Price the coverage gap
After-hours ROI calculator
Compare five ways to handle calls and messages after hours.
Five ways to cover the gap
What remains after the cost?
Compare each option with keeping your current coverage. Recovery, contribution, and payback are estimates; service quality still needs a real test.
Voice agent has the highest modeled 12-month improvement: $23,308. This is an assumption-based comparison, not a vendor recommendation or proof of service quality.
Get the one-page business case.
Signup unlocks a PDF with all five options, assumptions, sensitivity, and a safe pilot plan. Download it in this tab. Your numbers and personalized PDF are never uploaded or emailed.
- Five-option business case
- Your assumptions and sensitivity
- A practical pilot plan
Preview the result or example included in this kit
Voice agent has the highest modeled 12-month improvement: $23,308. This is an assumption-based comparison, not a vendor recommendation or proof of service quality.
52.0 lost inquiries/month imply up to $7,800 in job revenue at your close rate if all were reached instantly. That is an opportunity ceiling, not a measured loss or profit.
| Monthly unless noted | |||||
|---|---|---|---|---|---|
| Recovered inquiries | 44.2 | 42.9 | 27.8 | 14.3 | 41.1 |
| Recovered jobs | 13.3 | 12.9 | 8.3 | 4.3 | 12.3 |
| Job revenue | $6,626 | $6,441 | $4,169 | $2,146 | $6,168 |
| Job contribution | $2,650 | $2,577 | $1,668 | $858 | $2,467 |
| Total coverage cost | $1,800 | $950 | $420 | $480 | $750 |
| Net improvement | $1,150 | $1,927 | $1,548 | $678 | $2,017 |
| 12-month improvement | $13,705 | $22,868 | $18,423 | $7,740 | $23,308 |
| Setup payback | <0.1 months | 0.1 months | <0.1 months | 0.6 months | 0.4 months |
| 12-month ROI | 76% | 284% | 1159% | 302% | 370% |
Select an option to edit its assumptions and sensitivity. Scroll horizontally on small screens. Negative improvement means the option loses against current coverage. Options are alternatives, not additive.
Review all five assumption sets
| Option | Reach before delay (%) | Effective response delay (minutes) | Additional monthly cost ($) | One-time setup cost ($) | Existing staffing actually replaced (%) |
|---|---|---|---|---|---|
| In-house on-call | 90 | 10 | 1500 | 100 | 0 |
| Outsourced human service | 85 | 5 | 650 | 250 | 0 |
| Missed-call text-back | 60 | 20 | 120 | 150 | 0 |
| Chat agent | 30 | 15 | 180 | 400 | 0 |
| Voice agent | 80 | 2 | 450 | 900 | 0 |
What changes the answer for Missed-call text-back?
Each test changes one assumption by −20% and +20% relative to its current value, capped at field limits. Zero stays zero. The largest dollar swing signals what to verify first, not what is most uncertain. Equal swings are ties.
| Assumption | 12-month swing ($) |
|---|---|
| Inquiries | 8005 |
| Close rate | 8005 |
| Job value | 8005 |
| Assumption | −20%: annual improvement | +20%: annual improvement |
|---|---|---|
| Inquiries | $14,421 | $22,426 |
| Close rate | $14,421 | $22,426 |
| Job value | $14,421 | $22,426 |
| Margin | $14,421 | $22,426 |
| Reach | $14,421 | $22,426 |
| Time sensitivity | $17,853 | $18,812 |
| Monthly cost | $18,711 | $18,135 |
Volume tests hold quoted costs fixed. Real usage pricing, capacity limits, and seasonal demand can change the outcome.
Test a small intake pilot
Draft-only after-hours intake pilot: use approved business facts, allowed services, hours, consent rules, and a named human backup. Test with fictional callers first.
Collect the minimum callback details in an approved system. Draft an acknowledgement and a callback task; do not promise a price, arrival time, eligibility, or completed booking without an approved source and human approval.
Stop automation on opt-out, a request for a person, urgent or unsafe content, missing facts, a duplicate inquiry, a failed integration, or the approved cost limit. Route to the named human; never silently retry an external action.
Keep emergency, technical, medical, legal, and insurance judgments with qualified people. A financial score cannot authorize autonomous handling.
For 14 days, compare consented, deduplicated inquiries, actual response delay, paid jobs, contribution, total charges, failed handoffs, and complaints with the baseline. A person approves each send, booking, and pilot expansion. Do not paste private customer records into an unapproved AI tool.
Next, use the AI Receptionist Readiness assessment to check call flows, human fallback, and test-call criteria before a live pilot. Availability follows its separate publication status.
How the comparison works
Monthly inquiries = weekly inquiries × 52 ÷ 12. Recovered jobs = monthly inquiries × reach × 2−delay ÷ (60 × hours until half move on) × close rate. Percentages are converted to fractions. Timing decay is an editable Nerd Out assumption, not an empirical prediction.
Recovered contribution = jobs × job revenue × contribution margin. Monthly improvement = recovered contribution + staffing spend actually eliminated − additional coverage cost. Total coverage cost = current staffing − eliminated staffing + added cost. The model assumes existing service is preserved and these lost inquiries would otherwise produce no jobs.
12-month improvement = monthly improvement × 12 − setup. 12-month ROI divides that improvement by setup plus 12 months of added cost; no new spend means ROI is undefined. Setup payback divides setup by positive monthly improvement. A loss-making option never has payback, even with no setup fee.
Fractional jobs are expected averages. Costs stay fixed at your quoted volume. No taxes, financing, discounting, seasonality, capacity limits, cash collection delay, or lost value from replacing existing staff is modeled. This is not net profit or financial advice.
What does a 24/7 answering service cost?
Compare a written total for your volume: recurring base, included calls or minutes, overages, setup, integrations, transfers, staff oversight, and fallback. This calculator does not supply current quotes.
For examples of billing structures, review Ruby’s published plans and Smith.ai’s AI receptionist pricing. Neither is endorsed here; neither source supplies this tool’s fictional numbers. The contribution method follows the SBA’s break-even explanation. Sources and model reviewed 2026-08-30.
Which option fits the work?
- In-house on-call
- A person on your team handles intake. Include paid availability, actual call time, backup coverage, and supervision. Availability does not prove every inquiry is reached.
- Outsourced human service
- An external team follows your intake instructions. Confirm included calls or minutes, overages, transfers, languages, holidays, and escalation coverage in writing.
- Missed-call text-back
- A permitted text invites the caller to continue asynchronously. Reach excludes landlines, opt-outs, failed delivery, and people who will not text. Keep an approved human fallback.
- Chat agent
- Chat can only recover callers who actually reach and use the chat channel. Set reach to zero if you have no tested path from a missed call to chat. Do not count new website leads again.
- Voice agent
- An automated phone workflow handles bounded intake. Include telephony, model usage, transfer, monitoring, and human fallback costs. Test requests for a person and failed integrations.
For routine service callbacks, test whether the caller will use the channel. For urgent or regulated work, have qualified people approve triage and escalation before comparing price. No option’s recovery rate, response delay, or safety can be established by its label.
How this tool works
Put in-house coverage, a human service, text-back, chat, and voice AI on the same footing before you commit.
- Enter the inquiries you lose and your job economics
- Compare five options and test the assumptions
- Get the one-page business case
Before you buy coverage
A fast answer is useful only if it leads to the right next step.
An after-hours phone answering service receives inquiries outside your staffed hours and follows an agreed intake or escalation process. Compare what happens after the first response: whether someone can collect the right details, create a reviewable callback task, and reliably reach a human.
The best answering service for a small business depends on the work, caller preferences, approved procedures, and measured quality. Start with a written quote and a sample-call test. A higher financial estimate cannot compensate for failed handoffs or an unsafe workflow.
Questions owners ask
After-hours answering service questions
What does this calculator compare?
In-house on-call, an outsourced human service, missed-call text-back, a chat agent, and a voice agent. All five use the same lost-inquiry pool and conditional close rate, with editable reach, delay, new costs, setup, and actual staffing savings. They are alternatives; do not add their recoveries together.
Are the example costs current answering-service prices?
No. Every prefilled number is fictional. Replace them with a written quote and observed operating data. Include base fees, usage, transfers, integrations, supervision, and fallback. No recovery rate or timing assumption is an industry benchmark.
Does recovered revenue equal ROI?
No. Job contribution subtracts variable delivery costs through your margin. The model then adds only staffing spend actually eliminated and subtracts new recurring and setup costs. The 12-month ROI denominator is new recurring spend plus setup; zero new spend makes that ratio undefined. Taxes, financing, capacity, and cash timing are excluded.
Can a chat agent recover missed phone calls?
Only if callers actually move to chat through a tested, permitted path. Reach must reflect that channel change. Use zero reach if no path exists; do not substitute unrelated website leads or double-count jobs already won by callbacks.
How does response-time sensitivity work?
You set how long it takes for half of reachable callers to move on. The model halves remaining interest again for every interval. That exponential decay is a planning assumption, not a research-derived response curve. Replace it with a conservative estimate and verify it in a pilot.
Will you email or save my business case?
No. Signup unlocks a one-page PDF download in this tab and subscribes you to Nerd Out notes. Your inputs and personalized file never reach our server. Download before leaving; reload starts the fictional example. The free share image includes no private values unless you explicitly opt in.