How to calculate AGV payback for a warehouse
A realistic formula that accounts for operational change, not just payroll. And why estimates often miss.
When a client first sees the price tag of an AGV package for 5—15 machines — $80—400K — they ask “how soon does it pay back”. The answer depends on what you count.
The simple formula “one robot replaces one worker = divide cost by annual payroll” almost always gives a wrong answer. Real economics is more complex — and more interesting.
TL;DR
- TCO is more than hardware: integration, training and service add 25—40%.
- Count savings in labour-hours, not “headcount” — robots replace operations, not people.
- Typical AGV payback — 12—24 months. Less than 12 is suspicious; more than 30 means revisit the configuration.
What TCO consists of
TCO (total cost of ownership) for a 5—15 machine AGV project breaks into four blocks. Robot hardware — usually 60—70% of budget. Mapping, WMS integration, training — another 15—25%. First-year service — 5—10%. Reserve for adjustments and consumables — 5—10%.
Clients often look only at the first block and are surprised when the final quote is 30% higher. That's normal for a project delivery — bake the full TCO into your estimate from the start.
What savings to count
The most common mistake — counting savings by “heads”. A robot doesn't replace a worker; it replaces operations. One warehouse worker does many tasks: picking, movement, inventory, checks. The robot takes only some.
The right approach — count manual labour hours on operations being automated: how many hours per day go to picking and inter-zone movement. Multiply by hourly rate, get annual savings.
Secondary effects — fewer picking errors, faster order assembly, ESG metrics. They exist, but in year one they're hard to measure, so leave them out of the base case.
Base formula
Payback (months) = TCO ÷ (annual savings) × 12.
Annual savings = (manual hours per day robots replace) × 365 × (warehouse worker hourly cost).
For a mid-complexity AGV project, typical result — 14—22 months. If it's under 12, recheck whether you overestimated hours. If over 30, the configuration probably doesn't match the warehouse scale.
Why estimates often miss
The most common reason — adaptation period gets ignored. The first 2—3 months after launch, productivity is below target: operators adjust, routes get tuned, small integration issues surface.
Second reason — operational change. For robots to work efficiently, you usually need to remark storage zones, change order assembly, sometimes restructure shifts. These changes need management attention, not just hardware.
If a client accepts that payback comes 3—6 months later than the spreadsheet says because of these factors, the project almost always lands within budget. If they expect “buy, switch on, count” — there'll be disappointment.
Want a concrete estimate for your case?
The guide answers “how” and “why”. Concrete “how much and how long” is a separate conversation. Start with the 2-minute audit or just write to us.