carry is how the renewal conversations themselves went, which is the part that decides whether the second change cost anything.

Callyx.ai works on the calls a brokerage already records, so a capacity decision and its effect on service can be read from the same period.

Nothing about how brokers take calls changes, and no review task is added to anyone's day during the busiest weeks of it.

A renewal peak is one of the few demands in a brokerage year that can be anticipated, and the record describing last year's version of it is already in the phone system. Twelve months of call history, indexed by week and then cut by weekday and by hour, produces a forecast specific enough to roster against. Converting it into hours rather than calls is what makes it comparable to anything, and it is the step that lets a configuration change and a hire be weighed against the same number.

Three levers sit ahead of hiring. Routing decides how a surge is absorbed, through queue weighting, hold-time announcements, a callback option and overflow. Hours decide when cover is available, through scheduled overrides that end with the dates they cover. Allocation decides what broker minutes are spent on. They run from most reversible to least, and working through them in order keeps the slowest option from becoming the default.

A phone system with queue reporting switched on will tell you how heavy the peak was, how long callers waited and how many gave up, which on MaxoTel means the PBX statistics report and its plan availability. It will not tell you how the renewal conversations themselves went, which is the thing a capacity plan is ultimately protecting. That sits in the recordings, which is where Callyx.ai works: transcribing and analysing the calls a brokerage already holds, so the week that was busiest is also the week you can read.