Methodology

This page exists so you can check our work. If a number here is wrong or stale, we want to hear about it.

Constants

Constant	Value	Basis
Productive hours per shift	5.5 h	8 h shift minus breaks, meetings, training and idle time.
Shifts per FTE	4.6	Covers 7 days x 8 h of response window, with 1-2 weeks annual leave absorbed.
E-commerce AHT	7–12 min	Order status, returns, shipping, payment.
SaaS AHT	9–18 min	Billing, onboarding, how-to, bug triage.
Marketplace AHT	5–9 min	Listing and seller support.
Services AHT	8–15 min	Scheduling, quotes, client questions.
Automation ceiling — e-commerce	62%	Share of tickets that are repetitive enough to resolve without a human.
Automation ceiling — SaaS	55%	Lower because of account-specific and technical questions.
Automation ceiling — marketplace	50%	Disputes are more bespoke.
Automation ceiling — services	48%	Scheduling and quotes benefit most, but scope varies widely.

Formulas

Tickets per agent per day: floor(5.5 x 60 / handle time)
Baseline FTE: ceil( (monthly tickets / 30) / capacity x coverage factor )
Coverage factor: max(1, coverage hours / 5.5)
Automation coverage used: min(entered, industry ceiling)
Agents after automation: ceil( human tickets / 30 / capacity x coverage factor )
Net monthly effect: headcount saved - automation platform cost
Payback: annual platform cost / net monthly effect
Year-1 return: gross annual savings / annual platform cost

Where the ranges come from

Handle-time distributions and automation ceilings are compiled from published support-industry benchmarks and practitioner surveys. They are deliberately conservative and stated as ranges rather than averages, because a 12-minute AHT shop and a 4-minute AHT shop both exist and the difference is roughly a factor of three in headcount.

We do not claim these are precise. They are defensible planning figures. The single most valuable thing you can do is measure your own handle time for two weeks and re-run the model with the real number — that is why every input on the calculator is editable.

Limitations

• Seasonality, launch spikes and product launches are not modelled.
• Out-of-hours and weekend ticket distribution is averaged, not weighted.
• Quality, CSAT and lifetime-value effects of automation are excluded from the money figures.
• We do not model the cost of migrating and maintaining your help centre.