Resources / agent business case 01

The Run-Cost Model Tool

Put a number on an agent proposal before anyone commits to building it. Four ways of doing the same job, costed side by side, with the operating lines most business cases leave out. The number it lands on is cost per acceptable outcome.

4 options · 7 sections · 83 lines · 1 reference example

Built by Sunil Mathew, co-authored with Claude

01 What this tool is for
  • You are asked to put a number on an AI agent proposal. This tool gives you that number, and it includes the costs most business cases leave out.
  • Four options are compared side by side over one period: a rules workflow, the same rules rebuilt on a written spec, a model that drafts while a person approves, and a full agent with tools.
  • Build cost is kept apart from run cost. Quality sits beside cost, so the number you argue about is cost per acceptable outcome, not cost per case.
  • The result names the option with the lowest cost per acceptable outcome, says whether each option ever pays back its build, and lists the assumptions to check before you trust it.
02 How to use it
  1. Set the workload and the rates first. Every other line is multiplied by them.
  2. Fill in each option column by column. Leave a line blank if you do not know it. A blank keeps the result off the screen; a 0 is a claim.
  3. Watch the bar at the top. It shows cost per acceptable outcome for each option as soon as enough lines are filled.
  4. Read the result at the end. It names the leading option, the break-even month for each, and the assumptions that decide the answer.
  5. Open the Reference example tab to see every line filled for a real-shaped case, and load it into your model to start from there.
03 Rules for using it in a budget conversation

Five rules that decide whether the number is evidence or a guess.

  1. Fill it from a trial, not from a guess. Escalation share, review minutes and the retry multiplier come from logs of real cases. A demo produces none of them. Nothing here is a benchmark: there is no industry number for your escalation rate.
  2. Argue about cost per acceptable outcome, not cost per case. Cost per case rewards a system that quietly declines the hard cases. Cost per acceptable outcome charges it for them.
  3. One definition of an acceptable outcome, one reviewer, all four options. The acceptable-outcome rate must come from one written definition, applied to every option by the same person. Otherwise the comparison is theatre.
  4. Blank means unknown. Zero means measured. A blank keeps the result off the screen until you know the number. Type a 0 only when you have measured a 0. "This system never retries" is a claim.
  5. Change the rows, keep the shape. Add a cost line your workload has and this model does not: data labelling, a vendor licence, a second reviewer for high-value decisions. The shape is the point: build kept apart from run, quality beside cost, and the operating lines visible.
per acceptable outcome
Not started 83 lines to fill. A blank is unknown, not zero.
Rules Rules v2 Assisted Agent
0/83 lines set
Rules workflow

Plain code that follows fixed rules. No model anywhere in the path.

Rules v2, on the written spec

The same code, rebuilt on the decision rules the agent build forced the team to write down.

Model-assisted draft

A model drafts a recommendation. A person approves every one before it takes effect.

Full agent with tools

A model plans, calls tools and acts. People review a sample of cases and the exceptions.

1

Workload and manual baseline

What does this work cost today, done by hand?

shared by all four options

Every option is measured against this. If the manual baseline is wrong, every break-even month below it is wrong by the same amount.

free text

Such as INR or USD. Shown in front of every money figure. Not counted as a line.

count

One case is one unit of work: a purchase order, a ticket, a refund request. Take the count from last quarter, not from the plan.

months

The window the build has to pay for itself. Twelve is the usual budget horizon. Make it longer only if you can defend the longer horizon in the room.

minutes

How long one person takes to finish one case today. Time it on ten real cases.

currency per hour

Fully loaded: salary, benefits, seat. Use the rate of the people who actually do the work today.

Total cases in the period

Cases per month multiplied by months.

Manual baseline cost for the period

Total cases, times minutes per case, times the hourly rate. This is the number every option has to beat.

2

Team rates

What does an hour of each person cost?

shared by all four options

Three rates, shared by all four options. Engineers build and handle escalations; reviewers approve output and finish declined cases.

currency per day

What one engineer-day costs you. Used for the build and for upkeep.

currency per hour

Used for escalations: the cases an engineer has to look at.

currency per hour

Used for reviewing output and for finishing declined cases by hand. Often the same rate as the manual baseline.

3

Build cost, paid once

What does each option cost before it handles its first case?

one value per option

Engineering days, converted at the day rate. Evaluation work is a separate line because it is the one most estimates skip, and it is where a model-based option earns the right to be trusted.

Rules Plain code. No model. Rules v2 Same code, rebuilt on the written spec. Assisted Model drafts. A person approves each one. Agent Model plans, calls tools, acts.
Engineering days to build days

From the first line of code to a system that handles real cases. Include integration with the tools it calls.

Engineering days for evals and a test set days

An eval is a set of real cases with known-correct answers that the system is scored against. Building one is engineering time.

Total build cost

Build days plus eval days, times the engineering day rate.

4

Inference and tool calls, per month

What does the model bill actually include?

one value per option

Inference is the cost of calling the model. Most business cases price the successful calls and stop there. The retry multiplier is where the rest goes.

Rules Plain code. No model. Rules v2 Same code, rebuilt on the written spec. Assisted Model drafts. A person approves each one. Agent Model plans, calls tools, acts.
Model calls per case, successful calls per case

How many times a case calls the model when everything works. An agent that plans, then acts, then checks its work makes several.

Retry and failure multiplier multiplier, 1.0 = no retries often left out

Total model calls divided by successful ones. Count timeouts, tool errors and loops that re-enter, not only refusals. 1.0 means the system never retries, which no real system does.

Cost per model call currency

Average cost of one call, including the input tokens. Take it from a month of real invoices, not from the price list.

Tool and API calls per case calls per case

Lookups, searches, writes to other systems. Anything with a price per call or a rate limit.

Cost per tool call currency

Average across the tools it calls. Internal calls are not free: price the compute.

Model cost per month

Cases per month, times calls per case, times the retry multiplier, times the cost per call.

Tool cost per month

Cases per month, times tool calls per case, times the cost per tool call.

5

Human in the loop, per month

How much human time does each option still need?

one value per option

Three kinds of human time. Review is routine checking of output. Escalation is a case that needs an engineer. Declined is a case the system would not finish, which a person then does by hand. This section decides most comparisons.

Rules Plain code. No model. Rules v2 Same code, rebuilt on the written spec. Assisted Model drafts. A person approves each one. Agent Model plans, calls tools, acts.
Share of cases a person reviews %

100 for a model-assisted draft, because a person approves every one. A sample, such as 20, for a full agent.

Review minutes per reviewed case minutes often left out

Time it on real cases. Approving a good draft is quick; checking an agent that acted on its own is not.

Share of cases escalated to an engineer %

Cases that stop and need somebody who can read the trace. Measure it in a trial. A demo will show you 0.

Engineer minutes per escalation minutes often left out

Reading the trace, working out what happened, fixing the case. Longer for an agent, because there is more trace to read.

Share of cases declined or abandoned % often left out

Cases the system refuses or gives up on. They do not disappear: a person finishes them, at the manual rate.

Manual minutes per declined case minutes

Usually close to the manual baseline, because the person starts from nothing.

Review cost per month

Cases per month, times the share reviewed, times minutes per review, at the reviewer rate.

Escalation cost per month

Cases per month, times the share escalated, times minutes per escalation, at the engineer rate.

Declined-case cost per month

Cases per month, times the share declined, times manual minutes, at the reviewer rate.

6

Upkeep and operations, per month

What does it cost to keep the system trustworthy after launch?

one value per option

A rules workflow needs almost none of this. A model-based option needs all of it, every month, and none of it appears in a demo.

Rules Plain code. No model. Rules v2 Same code, rebuilt on the written spec. Assisted Model drafts. A person approves each one. Agent Model plans, calls tools, acts.
Observability and eval tooling currency per month often left out

Tracing, logging, the eval runner, the dashboards. Licence fees or the compute to run your own.

Eval maintenance engineer days per month often left out

Engineering days a month spent adding cases to the eval set and fixing the ones that go stale.

Model-version re-qualification engineer days per year often left out

Engineering days a YEAR spent re-running the eval set after a model or SDK update. Spread across the months by the model. Count every update the provider ships.

Prompt and regression testing engineer days per month often left out

Engineering days a month spent checking that a prompt change did not break something that used to work.

Incident and on-call load engineer days per month often left out

Engineering days a month on incidents caused by the system. Take it from the on-call log.

Upkeep cost per month

Tooling, plus every engineering day above at the day rate, with the yearly line divided by twelve.

7

Quality

How often does each option produce an outcome you would accept?

one value per option

This is what turns cost per case into cost per acceptable outcome. One written definition of "acceptable", applied to every option by the same reviewer.

Rules Plain code. No model. Rules v2 Same code, rebuilt on the written spec. Assisted Model drafts. A person approves each one. Agent Model plans, calls tools, acts.
Acceptable-outcome rate %

The share of cases that meet the written definition with no rework. Measure it on the same set of cases for all four options.

Acceptable outcomes in the period

Total cases, times the acceptable-outcome rate.

04 Your result

The outcome, read against the rubric.

Not started

Fill every one of the 83 lines to read the outcome. Until then this shows each total as soon as the lines that feed it are filled.

    Line RulesPlain code. No model.Rules v2Same code, rebuilt on the written spec.AssistedModel drafts. A person approves each one.AgentModel plans, calls tools, acts.
    Run cost per monthModel, tools, review, escalation, declined cases and upkeep, added up.
    Run cost for the periodRun cost per month, times months.
    Total cost for the period, build plus runThe build, paid once, plus the run cost for the period.
    Cost per caseTotal cost divided by total cases. The wrong number to argue about on its own.
    Cost per acceptable outcomeTotal cost divided by acceptable outcomes. The number to argue about.
    Net position against the manual baselineManual baseline minus total cost. Positive means the option saves money over the period.
    Months to break even on the buildBuild cost divided by the monthly saving against doing it by hand. "Never" means the option costs more to run each month than the manual process.
    Reference example

    An ordering agent across 40 sites, every line already filled.

    A retailer in India raises about 2,000 purchase orders a month across 40 sites. Each one takes a buyer 25 minutes by hand.

    The team built a full agent that reads the request, checks stock and supplier terms, and places the order. It worked. It also cost more to run each month than the buyers did.

    Building it forced the team to write down every decision rule the buyers had been carrying in their heads. A second rules workflow, rebuilt on that written spec, took 35 engineering days and wins on cost per acceptable outcome.

    Cost per case picks the original rules workflow. Cost per acceptable outcome picks the rebuilt one. That gap is the whole reason the second number exists.

    Every figure is in Indian rupees at Indian rates. If you work elsewhere, change the four rates in sections 1 and 2 first. The shape of the comparison holds; the absolute numbers will not.

    Back to your model
    The written spec is the asset. Run the code. Leading option: Rules v2, on the written spec · INR 52.78 per acceptable outcome

    The agent build paid for itself by forcing the team to write the decision rules down. Ship the rebuilt rules workflow, and keep the specification under version control like any other code. Revisit the model when volume or the price of a call changes.

    • Cost per case picks Rules workflow. Cost per acceptable outcome picks Rules v2, on the written spec. Argue about the second one: cost per case rewards a system that quietly declines the hard cases.
    • Full agent with tools breaks even in month 17.7, after the 12-month period you modelled. It pays back, but only if you can defend the longer horizon.
    Line RulesPlain code. No model.Rules v2Same code, rebuilt on the written spec.AssistedModel drafts. A person approves each one.AgentModel plans, calls tools, acts.
    Run cost per month INR 50,467 INR 37,733 INR 1,42,418 INR 3,67,967
    Run cost for the period INR 6,05,600 INR 4,52,800 INR 17,09,016 INR 44,15,600
    Total cost for the period, build plus run INR 11,05,600 INR 11,52,800 INR 27,29,016 INR 67,55,600
    Cost per case INR 46.07 INR 48.03 INR 113.71 INR 281.48
    Cost per acceptable outcome INR 59.06 INR 52.78 INR 127.76 INR 299.45
    Net position against the manual baseline INR 48,94,400 INR 48,47,200 INR 32,70,984 INR -7,55,600
    Months to break even on the build 1.1 months 1.5 months 2.9 months 17.7 months

    Check these assumptions first

    1. 01
      Rules v2, on the written spec: escalations are 35% of run cost. That line decides the comparison. Measure it in a trial before you trust it.
    2. 02
      Full agent with tools: upkeep is 50% of run cost. That line decides the comparison. Measure it in a trial before you trust it.
    3. 03
      Rules workflow: declined cases are 38% of run cost. That line decides the comparison. Measure it in a trial before you trust it.
    4. 04
      Model-assisted draft: upkeep is 41% of run cost. That line decides the comparison. Measure it in a trial before you trust it.
    1

    Workload and manual baseline

    What does this work cost today, done by hand?

    Line Unit Value
    Currency label free text INR
    Cases per month count 2,000
    Months in the period months 12
    Minutes per case, by hand minutes 25
    Cost per hour of the person doing it by hand INR per hour 600
    Total cases in the period worked out 24,000
    Manual baseline cost for the period worked out INR 60,00,000
    2

    Team rates

    What does an hour of each person cost?

    Line Unit Value
    Engineering day rate INR per day 20,000
    Engineer cost per hour INR per hour 2,500
    Reviewer cost per hour INR per hour 600
    3

    Build cost, paid once

    What does each option cost before it handles its first case?

    Line Unit RulesPlain code. No model.Rules v2Same code, rebuilt on the written spec.AssistedModel drafts. A person approves each one.AgentModel plans, calls tools, acts.
    Engineering days to build days 25354595
    Engineering days for evals and a test set days 00622
    Total build cost worked out INR 5,00,000INR 7,00,000INR 10,20,000INR 23,40,000
    4

    Inference and tool calls, per month

    What does the model bill actually include?

    Line Unit RulesPlain code. No model.Rules v2Same code, rebuilt on the written spec.AssistedModel drafts. A person approves each one.AgentModel plans, calls tools, acts.
    Model calls per case, successful calls per case 001.26
    Retry and failure multiplieroften left out multiplier, 1.0 = no retries 111.151.35
    Cost per model call INR 001.82.5
    Tool and API calls per case calls per case 0.20.215
    Cost per tool call INR 0.50.50.50.5
    Model cost per month worked out INR 0INR 0INR 4,968INR 40,500
    Tool cost per month worked out INR 200INR 200INR 1,000INR 5,000
    5

    Human in the loop, per month

    How much human time does each option still need?

    Line Unit RulesPlain code. No model.Rules v2Same code, rebuilt on the written spec.AssistedModel drafts. A person approves each one.AgentModel plans, calls tools, acts.
    Share of cases a person reviews % 12810020
    Review minutes per reviewed caseoften left out minutes 6624
    Share of cases escalated to an engineer % 10.81.54
    Engineer minutes per escalationoften left out minutes 20202535
    Share of cases declined or abandonedoften left out % 8432
    Manual minutes per declined case minutes 12121212
    Review cost per month worked out INR 14,400INR 9,600INR 40,000INR 16,000
    Escalation cost per month worked out INR 16,667INR 13,333INR 31,250INR 1,16,667
    Declined-case cost per month worked out INR 19,200INR 9,600INR 7,200INR 4,800
    6

    Upkeep and operations, per month

    What does it cost to keep the system trustworthy after launch?

    Line Unit RulesPlain code. No model.Rules v2Same code, rebuilt on the written spec.AssistedModel drafts. A person approves each one.AgentModel plans, calls tools, acts.
    Observability and eval toolingoften left out INR per month 008,00035,000
    Eval maintenanceoften left out engineer days per month 0013
    Model-version re-qualificationoften left out engineer days per year 00924
    Prompt and regression testingoften left out engineer days per month 00.250.51.5
    Incident and on-call loadoften left out engineer days per month 000.251
    Upkeep cost per month worked out INR 0INR 5,000INR 58,000INR 1,85,000
    7

    Quality

    How often does each option produce an outcome you would accept?

    Line Unit RulesPlain code. No model.Rules v2Same code, rebuilt on the written spec.AssistedModel drafts. A person approves each one.AgentModel plans, calls tools, acts.
    Acceptable-outcome rate % 78918994
    Acceptable outcomes in the period worked out 18,72021,84021,36022,560
    05 How the outcome is read

    Agree this before the numbers go in, not after.

    The leading option is the one with the lowest cost per acceptable outcome among the options that save money against the manual baseline over the period. The row that matches your result is highlighted.

    The full agent has the lowest cost per acceptable outcome and beats the manual baseline
    The full agent earns its keep
    Before you commit, measure the three lines that drive this number in a trial: escalation share, review minutes and the retry multiplier. If the trial moves any of them, run the model again before the budget conversation.
    The model-assisted draft has the lowest cost per acceptable outcome and beats the manual baseline
    Draft with a model, approve with a person
    Every case still passes a reviewer, so review minutes per case is the line that decides this. Time it on real cases, not on a demo. If review time doubles, check whether the full agent or the rules workflow now leads.
    The rebuilt rules workflow has the lowest cost per acceptable outcome and beats the manual baseline
    The written spec is the asset. Run the code.
    The agent build paid for itself by forcing the team to write the decision rules down. Ship the rebuilt rules workflow, and keep the specification under version control like any other code. Revisit the model when volume or the price of a call changes.
    The original rules workflow has the lowest cost per acceptable outcome and beats the manual baseline
    The rules workflow wins as it stands
    No model earns its cost on this workload at these rates. Keep the rules workflow. Check the acceptable-outcome rate first: if the rules workflow declines a lot of cases, its quality line may be the one that is wrong.
    No option saves money against the manual baseline over the period
    Nothing here beats doing it by hand
    Either the volume is too low to earn back a build, or a rate needs checking. Do not build for cost. Build only if there is another reason, and write that reason down in the decision record.

    Three flags sit beside the outcome and do not change it. Disagreement: cost per case and cost per acceptable outcome pick different options. Never breaks even: an option costs more to run each month than doing it by hand. Beyond the period: an option pays back, but after the months you modelled.

    06 The nine lines teams leave out

    The model bill is usually the smallest line.

    Each of these is marked often left out in the model above. Together they are most of what it costs to run an agent, and none of them appears in a demo.

    01

    Retry and failure multiplier

    Total model calls divided by successful ones. Count timeouts, tool errors and loops that re-enter, not only refusals. 1.0 means the system never retries, which no real system does.

    02

    Review minutes per reviewed case

    Time it on real cases. Approving a good draft is quick; checking an agent that acted on its own is not.

    03

    Engineer minutes per escalation

    Reading the trace, working out what happened, fixing the case. Longer for an agent, because there is more trace to read.

    04

    Share of cases declined or abandoned

    Cases the system refuses or gives up on. They do not disappear: a person finishes them, at the manual rate.

    05

    Observability and eval tooling

    Tracing, logging, the eval runner, the dashboards. Licence fees or the compute to run your own.

    06

    Eval maintenance

    Engineering days a month spent adding cases to the eval set and fixing the ones that go stale.

    07

    Model-version re-qualification

    Engineering days a YEAR spent re-running the eval set after a model or SDK update. Spread across the months by the model. Count every update the provider ships.

    08

    Prompt and regression testing

    Engineering days a month spent checking that a prompt change did not break something that used to work.

    09

    Incident and on-call load

    Engineering days a month on incidents caused by the system. Take it from the on-call log.

    07 What this model does not price

    A number you can take to a budget conversation. Not the whole argument.

    • It assumes steady volume, one workload and stable rates for the whole period.
    • It does not price revenue upside. A system that wins new business is not captured by a cost line.
    • It does not price the risk of a wrong action that cannot be undone. That belongs in the Agent Authority Review, not in a cell here.
    • It does not price the option value of the capability you build. If that is the argument, write it in the decision record in words, not as a number in this sheet.

    An agent proposal is also a commitment to operate an agent. This tool turns that sentence into a number. Sometimes the number says build it. The reference example is a case where it did not, and where the build was still worth doing, because it produced the written specification that made the cheap option good.

    See the cohort  All resources  The same model as an Excel workbook 

    Published 14 September 2026 · free to use and to pass on

    PDF Your model

    Get the PDF

    The PDF carries every line you typed, every total, the outcome and the assumptions to check. It is built against a name and an email address.

    Your figures go into the file and are not stored. Your address is used to send you this tool once. It does not start any other email.