What each number is
Volumes shifted, displaced and bounced back are arithmetic on your inputs. Resolution, displacement, eligibility and the residual complexity curve are assumptions; the opening profile is labelled heuristic and a driver still at it grades Directional. Net realizable is a conditional forecast: freed minutes become cash only through the capacity action you select, while bot fees are cash and netted in full.
Formulas
Monthly contacts × voice share × eligible share
The shift is capped at this pool; a larger request is scaled down and disclosed.
Shifted = contacts × shift points; resolved = shifted × resolution; displaced = resolved × displacement; bounced back = shifted − resolved
Resolved contacts that do not replace a voice call are new demand and earn no saving.
Voice effective minutes × (1 + curve × shifted ÷ voice contacts)
The calls left behind are harder. The curve is mild, moderate or severe.
(Voice contacts × voice minutes − remaining calls × residual minutes) ÷ displaced
Total voice minutes are conserved, so the curve fixes how simple the departing calls must be. The tool shows it so the curve can be checked.
Displaced × departing minutes − target channel minutes − bounced × departing minutes × (return factor − 1)
A bounced call always existed; only its extra friction is new cost. Chat and email minutes are handle time ÷ concurrency.
Net minutes × hourly wage × marginal load ÷ 60 × the capacity action's share − bot fees
Bot fees are paid on every bot attempt, resolved or bounced.
Net minutes ÷ (22 days × 8 hours × 60 × 70% productive)
A capacity equivalent of 7,392 productive minutes per FTE a month.
Chat FTE added × (training per agent + ramp weeks × 5 days × 8 hours × loaded wage × 30% ramp loss)
Payback is transition ÷ monthly net realizable.
The resolution rate of the largest shift at which net realizable crosses zero
Solved on whole points and interpolated between them.
Rules and bands
The read gives the break-even resolution and the gap to it.
Complaint, regulated, save-risk, vulnerable, identity or high-emotion volume needs a pilot before rollout.
If break-even is under 1%, the read says to check the bot fee and return factor first.
The curve is set too high for this shift; completeness holds Directional.
The verdict is a property of the answer and never caps a confidence axis. A bot under $0.10 a contact carrying volume is flagged as near-free and holds completeness Directional.
Every constant and where it comes from
Default so the tool opens on a runnable case. A volume still at this value grades evidence Directional.
Default so the tool opens on a runnable case. Voice share of the current mix.
Default so the tool opens on a runnable case. A voice handle time still at this value grades evidence Directional.
Default so the tool opens on a runnable case. Voice is one contact at a time.
Default so the tool opens on a runnable case. Chat share of the current mix.
Default so the tool opens on a runnable case. Chat handle time.
Default so the tool opens on a runnable case. Typical chat session concurrency.
Default so the tool opens on a runnable case. Email share of the current mix.
Default so the tool opens on a runnable case. Email handle time.
Default so the tool opens on a runnable case. Email worked one at a time.
Default so the tool opens on a runnable case. Bot share of the current mix.
Default so the tool opens on a runnable case. A bot fee still at this value grades cost evidence Directional when the bot carries volume.
Default so the tool opens on a runnable case. An eligibility still at this value grades evidence Directional.
Default so the tool opens on a runnable case. Planned shift into chat.
Default so the tool opens on a runnable case. Planned shift into the bot.
Default so the tool opens on a runnable case. Planned shift into email. Ships at zero.
Default so the tool opens on a runnable case. A chat resolution still at this value grades evidence Directional when chat carries a shift.
Default so the tool opens on a runnable case. A bot resolution still at this value grades evidence Directional when the bot carries a shift.
Default so the tool opens on a runnable case. An email resolution still at this value grades evidence Directional when email carries a shift.
Default so the tool opens on a runnable case. A chat displacement still at this value grades evidence Directional when chat carries a shift.
Default so the tool opens on a runnable case. A bot displacement still at this value grades evidence Directional when the bot carries a shift.
Default so the tool opens on a runnable case. An email displacement still at this value grades evidence Directional when email carries a shift.
Default so the tool opens on a runnable case. Friction of a recovery call after a failed shift.
Default so the tool opens on a runnable case. Transition training cost.
Default so the tool opens on a runnable case. Transition ramp length.
Mild complexity curve. Easy volume leaves and residual voice gets slightly harder.
Moderate complexity curve. The typical support environment and the shipped choice.
Severe complexity curve. Remaining voice work becomes materially harder.
Planning convention for productive time per FTE.
Planning convention for productive time per FTE and for ramp cost.
Share of paid time an agent spends on contacts, used to convert freed minutes to FTE.
Planning convention for ramp cost.
Productivity lost while a new chat agent ramps, priced into transition cost.
Displaced contacts implied to average under this handle time mean the complexity curve is too severe for the volume moved. Disclosed, and holds completeness Directional.
A bot fee at or below this, with bot volume shifted, makes any shift look costless and drives break-even toward zero. Disclosed, and holds completeness Directional.
A break-even below this reads as a shift profitable at any resolution. Framing only. It is a property of the answer and reaches no confidence axis by doctrine 5.5.
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.
The variable cost that disappears when one contact goes away. The only load a saving may be valued on, because fixed technology and facilities do not fall with volume.
Worked example
Computed by the tool's own engine at its opening profile, with avoided hiring as the capacity action (75% of freed capacity). The tool itself opens with no action chosen, which realizes $0.
Volume: 100,000 contacts a month, voice 70% at 7 minutes, 60% of voice eligible
Shift: 10 points to chat (85% resolve, 80% displace, 10 minutes at 2.5 concurrent) and 10 points to bot (65% resolve, 70% displace, $0.50 a contact)
Cost: $21.53 an hour, marginal load 1.18, loaded 1.3; return factor 1.2; moderate curve; $1,500 training and 4 ramp weeks
70,000 voice × 60% = 42,000; 20,000 shifted, within the pool
chat 6,800 and bot 4,550 displace 11,350 voice calls; 5,000 bounce back
voice rises 4.3% to 7.30 minutes, so the departing calls average 5.45 minutes
23,205 a month (3.14 FTE)
$7,369 labor at 75% less $4,025 bot fees = $3,344 a month
$12,771, paid back in 3.8 months
Approve; break-even at 71% chat resolution against 85% modelled
What this tool cannot tell you
- Resolution and displacement decide the answer, and both are usually estimated before a pilot. The break-even line shows how much room the plan has.
- The residual curve is a planning assumption. The departing handle time it implies is shown so it can be checked against the calls you plan to move.
- This is an operating-capacity model. It does not value what an interaction is worth to the business, and the full investment case belongs in the Business Case Builder.
- Chat and email minutes are handle time divided by concurrency; a queue that does not reach that concurrency costs more.