The basic formula
Annual savings ≈ tickets × handle time × hourly cost × deflection rate. Each input has an industry benchmark you can use as a starting point — but the real numbers come from your own ticketing system.
Channel mix matters
Chat deflection (1.00 modifier) is significantly higher than email (0.85), voice (0.65), or social (0.90). A 70/20/10 chat/email/voice mix produces dramatically higher ROI than a 30/30/40 mix at the same volume.
Build vs buy on ROI grounds
For most companies under 100K monthly tickets in non-regulated industries, SaaS wins on year-3 TCO. The break-even point shifts to custom build when complexity > 30%, regulated industries, or 6+ languages.
What good looks like
Mature deployments hit 60-80% deflection on tier-1 tickets within 6 months. Anything outside that band is either an over-promise (vendor) or an under-deployment (you).
Callout
How to use this calculator
Run the wizard with your real numbers; the free preview shows the savings range, deflection percentage, and payback period. Sign in to unlock the full breakdown, hidden-cost warnings, and a board-ready PDF.
Try it on your numbers
Plug in your real volumes and see the projection.
The calculator runs the same methodology described above — channel modifiers, hidden costs, and the build-vs-buy verdict. Free preview without an email.
Run the calculatorFrequently asked
Is the projection conservative?
Yes - we use vendor-published mid-points and apply a 15% conservative haircut for residual escalation, AI runtime overhead, and supervision drag.
Why is voice deflection lower than chat?
ASR error rates and intent ambiguity cap voice deflection at roughly 65% of chat effectiveness. Voice AI is improving rapidly but not yet at parity.
What counts as a simple ticket?
FAQ, account lookups, status updates, password resets - the 70-80% AI-deflectable bucket.
How accurate is the payback period?
It uses your actual cost inputs; expect ±20% accuracy on the mid estimate when ticket volume and handle time are reasonably well-known.
Will this work for B2B?
Yes. B2B tier-1 deflection patterns are similar to B2C, though absolute volumes are typically lower. The methodology applies at any volume.
Can I edit the assumptions on the results page?
Yes. Every assumption is editable and recomputes live in the browser - no server round-trip.