Custom Software for Security and Alarm Companies

You run a subscription business with trucks attached

Ask most alarm company owners what they do and they will describe installs. Cameras, panels, access control, pulling low-voltage wire through a building that was not designed for it. That is the visible work, and it is the part the crew talks about.

But the balance sheet tells a different story. The install is a one-time transaction that often barely covers its own cost, especially on residential work where equipment gets discounted to win the contract. The value is in the recurring monitoring revenue that follows — a small monthly number multiplied by thousands of accounts and by years of contract term. When someone buys an alarm company, they are buying that book, not the trucks.

Which means the most important operational question in your business is not “did we finish the install.” It is “is every account we monitor being billed correctly, and how many of them will still be here next year.” Almost no office software is set up to answer that.

Where recurring revenue quietly leaks

Recurring revenue is unusually forgiving of mistakes, and that is the problem. An unbilled account does not bounce. Nobody calls to complain that you forgot to invoice them. It simply sits there, monitored and free, until someone happens to notice. The common leaks:

• Accounts that go live at the monitoring center but never get activated in billing, because the install closed on a Friday and the paperwork handoff happened by memory.
• Contractual rate escalations that were negotiated, written into the agreement, and never applied — three percent a year on a book of two thousand accounts is real money that costs nothing to collect.
• Cancellations processed on request rather than on term, so accounts stop billing months before the contract they signed actually ended.
• Equipment added at a later service call — an extra camera, a second keypad, a cellular communicator — that changes the monthly rate on paper and never changes it in the system.
• Inspections and testing that are included in some agreements and billable under others, performed identically by the technician and billed identically by the office, which means one of the two is wrong.
• Service calls covered under warranty long after the warranty ended, because nobody knows the install date without pulling a folder.

Each of these is worth a small amount per account per month. That framing is exactly why they persist. On a book of any size, and across a contract term measured in years, small amounts per account per month is the whole business.

The monitored site is the unit of everything

The single most useful structural decision is making the monitored site the primary record instead of the job. Jobs are finished and closed. A site is a relationship that keeps producing revenue and keeps generating work for years.

Once the site is the record, everything hangs off it in one place:

• The as-built system — panel type, firmware, communication path, zone list, device locations, and what got replaced when.
• The contract — start date, term length, monthly rate, escalation terms, auto-renewal language, and the date the term actually ends.
• The people — account holder, authorized users, passcodes, emergency contacts, key holders, and who is allowed to request a change.
• The municipal side — alarm permit number, expiration date, and the false alarm history that determines whether the next one carries a fine.
• The full service history, so the fourth trip to the same door contact is visible as a pattern instead of showing up as four unrelated tickets.
• The money — what this site has paid, what it costs to serve, and whether the recurring rate still reflects the system that is actually installed.

That last line is where most owners find the surprise. Some accounts have absorbed six service visits in two years on a monthly rate set in 2019. Others have never called once. Treating them as the same account is a pricing decision made by accident.

The technician's day is the data

Everything the office needs is known by a technician standing at a panel. What devices went in and where, what the signal test returned, what the customer asked for that was not on the work order, whether the existing wiring was a mess that added two hours.

If that knowledge only makes it back as a handwritten sheet and a conversation, you will re-learn it the hard way on the next visit. Capture has to be fast enough to survive a real workday:

• Device-level as-built capture — scan or enter the serial, pick the location, done — so the system record matches the building instead of matching the proposal.
• Photos of panel interiors, wiring runs, and camera fields of view, timestamped and attached to the site, which saves the next technician an hour of tracing and settles disputes about what was installed.
• Signal test confirmation recorded as an event, so activation is triggered by a verified test rather than by an assumption.
• Change orders captured on site, at the moment the customer asks for the extra camera, with a signature — because that conversation is worth money only if it exists in writing.
• Labor and materials logged against the site so job cost is real, not a percentage assumption applied after the fact.
• Offline tolerance, since a good share of this work happens in mechanical rooms, basements, and metal buildings where signal dies.

The scheduling side of this is the same problem every trade faces — the right technician, with the right parts, at a time the building is actually open — and it is covered in more depth in field service scheduling software for small companies. What is different here is that the visit updates a permanent asset record, not just a ticket.

Attrition is a calendar problem before it is a sales problem

Cancellations rarely arrive out of nowhere. They cluster around predictable moments: the end of a term, a property sale, a business closing or moving, a competitor knocking during a renewal window, or a run of service calls that wore out the customer’s patience.

All of those are visible in data you already have, if anything is watching:

• A rolling view of accounts whose term ends in the next ninety days, so renewal conversations happen before the customer starts shopping.
• Accounts with more than a set number of service calls in a rolling year, which is your best predictor of a cancellation nobody saw coming.
• Accounts with repeated false alarms or permit problems, since municipal fines sour customers on the whole system.
• Sites where the panel has not communicated in weeks, which sometimes means a failed communicator and sometimes means the customer already left.
• Attrition rate by segment — residential versus commercial, self-generated versus acquired, by install year — because the average tells you nothing about which part of the book is bleeding.

A single point of attrition on a large book is worth more than a good month of new installs, and it is far cheaper to defend. That is the calculation most alarm companies never run because the numbers live in three systems.

Compliance dates that arrive whether you track them or not

Fire alarm inspection intervals, extinguisher and backflow schedules on combined-service companies, municipal permit renewals, technician licensing and continuing education, insurance certificates for commercial customers. None of these are optional, and all of them are commonly tracked in a spreadsheet that one person maintains.

The failure mode is not dramatic. A quarterly inspection slips to five months. A permit lapses and the next dispatch turns into a fine. A certification expires and a technician is on a job they should not be on. A system that generates the work from the due date — rather than waiting for a customer to call — turns compliance from a liability into a recurring, schedulable, billable revenue line.

Why the off-the-shelf options frustrate people

There are real alarm-industry platforms, and plenty of companies run on them well. The complaints that push owners toward a custom build are consistent:

• Recurring billing is rigid — one rate, one term, one escalation model — and the terms on an acquired book or a negotiated commercial agreement have to be handled by hand every cycle.
• The field app and the billing system are separate products from separate vendors, so the as-built list, the change order, and the invoice never quite agree.
• Reporting stops at revenue totals and cannot produce attrition by segment, cost to serve per account, or margin on an account over its life.
• Adding an adjacent line — access control, fire inspections, IT and networking, gate systems — means bending fields that were designed for burglar alarm accounts only.
• Per-account or per-user pricing scales with your book whether or not the capability you are paying for grows with it.

The honest framing is that custom is not automatically the right answer. A company running straightforward residential monitoring on a single rate through one central station will be well served by a packaged product at lower cost. The case for building gets strong when the book is mixed — multiple contract forms, acquired accounts, commercial fire alongside residential burglar, service lines sharing the same technicians — because that mix is precisely what general products handle by making you work around them. That is the same test laid out in custom software versus off-the-shelf.

What this connects to

A system for this business earns its keep by joining things that are currently separate:

• Accounting, so recurring invoices, install billing, and service charges post without anyone retyping them.
• Automated payment collection on the recurring side, since card and ACH on file is the difference between predictable revenue and a monthly collections chore.
• The central station, so account activation, status changes, and cancellations move as data instead of as email requests.
• A customer portal for commercial accounts, where a facilities manager can see inspection records, open a service request, and update a call list without phoning your office.
• Proposal and contract signing, so a signed agreement creates the account, the term, and the recurring rate in one step rather than three.

The accounting link is worth doing properly rather than as an afterthought. Recurring billing re-keyed by hand every month is both the most repetitive task in the office and the one where a small error repeats forever, which is exactly the problem a QuickBooks integration is meant to remove.

Built around how your operation actually runs

Brad Walker has spent more than twenty years building operational software, working with dispatch-driven and recurring-revenue businesses from his base in Wake Forest, NC. Alarm and security work always starts with the same questions: how many accounts do you monitor, how many of them are billed correctly, what does a technician capture in the field, and what is your attrition doing by segment. The answers decide what gets built and what gets left alone.

Engagements are fixed price, with the scope agreed before development starts. You know what you are getting, what it costs, and when your office and your technicians will have it.

Frequently asked questions

If you cannot produce a list of every account you monitor next to what each one is billed, or your renewal dates live in a spreadsheet one person maintains, that is a fixable systems problem. Start the conversation. The first step is a discovery call to map how accounts, installs, and recurring billing move through your operation today and where they are costing you.