What a $5M Contractor's Back Office Should Look Like

From Handy Ventures

Around $5 million in revenue, the back office stops being something you can improvise. Below that, a capable office manager, a bookkeeper, and an owner who works nights can hold it together. Past it, the volume of jobs, vendors, and payroll starts breaking the improvisation, and every contractor faces the same fork: keep hiring until the office can keep up, or figure out what the office is actually for.

Most choose the first path by default, one urgent hire at a time. This article is about the second path, and it starts with a list.

The eleven functions

Strip away titles and org charts, and a $5 million contractor's back office performs about eleven functions. Estimating turns opportunities into priced proposals. Sales tracking keeps the pipeline from living in the owner's memory. Project setup turns a signed contract into budgets, schedules, and files. Scheduling sequences crews and subs. Field communication moves information between the job and the office. Time capture gets hours onto jobs. Payroll pays people correctly and on time. Accounts payable receives, codes, and pays vendor bills. Billing turns completed work into invoices and progress claims. Job costing attaches every dollar to the job it belongs to. Reporting tells you what's true.

Every contractor performs all eleven, whether or not anyone owns them. The only variables are how well, how late, and at what cost. If a function has no owner, the owner of the company is performing it, badly, at midnight.

The default solution assigns humans to the list. An office manager takes four of the functions, the bookkeeper takes three, a project coordinator takes the rest, and when things slip, a fourth hire splits the difference. By $5 million, that's commonly three to five office salaries, somewhere between $250,000 and $400,000 a year, before software, benefits, or the management time the office itself consumes. Five to eight percent of revenue, spent keeping information moving.

The three kinds of work hiding in the list

The trouble with hiring against the list is that each function is not one kind of work. Look inside any of them and you find three different kinds of work braided together.

There's judgment: pricing the tricky estimate, deciding whether a change is worth the fight, reading a job report and knowing which number is lying. Judgment needs a person, arguably your best person.

There's specialization: work with real rules attached, done best by dedicated tools or trained professionals. The general ledger. Payroll tax. Takeoffs. You don't want creativity here; you want correctness.

And there's movement: taking information that already exists in one place and putting it in another. The estimate retyped into a budget. The timecard keyed into payroll and again into job costs. The delivery ticket matched to the PO and coded to the job. The schedule updated to reflect a phone call. Movement is most of the volume in a back office, and it's the part that eats the salaries, because it grows with every job you add.

The default office assigns whole humans to bundles of all three, which means expensive judgment gets spent on movement, and movement work expands to fill every seat you add. That's the design flaw, and no amount of good hiring fixes a design flaw.

The architecture

A better back office starts by sorting the eleven functions into their parts and matching each part to what performs it best.

Specialized software where specialization is the point: the accounting system keeps the ledger, the estimating tool does the takeoffs. These tools are good at their jobs. The mistake was ever expecting them to also connect to everything else.

Automation where information needs to move: the estimate becomes the budget without being retyped. Hours flow from the field to the job to payroll. Vendor bills land coded against the project they belong to. A schedule change reaches the people it affects. Every one of these is a route, and routes can run themselves.

People where judgment is required: someone who reads the exceptions, prices the changes, makes the calls that need a human voice, and decides. Fewer seats than the default office, but each seat spending its day on the work that justifies a salary.

Build it that way and the office at $5 million looks different. Not empty, but small and senior: an operations seat that manages exceptions instead of typing, finance expertise sized to what you need rather than a full department, and a set of routes underneath them that move the daily information on their own. The eleven functions all still happen. They just stop each demanding a person.

The point isn't headcount for its own sake. It's what the structure does next: when you grow from $5 million toward $8 million, the routes carry the extra volume, and the people carry the extra judgment. The office stops scaling with revenue, which, as covered elsewhere on this site, is precisely where contractor margins go to die.

Mark up your own list

Take the eleven functions and write them down the side of a page. For each one, mark honestly how it works in your shop today: J where judgment is doing the work, S where a specialized tool or pro has it handled, M where a person is moving information by hand.

Then count the M's, and notice whose salaries are attached to them.

That count is your overhead problem, stated precisely. It's also the good news on the page, because movement is the one kind of work you can take off people without losing anything. The judgment stays. The specialization stays. The retyping goes, and what's left is a back office shaped like the business it serves, instead of one shaped like its history of emergencies.

If it's a question you're sitting with right now, we'd rather talk it through than have you read about it.