← marketplace-ops-toolkit
marketplace-ops-toolkit · tool #5

Workforce Forecast Calculator

How many agents do you actually need? Multi-case-type forecast based on volume × complexity × productivity × shrinkage. Compares 100% in-house vs vendor-split scenarios and recommends the most cost-efficient mix. The math behind quarterly capacity planning.

Pick a scenario preset

Pre-loaded patterns for common ops planning situations. Click to load, then customize.

Case types

Break the queue down by case complexity. Each case type has its own daily volume and average handle time.

Case type name
Daily volume
Avg handle time (min)
Notes

Operating parameters

Productivity assumptions, target service level, and team economics. Adjust to match your operation.

h/day
%
×
%
$
$

Required headcount

The math: total case-minutes per day ÷ productive minutes per agent ÷ (1 − shrinkage) × peak ratio × (1 + service level buffer).

Total Headcount Needed
0
Adjust inputs above to forecast.
Total case minutes/day
0
Productive minutes/agent
0
Base FTE (no shrinkage, no peak)
0
After shrinkage adjustment
0
After peak coverage
0
After service level buffer
0

Per case type breakdown

How each case type contributes to the total. Useful for vendor split decisions — typically high-volume / lower-complexity cases go to vendors, gray-zone and high-stakes stay in-house.

Case type
Daily volume
Daily minutes
FTE share

In-house vs vendor split — cost comparison

Three staffing models with annualized cost. The recommended option is starred — usually a hybrid because gray-zone cases need in-house judgment but high-volume cases are cheaper at the vendor.

How this calculator works

The classic workforce forecasting question — "how many agents do we need?" — has a deceptively simple math at the core. The complication isn't the formula. It's that most teams pick numbers out of the air for the inputs and then defend the forecast politically rather than analytically.

This calculator forces explicit inputs across the five drivers that actually matter:

Real-world impact

Illustrative scenarios drawn from operator practice. Numbers are realistic order-of-magnitude estimates, not measurements from any specific deployment.

Case 1: Finance asking for next year headcount in 5 days
SetupHead of Ops at a regional fintech needs to justify a 2026 plan covering 4 case types (KYC, fraud review, ATO, disputes) with different complexity and volume curves.
ProblemExisting spreadsheet model was 14 tabs of stale assumptions, last touched 8 months ago, produced a single headcount number with no scenario range, putting roughly $1.8M in payroll asks at risk of being slashed.
Tool surfacedMulti-case-type calculator separated AHT and volume per type, layered shrinkage (22% real vs 15% assumed) and peak month uplift, produced in-house vs vendor split economics side by side.
OutcomePlan went in with three scenarios (lean, base, peak-ready) and finance approved base case at 31 FTE plus 8 vendor seats, vs the original ask of 38 FTE that would have been rejected.
Case 2: 60% vendor-heavy footprint that was bleeding cost
SetupCOO at a high-growth ecommerce platform had 24 in-house and 36 vendor reviewers across two sites, locked into a per-case vendor rate set 3 years prior.
ProblemVendor cost per case had crept to roughly $4.20 vs in-house fully loaded $3.10 once productivity and shrinkage were applied honestly; the platform was spending an extra $360K per year on the worse option.
Tool surfacedVendor split scenario showed crossover point at 45% vendor, not 60%, given current AHT and shrinkage; rebalancing recovered the spread without losing peak flexibility.
Outcome12 cases per day shifted in-house, vendor headcount renegotiated to a smaller flex pool, net savings around $230K annualized in year one with peak coverage preserved.

The formula

total_minutes = Σ (volume_i × handle_time_i) // sum across case types productive_min = productive_hours × 60 // per agent per day base_FTE = total_minutes / productive_min after_shrink = base_FTE / (1 − shrinkage) // PTO, training, sickness after_peak = after_shrink × peak_ratio // peak vs avg day final_HC = after_peak × (1 + sl_buffer) // service level cushion

What each input means

The vendor split decision

For most operations, a hybrid in-house/vendor model is more cost-efficient than pure in-house, but only for the right case types. The general rule:

The split calculator above assumes 60% of total volume can go to vendor at the vendor cost rate, and 40% needs in-house. Adjust those assumptions to your operation — they vary a lot by industry, case type, and your vendor capabilities.

What this calculator is NOT