Eight Capacity Planning Questions Every Multi-Location Contact Center Should Be Able to Answer
Your VP, CFO, or COO will ask at least one of these. Here's what a confident answer looks like and what it means if you can’t.
Multi-location contact center operations face a version of capacity planning that network-level reporting doesn't prepare you for. Volume shifts, site misses, new locations, allocation decisions – each one tests whether your plan reflects what's actually happening at each site. Run through them, see where you hesitate.
The Questions
Are we working to each location’s planning assumptions, or to a single network average
Which of our locations can realistically absorb more volume next quarter?
The plan looked right, but the site missed its capacity requirements. What happened?
If we shift volume to a lower-cost location, what does that actually save us?
How do we plan when some locations are captive and others are outsourced?
We’re adding a new location. What does the ramp plan look like, and when does it become productive?
How do we compare allocation options before committing to one?
Couldn’t we just build this in Excel or Google Sheets?
1. Are we working to each location’s planning assumptions, or to a single network average?
[Avatar: VP of Operations or COO – Operations Accountability]
This question surfaces after a site miss, during a QBR, or when headcount variances get escalated. Occasionally the CFO gets involved if the conversation moves into cost accountability. The honest answer tells you whether you’re working from a structural plan or an approximation.
Confident answer:
“We plan at the site level. Each location has its own shrinkage, average handle time, new hire learning curve, and attrition rate built into the plan. When we roll up to the network total, it’s an aggregation of individual site plans, not a shared assumption applied uniformly.”
If you can’t answer it:
Your plan is likely built on blended inputs. Start by pulling each of those four variables for every site and compare them to your network assumption. Where they diverge is where your plan is carrying untracked risk.
2. Which of our locations can realistically absorb more volume next quarter?
[Avatar: VP of Operations or VP of Sales – Growth/Commercial]
Volume growth decisions (whether from new business, seasonal peaks, or client expansion) shouldn’t be made by gut feel. A confident answer requires a view of each site’s available capacity, not just its current headcount.
Confident answer:
“Based on our current capacity plan, [Site X] has room to absorb [X] FTE of additional demand. [Site Y] is running close to its ceiling. If we’re taking on new volume, we route it to [Site X] first. [Site Y] only comes in if [specific condition].”
If you can’t answer it:
You’re likely routing volume based on historical patterns rather than a live capacity model. Get each site’s planned headcount, required headcount, and seat capacity into the same view. The gap between them, positive or negative, tells you what each location can actually carry.
3. The plan looked right, but the site missed its capacity requirements. What happened?
[Avatar: VP of Operations or End Client – Post Mortem Review]
This is the question that lands in your lap after a bad period. If you can’t trace it to a specific input, the explanation defaults to “forecast error” which rarely leads anywhere useful.
Confident answer:
“We reviewed the plan inputs for that site. [Site X]’s actual shrinkage came in at [Y]% against a planned [Z]%. That’s a [X] FTE difference. The plan was built on the right logic, the assumption is what needs revisiting.”
If you can’t answer it:
You’re likely reviewing performance at the network level, which obscures site-level causes. You need planned assumptions alongside actuals for each site, not just at the network level. Without that, you can’t tell whether you’re looking at an execution miss or a planning input that was off from the start.
Extended Reading: For the structural reason this happens, see The Headcount Gap Hiding Between Your Multi-Location Sites
4. If we shift volume to a lower-cost location, what does that actually save us?
[Avatar: CFO or COO – Cost Reduction Proposal]
Rate-card arithmetic makes every shift look attractive: hours moved times the rate difference. It’s the calculation a spreadsheet produces in seconds, and it’s usually wrong because it treats an hour of demand as identical everywhere.
Confident answer:
“The receiving site carries its own handle time, shrinkage, and occupancy profile, so the same volume requires [X] more FTE there than at the sending site. Once we account for the true FTE requirement and the ramp period, the landed saving is [X]%, not the [Y]% the rate card implies. It may still be the right move, but at a different number and on a different timeline, and we make the decision on that number.”
If you can’t answer it:
You’re pricing the move on rate difference alone. Model the shift using the receiving site’s own operating assumptions rather than the sending site’s, then compare fully loaded cost against fully loaded cost. The spread between rate-card savings and landed savings is often the difference between a sound decision and an expensive one that looked sound in a spreadsheet. Cost is one input. Service quality, and customer and employee experience are harder to model but carry real operational risk if the shift goes wrong.
5. How do we plan when some locations are captive and others are outsourced?
[Avatar: VP of Operations or COO – Mixed Delivery Model]
Captive and outsourced sites don’t just carry different costs, they respond to different levers. You control hiring and scheduling at a captive site; you control volume commitments and contract terms at a BPO. A plan that treats them identically will eventually pull the wrong lever.
Confident answer:
“Both in-house and outsource sites sit in the same plan, but each has its own cost structure and constraints. Captive sites carry fully loaded labor, hiring lead times, and training pipelines. Outsourced sites carry contracted rates, minimum volume commitments, and ramp obligations. When volume moves between them, we see the capacity impact and the contractual impact in the same view, including whether a shift takes a BPO partner below its committed floor.”
If you can’t answer it:
You’re likely running two disconnected plans or one plan with blended assumptions that describes neither operation. Bring both into a single model with their actual cost structures. Pay particular attention to minimum-commitment clauses: shifting volume away from an outsourced site below its floor rarely saves what the simple math suggests.
6. We’re adding a new location. What does the ramp plan look like, and when does it become productive?
[Avatar: VP of Operations or COO – New Location Setup]
A new site doesn’t arrive at full productivity; it earns it, cohort by cohort. The meaningful question isn’t when the site opens; it’s when it becomes net productive against its cost, and who carries the volume until then.
Confident answer:
“The ramp plan models hiring lead time, training duration, and the learning curve for each hiring cohort. [Site X] starts taking volume in week [Y], reaches [Z]% proficiency by month [A], and becomes net productive in month [B]. Until then, the network plan explicitly carries the overflow at existing sites; that capacity is reserved, not assumed.”
If you can’t answer it:
You’re probably applying a steady-state productivity assumption from day one, which overstates early capacity and understates early cost. Model each cohort’s learning curve separately and roll it into the site plan. The gap between assumed and actual productivity in the first two quarters is where new-site business cases quietly fail.
7. How do we compare allocation options before committing to one?
[Avatar: VP of Operations or COO – Before Making Big Decision]
Most multi-location decisions aren’t “can we do this?” They’re “which version of this is better?” The organizations that decide well are the ones that can put the options side by side before anyone is anchored to one.
Confident answer:
“We run the options as parallel scenarios, same demand, different allocations, with every site keeping its own assumptions in each one. We compare required FTE, fully loaded cost, and service risk across scenarios before anything is committed. The decision becomes a comparison, not a debate.”
If you can’t answer it:
Scenario comparison in a spreadsheet means duplicating the workbook, and duplicated workbooks drift. A correction made in one copy never reaches the others. If comparing three allocation options takes three days of rebuilding, the organization will default to whichever option was modelled first. Speed (and accuracy) of comparison is itself a planning capability, and it’s often what determines whether the right option actually gets considered.
8. Couldn’t we just build this in Excel or Google Sheets?
[Avatar: CFO or COO – New Platform Evaluation]
An honest question deserves an honest answer: for a single site with stable assumptions with few queues/skills, a well-built spreadsheet can work, and many capable planners have run one for years. Multi-location is where it structurally breaks; not because planners lack skill, but because the problem changes shape.
Confident answer:
“Three things break at network scale. First, structure: each site needs its own shrinkage, handle time, attrition, and learning curves, and spreadsheets push you toward blended averages because maintaining separate assumption sets by hand doesn’t scale. Second, comparison: every scenario needs a copy, and every copy drifts. Third, the math itself: spreadsheet capacity models rest on averages or constrained Erlang-style formulas, which are fine for what they were built for, single-skilled queues in steady state, but a multi-location network with mixed sites, ramping cohorts, and shifting volume isn’t that. A simulation engine models how the network actually behaves with multi-skilled agents, site by site and interval by interval, then rolls it up without blending anything away.”
If you’re doing it anyway:
Keep one canonical workbook under version control, document every site-level assumption and its source, and reconcile plan against actuals by site every period. And know the threshold: once you’re past two or three sites, carrying a captive/outsourced mix, or making reallocation decisions more than occasionally, the spreadsheet stops being a planning model and starts being the largest unmanaged risk in the plan.
Where To Go From Here
If you worked through these and hesitated on one or two, the gap isn't in your planning skill; it's in what your planning environment is built to support. Most multi-location operations have at least one of these they've been working around without naming it. Naming it is the first step.
If you'd like to talk through what closing that gap looks like for your operation, we're happy to have that conversation.
