What each number is
Every figure is arithmetic on your inputs. The opening case, the attribution stances and the target planning ranges are planning values, labelled as heuristics; the wage is the BLS median and the benefits load the shared load. The return, payback and three-year value are a conditional forecast: freed capacity becomes cash only through the capacity action you select, and no action selected realizes $0.
Formulas
Loaded = hourly × (1 + benefits). Savings rate = hourly × the shared marginal load (1.18, never above the loaded rate). Marginal per contact = AHT ÷ 3,600 × savings rate, unless one is entered
A pulled marginal more than 10% from the derived one is flagged as from a different operation.
Contacts × 12 × containment × marginal
Deflected contacts leave the handled pool before any per-contact saving.
Handled contacts × (talk and hold × reduction + after-call work × ACW reduction) ÷ 3,600 × savings rate
After-call work is a slice of AHT, so the two reductions never count the same seconds.
Repeat contacts × (1 − new failure rate ÷ old failure rate) × marginal
Repeats are your measured same-reason share, or (1 − FCR) ÷ (2 − FCR) on the handled pool.
Agents × attrition × reduction × (recruiting + training days × 8 hours × savings rate)
Recruiting is cash; training time is capacity, valued at the savings rate like every other freed hour.
Each lever × its stance share. Expected: 85%, 90%, 80%, 65%; conservative: 70%, 80%, 65%, 50%; aggressive: 100%, 100%, 100%, 100%
Order: containment, handle time, FCR, attrition.
Attributed capacity × the capacity action's share + attributed attrition cash
Cash out the door is never scaled. No action realizes $0 of capacity.
Month t earns the net × a ramp factor (0 through migration, then linear over the ramp), plus any displaced spend, less the platform fee; one-time cost at month 0
Payback is the first month cumulative cash turns positive.
(Three-year benefit − three-year cost) ÷ three-year cost, where cost = implementation + exit + backfill + platform fee × agents × 36
Displaced spend counts as a benefit, never netted from the cost.
Rules and bands
Above any, the benefit stream caps at Planning-grade. Internal planning ranges, labelled.
The benefit stream caps at Planning-grade.
The benefit stream grades Directional, Planning-grade or Finance-grade. A baseline pulled from another tool grades by the origin grade it carries, never above Planning-grade. An unanswered question with a baseline edited reads as your estimate.
A finding in the read. It never caps a confidence axis.
The return, payback and whether the case pays back never cap a confidence axis. Evidence is the weaker of the cost stream (the investment's evidence) and the benefit stream (stance, target ambition and where the baselines come from).
Every constant and where it comes from
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Default so the tool opens on a runnable case.
Aggressive stance: full modelled saving, no attribution haircut. Caps the benefit stream at Planning-grade. Containment.
Aggressive stance: full modelled saving, no attribution haircut. Caps the benefit stream at Planning-grade. Handle time.
Aggressive stance: full modelled saving, no attribution haircut. Caps the benefit stream at Planning-grade. First contact resolution.
Aggressive stance: full modelled saving, no attribution haircut. Caps the benefit stream at Planning-grade. Attrition.
Expected stance: each lever discounted for real-world attribution. Containment.
Expected stance: each lever discounted for real-world attribution. Handle time.
Expected stance: each lever discounted for real-world attribution. First contact resolution.
Expected stance: each lever discounted for real-world attribution. Attrition.
Conservative stance: a heavy haircut on the soft levers. Containment.
Conservative stance: a heavy haircut on the soft levers. Handle time.
Conservative stance: a heavy haircut on the soft levers. First contact resolution.
Conservative stance: a heavy haircut on the soft levers. Attrition.
Top of the 10 to 25 percent internal planning range for containment without a proven pilot. Above it the benefit stream caps at Planning-grade.
Top of the 8 to 15 percent internal planning range for handle-time reduction. Above it the benefit stream caps at Planning-grade.
Top of the 5 to 10 point internal planning range for FCR improvement. Above it the benefit stream caps at Planning-grade.
Attrition reduction above this is hard to attribute to a platform. The benefit stream caps at Planning-grade.
Implementation per agent used only for the comparison case the read prints beside yours. Feeds no figure of your case.
A paying case whose three-year net is under this share of its benefit is read as fragile. A property of the answer; it reaches no confidence axis.
A pulled marginal cost more than this far from the one derived from handle time and wage is flagged as from a different operation.
Converts training days to paid hours.
The one agent wage benchmark the platform cites. Staffing, Cost per Contact and Channel Shift read it. It is the occupation median, no figure of the user's own, so a driver still at it grades evidence Directional. Source: US Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 (released 15 May 2026), SOC 43-4051 Customer Service Representatives, national median hourly wage.
Wage plus benefits and employer payroll burden, and nothing else. The narrowest of the three loads. Use it wherever a wage becomes a loaded hourly rate for unit metrics.
Worked example
Computed by the tool's own engine at its opening case: the expected stance, phasing on (9 months migration, 6 months ramp) and no capacity action, which is how the tool opens. The last step turns on hiring avoidance (75%) to show what realization does.
Operation: 200 agents at $21.53 an hour with 30% benefits, 120,000 contacts a month, AHT 420 seconds with 45 after-call, FCR 72%, attrition 35%
Targets: containment 15%, handle time 12%, ACW 30%, FCR +8 points, attrition down 20%
Investment: $750,000 implementation, $135 per agent a month
$27.99 an hour loaded; $2.96 marginal per contact at the savings rate
containment $640,216, handle time $505,313, FCR $226,743 (76,500 repeats), attrition $108,754 (14 hires); $1,481,026 in all
$1,219,200 capacity and $31,850 cash at the expected stance
net $31,850 a year, the attrition cash only; three-year cost $1,722,000; return -96%
$914,400 capacity realized, net $946,250 a year; three-year benefit $1,931,927, return 12%, payback month 32
Checked against
Cases whose answer is known outside this site. The test suite computes each one with this tool's own engine on every change.
The tool's opening case at the expected stance, phasing on, no capacity action. Result: net $31,850 a year, three-year cost $1,722,000. ContactCenterCX tracker fixture, reconciled to the dollar on a live PDF. Internal, and named as such.
What this tool cannot tell you
- The baseline answer is one question for four figures. A single baseline from a weaker source should be answered at that source's level; the tool cannot tell which figure came from where unless it was pulled from another tool.
- Freed agent time is not cash. The case shows it three ways: gross, attributed and realized, so the gap is visible.
- The ramp is linear after migration. A program that lands in steps pays back later than it shows.
- Each lever's target is yours. The planning ranges flag ambition; they do not say a target is wrong.