Home Services: Win More Jobs From the Calls You Already Get
For HVAC, plumbing, roofing and remodeling companies: AI reviews every call, tracks each estimate by stage and turns promised callbacks into tasks.
In home services, the sale rarely happens on the first call. A homeowner calls because the AC is blowing warm air, the CSR books a visit, a technician or comfort advisor goes out and presents options, and then someone has to get a yes and put the install on the schedule.
That can take days or weeks, and every handoff is a chance to lose the job. The estimate was promised “by Friday,” the homeowner didn’t hear back, and the company that called first got the work. In the CRM, it’s logged as “went with someone else.”
Most of those jobs aren’t lost on price. They’re lost to a missed call, an estimate nobody followed up on or a call that ended without a date. Omniscore reviews every call and shows you where that happens.
Where home service jobs get lost
- Missed calls nobody returned. In the first hot week of summer, every line is busy. A homeowner with no AC doesn’t wait for a callback — they call the next company on the list.
- Calls that end without a next step. “Let me talk to my husband and call you back” — and nobody agrees on when. Without a date for the visit or the next call, the lead is as good as closed.
- Unsold estimates with no follow-up. The comfort advisor leaves three options on the kitchen table, the homeowner wants to think it over, and nobody calls to walk through them. A week later, a competitor’s lower bid is on the same table.
- “That’s too expensive” with no answer. The rep states the price and goes quiet instead of explaining what’s included — permits, warranty, cleanup — and asking what the homeowner is comparing it with.
- Promises nobody wrote down. “I’ll check with the install crew and call you back with a date” — said on the call, forgotten right after it.
- Jobs stuck in a stage. Without a time limit, an estimate that has waited three weeks for approval looks the same as one sent yesterday.
How Omniscore handles it
Omniscore connects to your phone system through its API or an SFTP folder (or you upload recordings) and to your team’s mailboxes over IMAP. CSRs and estimators keep their numbers and work as usual, and every conversation goes through the same steps:
- Every call reviewed in about two minutes. A transcript with speakers and timestamps, a score against your call script with a quote for every point, a summary, the outcome and the next step — for every call, not a 2–5% sample. More on the AI call analysis page.
- Booking rate, call by call. Set a booked visit as the goal of a call, and the dashboard shows what share of homeowners who wanted to book actually got on the schedule. A call where the booking was lost on the CSR’s side stays on the Today screen until someone reaches the customer.
- The call finds its job. AI links calls and emails to the customer’s deal in the CRM. If the homeowner has two open jobs — say, a water heater replacement and a new AC system — the call goes to the “Review” queue, and the rep picks the right one in one click.
- Promises become tasks. “Send the estimate by Friday,” “call back Thursday once she’s talked to her husband” — AI creates a task with a due date for the rep who made the promise, and anything overdue turns red on the task board.
- A pipeline with your stages and time limits. We set up your process — for example, new lead → visit scheduled → estimate sent → approved → installed — with an SLA for each stage. A job that sits too long is highlighted on the board and shows up on the Today screen. Lost jobs close with a reason from a fixed list — “price,” “hired another contractor,” “repaired instead of replacing,” “postponed” — so after a month you know what’s really costing you work.
- Missed calls under control. The dashboard shows missed and unreturned calls for each rep and by hour of day, so you know when the phones need another person. Only a connected call with that customer on the same day counts as a callback — not a voicemail.
The next step is the heaviest item on the scorecard
When a sale takes several steps, every call has to end with a specific agreement: who does what, and when. That’s why the next step gets the most weight on a home service scorecard — or becomes a critical item, and a call without one scores zero and lands under critical fails on the dashboard. A sample scorecard for an inbound call:
| Criterion | What AI checks |
|---|---|
| The problem | Found out what’s wrong, how old the system is and how urgent it is — no heat, no AC, water on the floor |
| Decision makers | Asked who else is part of the decision and whether they’ll be home for the visit |
| Price | Gave the service call fee or a price range and explained what the final estimate depends on |
| Visit time | Offered at least two specific arrival windows |
| Price objections | On “that’s too expensive” or “another company quoted less,” asked what they’re comparing and answered on the merits |
| Next step | Agreed on an action with a date: the visit, the estimate, a call to go over it |
| Promises and tone | Critical: rudeness, or promising a date or discount the company can’t deliver, zeroes the score |
Every point comes with an explanation and a quote from the call. If the homeowner never pushed back on price, that item is marked “not applicable” and doesn’t lower the score. You edit the scorecard yourself, without a developer.
Pilot results at an installation and service company
Nine reps and 6,300 conversations over three months, May–August 2026. No new hires and no new script — the only change was that the manager could see every call instead of one in forty. The metrics are calculated across all conversations.
| Metric | Before | After |
|---|---|---|
| Calls reviewed by the manager | a 2–5% sample | 100% |
| Close rate | 12.8% | 15.4% |
| Calls that followed the full checklist | 56% | 77% |
| Calls with an agreed next step | 47% | 66% |
| Price objections handled | 34% | 52% |
The ad budget and the team stayed the same.
What the owner or sales manager sees
- Standards compliance — how often each scorecard item is followed, across the team and for each rep. A row that’s red across the board is a training gap; one red cell is a conversation with one person.
- Agent leaderboard — score, critical fails, unreturned missed calls and the trend for every CSR and estimator. Sort by “Needs attention” to see who to talk to first.
- Pipeline report — how many jobs reach each stage, where the biggest drop is, how long jobs sit in each stage and how many are past SLA.
- The Today screen — customers waiting for a reply, missed calls not returned, conversations at risk and stalled jobs, in one list every morning.
- Ask Omni — “Who haven’t we called back?” Omni answers from your data with links and suggests the next action as a card, but never changes anything on its own.
Pricing and pilot
The Omni core is $149 a month per workspace: AI review of every call, the Today screen, reports and Omni for a team of up to 10 (each additional 5 people is $59 a month). Add the CRM ($79 a month), Tasks ($25) and Email ($15) when you need them. A second branch or brand with its own team and reports is another $19 a month. AI review is billed from a prepaid balance as you go — about 2¢ for a 1–2 minute call. Prices exclude sales tax. Build your setup in the calculator on the homepage; all the terms are on the pricing page.
The pilot is free for 14 days on your real calls: up to 1,000 minutes, no prepayment and no contract. We connect your phone system and mailboxes, turn your call script and your stages from lead to install into a scorecard and a pipeline, and check the AI’s scores against your manager’s on 20–30 calls. First results come in 3–5 days, and at the wrap-up meeting you decide whether to continue. Request a pilot.