A 70% AI resolution rate is measured on the conversations the bot handles, which are only part of your customer demand. This tool separates three rates, each with its own denominator: coverage (how much demand a bot could handle), resolution (how much of the bot's traffic it resolves) and durable automation (how much of your total volume goes away for good). It then values what is left at the cost that actually leaves your budget. Sometimes the program pays. Sometimes the claim is inflated, and the right move is to renegotiate, fix the foundation first, or buy nothing.
Three rates, each with its own denominator. The vendor's headline is apparent resolution. Your budget moves with durable net automation. The starting values are illustrative planning values and describe no particular operation or vendor. Replace them with your own figures.
The vendor's 65% resolution is a share of AI-involved conversations. Against your total demand it is 29.3%. A resolution rate means little until you know its denominator.
Freed agent time counts as $0 until you say how it becomes cash: reduced overtime, avoided hiring, less outsourced volume or fewer seats. Operating cost and the escalation premium are cash either way and are already counted. Choose the action above and the decision appears.
What decided this: net -$35,275 a month, upside case -$27,107 a month, eligibility 55%, evidence internal estimate or benchmark, capacity action not selected. Change any of them and the verdict can change.
Bound by evidence and realization. Monthly volume, AI-eligible demand, apparent resolution, repeat and false resolution and escalation premium are still at the tool default. Marginal cost and platform fee are still at the tool default. No capacity action is selected, so no freed capacity converts to cash. Net savings carry a plus or minus 25% band at this evidence level, -$26K to -$44K per month.
This grade is self-declared. It reflects what you have told this tool about your sources. No document, payroll file or pilot dataset has been reviewed here. Independent validation of the inputs is a separate step.
This is the one tool on the site that produces a realistic deflection rate. Two figures are now available to the other tools, and each has its own denominator.
Every subtraction, in order, starting from the vendor-style figure: the resolution rate applied to all volume at loaded cost. The steps reconcile exactly to net monthly savings.
The escalation premium is a directional starting value. At 0% the net would be -$13,700 a month. At 50%, double what you entered, it would be -$56,851. That is a $43,151 swing across the plausible range, so measure your own handle time on escalated contacts before you lean on this line.
Each scenario states the eligibility, resolution and repeat assumptions behind it, so you can see exactly what has to be true to reach it.
Start with the denominator. The quoted 65% resolution is measured on the conversations the bot takes part in. Against your total demand the bot removes 29.3% for good, because 55% of your volume is eligible and 18% of apparent resolutions come back. A finance team that approves against 65% and sees 29.3% in the results will notice the gap within a quarter.
You have not supplied a marginal cost per contact, so the tool assumed $4.20, sixty percent of loaded. Every dollar figure on this page moves in proportion to that number. It is the one input worth thirty minutes with your finance partner before you share this analysis.
At these assumptions the program never breaks even. Operating cost and the escalation premium exceed the realistic savings at any resolution rate. Either the platform cost is too high for this volume, or eligibility is too low and repeats too high to overcome.
You have selected no capacity action, so realized savings are $0. The only cash moving is $13,700 of operating cost and $21,575 of escalation premium, a monthly loss of $35,275. Freed handle time becomes money when you reduce overtime, slow hiring, cut outsourced volume or reduce headcount. Choose one, or present this as added capacity and leave savings out of it.
Year 1 net of -$369,366 carries no implementation cost, because you entered none. Bot programs are rarely free to stand up. Whatever the vendor charges for integration, content build and professional services comes straight off that figure and pushes payback later than shown.
Ramp sensitivity. Year 1 reads -$369,366 over a 6 month ramp and -$423,305 at full run rate from month one, a difference of $53,939.
Decision: Choose a capacity action first. Freed agent time counts as $0 until you say how it becomes cash: reduced overtime, avoided hiring, less outsourced volume or fewer seats. Operating cost and the escalation premium are cash either way and are already counted. Choose the action above and the decision appears.
Published figures to hold your result against. Each measures something slightly different, so read what it counts before you compare. None of them changes your result or its grade.
Three different questions asked of different people; none is a containment rate. Vendor-defined resolution rates are left out.
What centers handle and what customers say they prefer are different measures; they are shown apart.
The report downloads immediately. No email required, no wall.
Method 1.2, published 30 September 2026. How this is calculated
A person on our team sends it, usually within one business day. Your address is used to send this report and to reply if you ask a question. Nothing you entered leaves your browser unless you send it here, and the download never asks for it.
Each result raises a sharper question. This is the diagnostic that answers it.
Send this analysis for an independent review. You get a written read on what the numbers support, what they do not, and which diagnostic is worth running next. No sales call. No vendor introduction.