Security and Alarm Installer Automation Stack in 2026: Surveys, Takeovers, and RMR

Alarm companies run two businesses through one phone number

Almost every trade sells one kind of revenue. Security and alarm companies sell two simultaneously, and they behave nothing alike.

The first is project revenue — an install with equipment cost, labor hours, wiring, permits, and a completion date. The second is recurring monthly revenue from monitoring, which is the reason a well-run alarm company is worth a multiple that a well-run plumbing company is not. RMR is an annuity: it accrues whether or not anyone answered the phone today, it compounds with every site added, and it is the number a buyer actually values. As of August 2026, that dual-revenue shape is still the single biggest reason generic field-service software fits this trade badly.

Running both through one phone line, one calendar, and one billing system is where most alarm companies quietly break. The pattern separating the operations that scale is that they stopped treating every inbound call as the same event and stopped treating monitoring billing as an afterthought bolted onto invoicing. This guide covers the whole stack — call triage, site surveys, takeovers, wired versus wireless implications, install scheduling, monitoring and RMR, false alarms and permits, commercial access control, and the after-hours alarm call — mapped to the Run with Jarvis platform.

Why alarm companies lose calls in a particular way

Technicians are in attics, crawlspaces, and ceiling grids. Pulling cable above a drop ceiling or terminating in a low-voltage closet is not a state you interrupt for a phone call. In a shop with three technicians and no dedicated office, the phone is essentially unattended between 8 AM and 5 PM.

Three different callers share one number and need three different conversations. A homeowner shopping for cameras, a landlord who just bought a building with an abandoned panel on the wall, and an existing customer whose siren has been shrieking since 2 AM all dial the same number. Handling them identically fails all three.

The urgent calls arrive outside business hours by definition. Alarms go off at night. Break-ins happen at night. A customer who cannot reach anyone at 3 AM will call the company on the yard sign next door tomorrow, and the churn of a monitoring account costs vastly more than the call did. The general handling pattern is in the after-hours playbook.

Sales calls are long and consultative. A camera-coverage conversation is not a two-minute price quote. Missing it and calling back the next day loses to whoever picked up, which is the decay described in the speed-to-lead guide.

Layer 1: An AI receptionist that triages before it sells

The foundation is KeyBot, the AI phone agent. It answers on the first ring in English or Spanish, at any hour, and — critically for this trade — its first job is not to sell, it is to classify.

The classification tree that matters:

Existing customer with an active alarm or trouble condition. Highest priority, service path, immediate escalation to on-call. Never route this into a sales script.

Existing customer with a non-urgent service need — a low battery beep, a sensor that stopped reporting, a keypad that needs a code changed, a camera that went offline. Books a service appointment against real technician availability.

New install prospect, residential. Needs a consultative conversation and, almost always, a site survey appointment rather than a phone quote.

New install prospect, commercial. Different scale, different decision process, different products — access control, door hardware, network video, sometimes fire integration. Should be captured and routed to whoever handles commercial rather than quoted like a house.

Takeover of an existing system. A distinct product with a distinct interview, covered below.

Monitoring-only inquiry. Someone with functioning equipment who wants a monitoring contract, which is the highest-margin call in the entire trade and the one most often handled poorly.

Parallel answering and script fidelity are both real advantages here, but the classification is the thing. An answering service takes a message and hands you a name, which is close to useless when the difference between two of those categories is a technician being woken up or not. A phone tree is worse still, because it asks a frightened customer to navigate a menu.

Layer 2: The site survey is its own appointment type

This is where alarm companies most often mismanage their calendar.

A site survey is not an estimate visit and it is not an install. It is a walkthrough that produces the information the install depends on:

• Camera count and placement, which is a coverage question rather than a quantity question. Where the entry points are, where the driveway sightline is, which corners cover the most ground, whether an angle looks into a neighbor's window, whether there is a mounting surface at the right height.
• Existing infrastructure. Panel location, existing wiring runs, whether there is conduit, whether the attic is accessible, how the exterior walls are constructed, whether stucco or brick makes a run impractical.
• Network capacity and placement. Where the router is, whether there is PoE capability, whether recording will be local, whether the upload bandwidth supports the camera count, whether there is a suitable location for a recorder that stays cool and locked.
• Power availability, including where transformers can live and whether a circuit is available near the panel.
• Lighting conditions at night, which determine camera selection more than resolution specs do.
• Who has authority. Whether the person walking with you can sign, whether it is a rental, whether an HOA or a landlord has to approve exterior mounting.

That is a real appointment with a real duration, and it needs to occupy a slot on a specific person's calendar — usually a senior technician or the owner, not a junior installer. If your calendar treats "estimate" as a generic thirty-minute block, you will either underrun a complex commercial walkthrough or waste a lead technician's morning on a two-camera doorbell job.

GetTimePad holds separate appointment types with separate durations and separate assignable resources, so the AI can book a residential survey into a wide window and route a commercial survey to the person who handles commercial. The booking mechanics are in how AI appointment booking works.

The survey outputs belong in the CRM record, attached to the site, permanently. Camera positions, panel model, zone list, wire routes, network details, gate codes, and the customer's stated concerns. That record is what makes the second visit cheap, and it is what makes a service call two years later something other than an archaeology expedition. IntelliDrive is the CRM, dispatch, and invoicing layer holding it.

Layer 3: Takeovers, which are a separate product

A takeover — adopting an existing panel and sensors rather than installing new — is one of the fastest paths to RMR in the trade, and it is priced entirely on information a scripted interview can collect.

The takeover intake:

• What brand and model is on the wall, and roughly how old is it? The keypad usually says. Age determines whether reuse is realistic.
• Is it hardwired, wireless, or a hybrid? Hardwired sensors are frequently reusable with a new panel; proprietary wireless sensors usually are not, which turns a takeover into a replacement.
• Does the customer have the installer code? Without it, some panels cannot be reprogrammed and must be replaced. This single question determines the price more often than any other.
• How many zones and what devices? Door and window contacts, motion sensors, glass breaks, smokes, keypads, sirens.
• Who monitors it now, and is there a contract in force? A customer under contract with a competitor is a future opportunity with a date attached, not a lost lead — and that date belongs in the CRM as a follow-up rather than in someone's memory.
• Does it still communicate? A panel with a dead phone line or a decommissioned 3G radio needs a communicator regardless of anything else.
• Is there an existing permit on file with the municipality?

Getting those seven answers on the first call means the site visit either becomes unnecessary or becomes a short confirmation. Not getting them means a technician drives out to discover a panel nobody can program.

Layer 4: Install scheduling, wired versus wireless

The wiring decision drives almost every scheduling variable in this trade.

Wireless installs are faster, cleaner, and largely predictable in duration — a defined number of sensors, a panel, a communicator, a walk test. Retrofit-friendly because no walls open.

Wired and structured-wiring installs are a different animal: attic and crawlspace time, fishing walls, drilling plates, terminating at a panel or a rack, patching and painting, and enormous duration variance depending on construction. A two-story home with a finished basement and no accessible chases can turn a one-day job into three.

New construction pre-wire is a third pattern entirely, and its scheduling constraint is other trades. Pre-wire happens after framing and before drywall, in a window the builder controls and moves. Your calendar has to tolerate a job whose date is set by somebody else's inspection, which means the appointment needs to be reschedulable without losing the equipment reservation or the labor allocation.

Practical scheduling implications:

• Multi-day installs need to be modeled as multi-day, not as three unrelated appointments that a routing optimizer might scatter.
• Two-technician jobs need to reserve two technicians, which most simple calendars handle badly.
• Equipment availability gates the schedule. Booking an install before the recorder and cameras are on the shelf produces a rescheduled customer, so the CRM record and the install date have to reference the same equipment list.
• Arrival windows and status texts matter more than usual because these are long appointments in occupied homes. GPS dispatch with automatic ETA and arrival SMS removes the where-is-my-technician call, and for a business running two or three long installs a day that is most of the day's inbound noise. The multi-technician dispatch pattern is in CRM and dispatch software for multi-tech operations.

Layer 5: Monitoring, RMR, and why billing has to be two systems in one

This is the layer that separates alarm companies from every other trade in this series.

Install revenue is a project. Equipment, labor, possibly a deposit, possibly milestone billing, a final invoice, sales tax on materials, and done.

Monitoring revenue is an annuity per site. It bills monthly or annually, it continues indefinitely, it has to survive a card expiring, and it has to be reconciled against what the central station is actually billing you per account. RMR that quietly stops billing is the single most damaging silent failure in this business, because nobody notices a missing $42 charge the way they notice a missing $4,200 invoice — and twenty of those going unnoticed for a year is real money against the company's valuation.

What the operations layer has to do:

• Hold both revenue types against the same customer and site record, so you can see what a site is worth in total rather than only what its last invoice was.
• Run recurring charges on a schedule with stored payment methods and automatic retry when a card fails, plus an alert when it keeps failing. Payment-link mechanics are in getting paid faster with payment links.
• Sync to accounting cleanly, because deferred monitoring revenue and one-time install revenue are different lines in the books.
• Track contract terms and renewal dates per site, so a three-year monitoring agreement reaching term is a scheduled conversation rather than a surprise cancellation.
• Surface per-site profitability, because a site with a low monitoring rate and four service visits a year is losing money, and you cannot see that without joining service history to recurring revenue.

The upsell path lives here too. An existing monitored customer is the cheapest camera sale, doorbell sale, smart-lock sale, and sensor-addition sale you will ever make. That only happens if someone knows what each site already has, which is a CRM property rather than a sales-talent property. Asking a plain-language question like "which monitored sites have no exterior cameras" is exactly the kind of thing the Jarvis brain answers without anyone building a report.

Layer 6: False alarms, permits, and the compliance surface

Alarm work carries a regulatory and municipal layer most trades do not.

Alarm permits. Many municipalities require a registered permit per monitored address, with a fee and a renewal, and they fine for dispatches to unregistered addresses. That permit number belongs on the site record with its expiration date.

False alarm ordinances. Most jurisdictions escalate fines per false dispatch within a period, and some suspend response entirely after a threshold. A customer with recurring false alarms is a churn risk and a service problem at the same time, and the pattern is only visible if service history is logged per site rather than remembered per technician.

Installer licensing. Low-voltage and alarm installation licensing is state-regulated and often requires individual technician registration alongside company licensure. Renewal dates per technician belong somewhere that alerts before they lapse, not on a certificate in a drawer.

Central station verification procedures, including who is on the site's call list and what the verbal passcode is. Getting that wrong on a real alarm is the worst possible failure, and getting it wrong on a false one produces a police dispatch and a fine.

For consumer-protection context around contracts, cancellation, and door-to-door sales practices — a category with real enforcement history in this industry — the Federal Trade Commission publishes guidance at ftc.gov. The Bureau of Labor Statistics maintains occupational data for security and fire alarm installers at bls.gov, which is useful when you are modeling technician cost. Your state licensing board and your municipality's alarm ordinance will govern more of your day-to-day than either.

Layer 7: The 3 AM call, handled as service and not sales

Treat this as its own workflow, because it is the moment your monitoring customer decides whether they keep paying you.

The call arrives frightened or annoyed. What has to happen:

1. Identify the customer and the site immediately from the calling number, and pull the panel details and zone list.
2. Determine state: is the panel in alarm right now, in a trouble condition, or has it already been silenced? Is there a police response en route?
3. Capture which zone or device, because "the alarm went off" and "the basement glass break went off" are different problems.
4. Escalate to the on-call technician with all of the above already collected, rather than waking someone up to start an interview from zero.
5. Log it against the site so the false-alarm pattern becomes visible.
6. Schedule the follow-up service visit in the same interaction rather than leaving it to a morning callback that may not happen.

The thing to avoid is the sales script. Nobody at 3 AM wants a package comparison. A receptionist that cannot tell an emergency from a lead will eventually put a monitoring customer through a quote flow during a break-in, and that customer will leave.

Two adjacent low-voltage and project-based stacks worth reading alongside this one: the electrical contractor automation stack for the licensing, permitting, and multi-day project scheduling patterns, and the fence company automation stack for the site-survey-to-install project flow, which is structurally similar even though the product is not.

What each plan gives a security and alarm company

Every plan is month-to-month with zero setup fees, unlimited users, and no per-call fees.

The arithmetic, including the part most trades do not have

Start with installs. Assume a company completes 22 installs a month at a $2,450 average across residential camera packages, alarm installs, and small commercial jobs. That is $53,900. Assume the phone rings 190 times a month and, because everyone is on a job site, 55 calls go unanswered.

Assume only 3 of those 55 would have become installs. At $2,450 that is $7,350 in recovered monthly revenue against a Core plan at $500 plus roughly 400 overage minutes at $0.45, which is $180, for $680 total. That is more than ten times the cost, and installs are the less valuable half of the story.

Now the RMR half. If each of those 3 recovered installs attaches a monitoring account at $45 a month, that is $135 in new RMR — every month, permanently, compounding as it repeats. Twelve months of recovered installs at that rate is roughly $1,620 in monthly recurring revenue added in a year, which is the number that actually moves what your company is worth. No other trade in this series has a missed-call cost that keeps compounding after the call.

The minute math. Alarm calls run long: a consultative camera conversation or a takeover interview realistically takes six to eight minutes, while service calls are shorter. At a blended six minutes, Core's 500 included minutes cover about 83 calls. A company taking 190 calls a month runs roughly 640 overage minutes at $0.45, about $288, for an effective $788. Once you are consistently past about 1,100 minutes a month, Pro's 1,000 minutes at a $0.40 overage is the cheaper structure, and it adds attribution. The framework is in the AI receptionist ROI guide.

Where attribution starts earning its keep

Security is an expensive advertising category, and install leads are worth enough per unit that measuring them properly matters more than in low-ticket trades. Raw call volume is badly misleading, because security search terms attract people wanting a locksmith, people wanting a doorbell camera installed that they already bought, and existing customers of other companies calling the wrong number.

Pro adds Dynamic Number Insertion so each call ties to the campaign, keyword, and ad that produced it, and booked surveys feed back to Google Ads so bidding optimizes toward surveys and installs rather than rings. Transcription and lead scoring let you separate a $6,000 commercial access-control lead from a $199 camera-mounting request that arrived through the same keyword.

A sensible rollout order

Answering and triage first, with the after-hours alarm escalation path built before anything else. That is where both the recoverable revenue and the churn risk live.

Then site survey booking as a distinct appointment type, routed to the right person with the right duration.

Then CRM site records holding panel details, equipment lists, zone lists, permits, and service history.

Then dispatch, GPS, and arrival texts for long install days.

Then install invoicing and monitoring recurring billing in one place, with card-failure alerts, because that is the layer protecting your RMR.

Then review automation.

Only after all of that, if you are advertising, move up for attribution or campaign management.

The short version

Alarm companies are decided by who answers the 3 AM call, who classifies an inbound call correctly before selling anything, who walks the property before quoting cameras, and who never lets a monitoring charge quietly stop billing.

Triage before you sell. Make the site survey its own appointment with its own duration and its own assigned person. Run the takeover interview on the phone so nobody drives out to an unprogrammable panel. Model wired, wireless, and pre-wire installs as the different jobs they are. Hold project revenue and monitoring RMR against the same site record, and alert on failed recurring charges.

If you want that modeled against your own install average, monitoring rate, account count, and call volume, reach out or review the plans on the pricing page.