What could AI return for your team?
A realistic estimate built on fully-loaded labour cost, the automatable share of work, a conservative efficiency gain and a first-year adoption ramp — not back-of-napkin math.
Your scenario
Share of their work that is repetitive / rules-based
How much of that automatable time AI realistically reclaims
Scales your efficiency assumption — Expected uses it as-is.
One-time build, integration & rollout
Estimated return
Customer Support · Expected · 10.5 FTE reclaimed
Free consultation — no obligation
- • Fully-loaded cost = salary × 1.3 (benefits + overhead)
- • 1,800 productive hours / person / year
- • Year 1 ramped to 60% (rollout & adoption)
- • Annual run cost = 20% of implementation ($24.0K)
Methodology & sources
- Fully-loaded cost (×1.3): U.S. BLS — Employer Costs for Employee Compensation (benefits ≈30% of total compensation); we use a conservative 1.3.
- Working hours (1,800/yr): a standard 2,080-hour FTE (40h × 52) less typical leave, holidays and sick time; consistent with OECD — Average annual hours worked.
- Efficiency-gain ranges: informed by McKinsey — Economic potential of generative AI (2023), Brynjolfsson, Li & Raymond — Generative AI at Work (NBER 2023, ≈14%) and GitHub Copilot research. Presets use conservative mid-ranges.
- Adoption ramp & run cost:standard enterprise software TCO & change-management conventions (annual maintenance commonly ≈15–20% of build cost).
These are directional planning assumptions drawn from public research and industry standards — not guarantees. Actual results depend on scope, data quality and adoption, and you should validate them against your own data and the latest research. Zero Gravity Cybernetics accepts no obligation where inputs do not reflect your real situation.