TL;DR
Someone is going to ask for a number: calls per month, agents needed, budget to approve. Start with the honest answer, which is that it's never one figure, it's three. One is the baseline you're already handling, counted from your own records rather than rated on a satisfaction scale. The second is a multiplier tied to a specific growth event, not a general sense that things are about to get busier. The third is a range for the volume that shows up unplanned, stated as a rate per event rather than a monthly average. Work through those three in order and you can turn "we're going to scale" into a baseline, a multiplier, and a range you can defend in a budget meeting.
Fill In Your Volume Worksheet
Work through these five steps in order, using your own records, before you talk to anyone about staffing or price. Each step corresponds to a mistake teams make when they skip straight to a single number, explained in the sections below.
| Step | What to record | Your answer |
|---|---|---|
| 1. Current baseline | Calls per day or week, per use case, and how many people currently handle it | |
| 2. Split the baseline | Which of that volume is calendar-driven (scheduled campaigns, fixed cadence) and which is event-driven (incidents, complaints, ad hoc triggers) | |
| 3. Driver rate and range | For the event-driven portion, a rate per event ("each incident generates roughly N calls") and a low/high range of events, not an average | |
| 4. Upstream drivers | Known drivers outside your own team's control (a launch, a campaign, a digital push), how many are planned, and when | |
| 5. Final range | A low, expected, and high monthly figure, with the event-driven portion flagged separately |
Fill every "Your answer" cell from your own records before reading further. The rest of the page explains why each row matters, not how to guess it.
Ask For Calls, Not Satisfaction
Narrow the question you ask your own team down to a count, not a rating, before you try to estimate anything else. Teams asked a broad question about customer experience will answer with a broad description of experience; teams asked for a number will give you one, usually within the same breath.
A facility management provider serving corporate clients spent several minutes describing satisfaction, response time, and value for money before anyone asked the actual question. Once the question narrowed to call volume, the operations contact gave an estimate of about six calls a day on average, handled by a client-management team of one or two people. An edtech platform building course tools for universities and academies was also asked about their volume. Its representative stated outbound outreach at about fifty conversations a week, close to two hundred and fifty a month, run by one or two people doing B2B outreach, and that figure was used to build a pricing estimate that our team sent over after the call.
What both teams had was a small, steady, well-understood current load, and the only thing standing between that load and a usable baseline was a question specific enough to answer. If a baseline number isn't on the table within one short conversation, the question may still be too broad, though sometimes the underlying volume really is too variable to reduce to a single figure yet.
Don't Average Incident-Driven Calls
Where volume is triggered by events rather than a schedule, replace the monthly average with a driver rate and a range. A flat monthly figure assumes calls arrive steadily across the month, but incident-driven work doesn't behave that way. A quiet week produces almost nothing, and a bad week produces a spike that any average will understate on the days that matter.
The customer experience manager at a commuter shuttle service discussed monthly campaign volume for pricing purposes rather than committing to a flat monthly call figure. Some of her team's calls are triggered by road incidents, on top of the scheduled onboarding calls, surveys, and check-ins that already run on a calendar. On a bad day, five incidents on the road mean five separate outbound campaigns have to go out that same day, regardless of what the rest of the month looks like. She agreed to supply a range instead of a single figure, and that range, not an average, was the right thing for a vendor or budget owner to be asking for from the start.
A monthly average smooths away the week that actually breaks your staffing plan.
Replace it with a rate: state that each incident generates roughly a fixed number of calls, then multiply that rate by the low and high number of incidents you actually see in a normal week, never by one assumed number.
Turn Ambition Into A Multiplier
When a known growth event is coming, force the direction of change into a number before it reaches a vendor or a budget line. Knowing that volume will rise is not the same as knowing by how much, and the gap between those two is where estimates stall.
A Nigeria-based finance house illustrated the mistake cleanly. Its chief operating officer described a small team of three to four staff handling all inbound and outbound customer communication today, and called that volume manageable. She also described a planned nationwide push to onboard far more customers, after which, in her words, the business would "need to communicate at scale." When scaling up, ensure you define what "communicating at scale" means in terms of calls, agents, or minutes. The ambition was clear. The multiplier was not.
The fix is mechanical rather than clever. Once you can name the event, ask how many times current volume it will produce, and keep asking until someone commits to a number, even a rough one. An intention without a multiplier cannot be priced, staffed, or defended in a budget meeting. It can only be deferred.
Model The Upstream Handshake
When another team's calendar generates your calls, build that handshake into the estimate instead of extrapolating from your own history. Contact centre volume is often treated as a simple function of customer count, but where marketing or digital activity drives the contact, the real schedule lives in someone else's calendar, not yours.
A retail bank's digital lead described this dependency directly. The bank wants to scale its large customer base significantly using digital channels, and worried openly about sustaining contact centre operations if headcount has to grow in step. He explained that digital campaigns, like account opening promotions, generate interest, questions, and technical issues that flow straight into the contact centre. The centre's workload is set partly by another team's launch calendar, not by its own growth curve.
Ask the team that owns the upstream activity for their own calendar before you build yours. Their launch dates are your peak days, and no amount of historical call data from your own team will show you a campaign that hasn't run yet.
Insight7 helps operations teams turn exactly these event-driven and upstream patterns into a documented baseline, multiplier, and range instead of a guess: https://insight7.io/book-demo/
Hand Over A Range
Give whoever prices or staffs a low, expected, and high figure, and say out loud which part of it is event-driven. A single number invites a false sense of precision, and the moment it turns out wrong, in either direction, it stops looking like an estimate and starts looking like a mistake.
A single number isn't precision, it's a guess wearing a decimal point.
The range only earns its keep if you explain why it's wide. Say plainly that the low end assumes a quiet month and the high end assumes a bad one, and name which part of the range moves with incidents, campaigns, or launches rather than with your team's steady workload. Without that explanation, the first bad week reads as a forecasting failure instead of the event-driven swing you already flagged, and the next number you hand over earns less trust, not more.
Questions You'll Get Asked
How do I estimate call volume for something that hasn't launched yet?
Borrow the driver rate from the closest comparable event you can actually measure, a smaller regional campaign or a past onboarding push, and apply it at the new scale. Treat the result as the high end of your range until the first real event gives you an actual number to correct it with.
What counts as event-driven versus normal volume?
If the call would not exist without a specific trigger, an incident, a campaign, a launch, a complaint, it is event-driven. If it happens on a predictable cadence regardless of what else is going on that week, a scheduled check-in or a recurring survey, it belongs in the calendar-driven baseline. Mixing the two into one average is what produces a number nobody can defend.
Does automating part of the workload change how I should estimate volume?
Not the method. What changes is which part of the range you can afford to let swing, because automated capacity can flex with a bad week in a way a hiring plan cannot.
What if whoever is pricing this refuses to accept a range?
Hand over the expected figure as the headline number and attach the low and high as a single flagged sentence underneath it, not as three competing figures. That satisfies the request for one number without hiding the swing that will matter the first time volume actually moves.
One Number Isn't The Goal
Once you stop defending a single figure, the conversation with whoever prices or staffs your team changes shape. You are no longer arguing about whether the number was right. You are explaining which part of it moved and why, which is a conversation any operator can win. If you want help turning your own baseline, multiplier, and range into something a vendor or a budget owner can act on immediately, book a session at https://insight7.io/book-demo/.
