How owners find where a growing year’s profit went by comparing each completed job with the estimate that won it, whether they ever sell or not.
A good year can be harder to read than a slow one.
The third quarter of 2026 closed on September 30, and the year-to-date numbers are reaching owners’ desks this month. In a growing company they often say two things at once: revenue is up, and profit has not grown at the same pace. This piece is about finding out why, one completed job at a time, before next year’s work is estimated.
If that is your statement, many of the industry’s largest companies reported the same trend. Together, the companies in SDM’s ranking of the largest systems integrators reported revenue up 15 percent for 2025. Just over half of them, 53 percent, said their net profit margin increased. In each of the two years before, 60 percent or more said so.
A financial statement can show that the gap is there. It cannot show where it opened. A job closeout comparison can: the estimate that won a job, with a second column added for the actual results. Every completed job gets one. The information already exists in the company’s own records, and the work pays off whether you ever sell the company or not.
A practice your technicians already know
A commercial project ends with a set of as-built drawings: the system as it was installed, which is seldom exactly the system as it was designed. A wall turned out to be concrete. The customer added a door. The corrections are no criticism of the designer. The drawing is corrected because the next technician on that site needs it to be true.
The estimate that priced the job usually stays as written, because the next one is already due. A job closeout comparison gives the estimate the same treatment: what the job actually required, beside what the company expected.
That matters because a lower margin can hold two different things. Part of it can be deliberate. A company may price an installation close on purpose, to begin a monitoring or service agreement that will bill monthly for years. That part is an investment.
The rest is hours and materials nobody planned, on jobs that were estimated one way and built another. One integrator in SDM’s ranking traced part of its 2025 margin pressure to price increases between the time of proposal and the time of install.
On the financial statement, the two look the same. A closeout comparison tells them apart, line by line. After a price increase, for example, the same material sits on the same line at two different costs: one in the estimate column and a higher one in the actual column.
What the comparison gives back
The first return goes to the people who do the work. A job estimated at the hours it takes is a job staffed at the hours it takes, on a schedule the crew can keep. The last week on site is not likely the week a crew makes up for a number quoted months earlier, before anyone had seen above the ceiling. Technicians notice a well-estimated job before anyone in the office does.
The second goes to whoever writes the estimates. An estimator sends a number out and seldom learns how it came back. A second column that shows them the actual results, job after job, is a feedback loop the role almost never gets, and a site that differed from the drawings is nobody’s miss.
The third is yours. After you have reviewed dozens of completed jobs, patterns begin to emerge, revealing which types of projects consistently achieve their estimated profit margins and which routinely fall short. That is a growth decision with a record behind it: where to add a crew, and which kinds of work to pursue.
What goes in the second column
Start with the estimate exactly as it was sent, and leave it alone. The second column runs beside it, line for line, with the actual results. A few more lines go underneath, for things no estimate carries.
- Contract value and change orders: the amount signed, then each change to it, marked as billed, approved and waiting to be billed, or done as a courtesy.
- Labor hours: the hours estimated and the hours worked.
- Equipment, materials, and outside costs: what was estimated, and what the purchase orders, subcontractor invoices, permits, and rentals total.
- Return visits: each visit after handover that belongs to the installation, with its hours.
- Three dates: the last day on site, the final invoice, and the first recurring invoice.
- Recurring revenue: the monthly monitoring, service, or inspection amount the estimate assumed, and the amount now billing.
- One sentence: what caused the largest difference, and what the next estimate for this kind of work should assume.
Total each column and there are two margins to compare: the one the job was estimated at and the one it actually earned. Complete the comparison when the job closes, while the details are fresh. Return visits and the first recurring invoice are added as they happen. Whoever keeps the books prepares it with the project manager, and the estimator sees the result.
The records are seldom in one place. Often the estimate lives in one system, the hours in another, the purchase orders in a third, and the invoices in the accounting file. In a growing year, the crew is often on the next job before anyone sets them side by side.
Four lines that deserve a closer look
Each one points to a different decision.
Labor hours
Labor is the line most worth a second column, because its cost has been moving. The Bureau of Labor Statistics put the median wage for security and fire alarm systems installers at $28.88 an hour in May 2025, up from $27.13 two years earlier. That is a little more than 6 percent, and it is the market paying for a skill it is short of.
The figure is a national median, and your own payroll is the number that counts. The point for the estimate is narrow: a labor rate set in 2023 is worth checking against what your technicians cost today, with benefits and payroll taxes included.
The count of hours matters as much as the rate. An overrun in the second column is a reading on the estimate. It may say the site was not what the drawings showed, or the equipment arrived after the crew did, or a specific kind of system takes longer than the template assumes. Each has its own fix, and none of them is asking the crew to hurry.
Return visits
After a system is handed over, a technician often goes back. A door needs a small adjustment once the building is back in daily use. The customer’s staff want a second walk through the system after a week of using it. Those visits are part of finishing the job, and in many companies their hours land on the service schedule, because service is who gets the call.
The effect can run in two directions: installation looks more profitable than it was, and service looks less. An owner reading those two numbers could reasonably put the next dollar in the wrong place. Counting return visits with the job they belong to puts the hours where the decision was made.
They are also the customer’s first experience of the company’s service. We wrote in the attrition ledger that the reason a customer leaves is worth recording, and service is one of the categories. The first 90 days are when a new customer learns what kind of company they hired.
Three dates
The last day on site, the final invoice, and the first recurring invoice. In a busy quarter they can drift apart for ordinary reasons: a signature is outstanding, a punch list is holding the final invoice, or the service agreement is sitting in a folder until someone tells billing the system is live.
Each gap means something different. The first is cash: the work is finished and the final invoice has not gone out. The second is the recurring base, which grows on the day billing starts, whatever day the agreement was signed.
Both usually close at the handoff between operations and accounting, with one decision: who tells the office a job is done and its system is live, and on what day. A clean handoff serves the customer too, because the first invoice arrives when the agreement said it would. A job that reaches its first recurring invoice before January begins 2027 with a full year ahead of it.
The recurring revenue a job starts
Job costing usually stops when the job does. A commercial security project is often two sales: the installation, and the monitoring, service, or inspection agreement meant to follow it. The estimate assumed the second. This line records whether it started, and at what monthly amount.
It also answers the question the financial statement could not. A job priced close on purpose, for a customer the company plans to serve for fifteen years, is an investment once that agreement is billing. Where the agreement was proposed and never signed, the same job is simply a low-margin one.
Nobody decided that. The service proposal was going to follow the installation, and the installation ended on a busy week. What the line asks of an owner is a choice: whether the service agreement goes out with the installation proposal or after it, and who sends it.
Start with five completed jobs
Pull the five largest jobs the company completed in the last twelve months and add the second column to each. No new system is needed to begin. Then read them together by kind of work, never by person: fire alarm beside access control, new construction beside retrofit, a single site beside a rollout across many.
What comes out is a short list of assumptions to change, and it goes two places. The estimating template gets the hours and costs the completed jobs reported. Next year’s plan gets a margin the company has already achieved.
We wrote in the revenue bridge that the project side of the business walks on its own page, with a margin assumption written on it. This is where that number comes from. The proposals going out before the end of 2026 become the jobs of early 2027, so the template is worth correcting before they are written.
What a serious reader sees
A banker, an insurer, a successor, or a partner tends to ask one thing about a company’s margin: does the company know why the number is what it is? A stack of job closeout comparisons answers before anyone asks. It shows a company that estimates from its own record and corrects its assumptions in writing. In the five fundamentals we called that a story the numbers back up.
Where we stand
A company that corrects its estimates is keeping a promise to the people who build its jobs and the customers who pay for them: the next number will be a closer one. We built Halo Service Partners on the conviction that in commercial security, trust is the product, and a job closeout comparison is one place that trust is written down. Continuity is the center of how we partner, because the trusted local company is the thing worth protecting. The name stays, the team stays, and the customer relationships stay. What changes is what stands behind them: shared resources across the network, finance among them, so an owner taking a closer look at job margins is not doing it alone, and a company can take the step it has been weighing. Founders choose their own path forward: step back, stay on to lead, or take a larger role across the network.
ProTech Security and Verified Security both run under their own names today, with the leaders and teams who built them still serving the customers who know them.
If you’re thinking about the future
If you own a commercial security company and you are thinking about its future, in any direction, we are glad to compare notes: on how companies like yours estimate and finish their work, on the market, and on what your next chapter could look like. The five closeout comparisons are worth doing for their own sake, before next year’s plan is final. If they leave you with a question you would like to talk through, bring it with you.