Most contractors think of an answering service as a phone-coverage cost. The real value is on the revenue side: each missed call that converts into a captured lead is a job that wasn't going to happen otherwise. This post walks through the ROI math on automating missed-call follow-up, with worked numbers for a 1-truck and a 10-truck shop.
OnCrew at a glance
OnCrew is an AI receptionist for home service contractors such as HVAC, plumbing, electrical and roofing shops. It answers forwarded calls in the company's name, takes the problem, the service address and a callback number, flags emergencies, texts the on-call tech and emails the transcript and recording.
| Plan | Price | Included minutes | Extra minutes | Alert texts |
|---|---|---|---|---|
| Starter | $49 a month | 150 | $0.49 each | 150 a month, then $0.05 each |
| Pro | $149 a month | 500 | $0.39 each | 500 a month, then $0.05 each |
| Multi-Truck | $349 a month | 1,500 | $0.29 each | 1,500 a month, then $0.05 each |
14-day free trial. No credit card required to start. No setup fee. Same price nights, weekends, and holidays. Cancel from your dashboard anytime.
Call (213) 319-5075: Emily asks what kind of business you run and answers as its front desk. You can ask her anything about OnCrew, too.
The starting point: voicemail callback rates
These are the model's assumptions, not measured industry statistics. Replace them with your own numbers:
- Voicemails left on contractor lines: 30-60% of missed calls (the rest hang up).
- Voicemails that get a callback from the contractor: 40-70% (the rest disappear into queue).
- Voicemails returned within 1 hour: 15-30% (the timing that converts).
- Voicemails returned same day: 50-75%.
The compound result: in this model a missed call has roughly a 20-30% chance of becoming a captured lead through voicemail. The other 70-80% drop out.
What we can measure on our own lines: on the lines OnCrew answers, 30.1% of conversations came outside weekday business hours, and those after-hours conversations were emergencies 28.3% of the time, against 5.8% in weekday daytime (1,277 conversations on the lines of nine businesses that have paid for OnCrew over the 120 days from June 19 to October 16, 2026; shares of the calls each shop sends OnCrew, not of every call it gets; method on the missed call statistics page).
What captured intake changes
In this model a real intake (live or AI) captures 90% of callers who stay on the line. The customer talks to someone, gives their information, and gets a commitment for callback or a scheduled appointment. In the model, the drop-out rate goes from 70-80% to about 10%.
That's a 3-4x lift in captured-lead rate per inbound call.
The conversion-to-job math
Assumed close rates by trade for this model (rules of thumb, not measured data; use your own):
- Plumbing emergency: 70-80% conversion (urgent need, willing customer)
- HVAC emergency: 70-80% (same)
- Routine plumbing: 40-50% (price-shopped)
- Routine HVAC service: 50-60%
- New install / replacement estimate: 25-35%
- Roofing repair: 35-45%
- Roofing replacement estimate: 15-25%
- Electrical service: 50-60%
- Electrical estimate: 30-40%
Compound this: captured-lead-rate × conversion-rate = job rate per inbound call.
Worked math: 1-truck plumbing shop
Inbound calls: 80/month. Missed during work hours + after-hours: 30/month (37%).
Voicemail baseline:
- 30 missed × 50% leave voicemail × 50% return within day × 50% conversion = 3.75 jobs/month from missed calls.
- Average plumbing job: $350.
- Monthly revenue from missed-call recovery: $1,312.
AI receptionist:
- 30 missed × 90% captured × 60% conversion = 16.2 jobs/month.
- Monthly revenue from captured intake: $5,670.
- Monthly answering service cost: $49 (Starter, 150 included minutes; the 30 forwarded calls at OnCrew's measured average of about 1.2 minutes a call come to about 36).
- Net monthly lift: $5,670 - $1,312 - $49 = $4,309 in incremental revenue.
In this model that is about $4,300 a month of added revenue against a $49 plan. The model is only as good as its assumptions, so run it with your own close rate and ticket.
Worked math: 10-truck HVAC shop
Inbound calls: 400/month. Missed: 100/month (25%).
Voicemail baseline:
- 100 × 50% × 50% × 55% = 13.75 jobs/month recovered.
- Average HVAC job: $475.
- Monthly revenue from voicemail callback: $6,531.
Live receptionist (current):
- 100 captured × 90% × 60% = 54 jobs/month.
- Monthly cost: $915 (Smith.ai's published receptionist plan: $810 for 90 calls, plus 10 calls at $10.50, read on smith.ai 2026-09-22).
- Net monthly lift vs voicemail: (54 - 13.75) × $475 - $915 = $18,204.
AI receptionist:
- 100 captured × 90% × 60% = 54 jobs/month (same conversion).
- Monthly cost: $149 (Pro, 500 included minutes, enough for the same 100 captured calls at OnCrew's measured average of about 1.2 minutes a call, about 120 minutes).
- Net monthly lift vs voicemail: 40.25 × $475 - $149 = $18,970.
ROI comparison: in this model AI delivers the same job count at $766/month lower cost than live. The model assumes the same conversion rate for both; the cost difference is the gap.
Where the math gets nuanced
The "same conversion rate" assumption above is the contested part. Some scenarios where conversion differs:
- High-touch residential brand. Live phone empathy may lift conversion on relationship-driven calls. If your average ticket is $5,000+, that lift matters.
- Premium service contracts. Live receptionist's polish can affect retention.
- Emergency calls. AI answers at 2 AM the same way it does at 2 PM, which is where night-shift staffing is weakest.
- Estimate-heavy paint or remodel. Both AI and live capture intake well; conversion depends on the estimator, not the intake.
The honest answer: we have not measured AI against live conversion, and we have not seen a study that did it well for the trades. Treat any difference as something to test in your own pilot; it matters most in brand-experience contexts.
The compounding effect of speed
The verified research points the same way. In the 2007 Lead Response Management Study, the odds of qualifying a lead dropped 21 times when the response slipped from 5 minutes to 30 minutes, and Harvard Business Review's 2011 audit of 2,241 companies found firms that made contact within an hour were about seven times as likely to qualify the lead. Both are verified at their publishers on the missed call statistics page.
AI's structural advantage here is simple: the call is answered while the caller is still on the line, so there is no callback gap, and the details reach your team typically within seconds of the call ending (measured median on oncrew.ai/trust). A voicemail waits until someone listens to it.
What the math doesn't capture
Numbers don't show:
- Owner sleep quality. A 1-truck owner who stops getting woken for routine calls has time and energy back. Hard to dollarize but real.
- Reputation effects. Customers who get answered consistently leave better reviews. Compound over years.
- Team capacity. Dispatchers freed from intake handle higher-value tasks (estimating, route optimization).
- Owner ability to grow. Time spent answering calls is time not spent selling, hiring, or building the business.
These second-order effects are usually the deciding factor for shops that switch.
The simple test
Pull your last 90 days of inbound calls. Count missed calls. Apply the math above with your real conversion rates. Compare against the cost of an AI receptionist plan.
If the incremental revenue is more than 5x the cost, the math is decisive. If it's 2-5x, the math says go, but pilot for two weeks to confirm. If it's under 2x, your call mix probably doesn't fit the AI category. Find out why before you buy.
For more, see the contractor missed-call cost benchmark, the contractor missed-call playbook, the verified missed call statistics roundup, and the missed call calculator.
FAQs
What if my missed-call rate is already under 10%?
Then the ROI math is much smaller. Answering services aren't the right investment for shops with strong existing phone coverage. Look at other operations bottlenecks instead.
Does this math include the cost of the trucks rolling on captured leads?
No. The revenue numbers above are gross, not net. Subtract your job costs (parts, labor, overhead) to get the profit lift. In both worked examples, the gross profit at a 30-40% margin is still many times the plan cost.
How fast does the ROI show up?
Calls are answered from the first day your line forwards, so your first month of calls is the first real read. Compare that month's jobs from captured calls against the jobs you got from missed-call callbacks before.
What if the AI doesn't convert as well as expected?
Run a 14-day pilot with full transparency on which calls converted. If the conversion drops more than 10% from your live baseline, the AI script needs work: listen to the failed calls and refine. If after refinement it still underperforms, the trade-specific fit may not be there.