There’s a meeting that happens every quarter at most mid-market contractors. Cash conversion is on the agenda. DSO is trending the wrong way. Somebody pulls up an aging report, somebody else suggests tightening collections, and a decision gets made about dunning cadence or early-pay discounts.
Look around that room. Almost everyone in it works in finance.
That’s the problem, and it isn’t a staffing oversight. It’s a category error baked into how the metric gets reported. Billing lag lands in a finance dashboard, so it gets diagnosed as a finance problem — and the two weeks that went missing before accounting ever saw the job never come up, because nobody in the room was there when they disappeared.
The clock does not start when you think it starts
Ask when the billing cycle begins and most organizations will answer, in effect, when the invoice is generated. That’s when the system creates a record, so that’s when measurement begins.
But an invoice is a claim about work that already happened, and it can’t be issued until someone can describe that work accurately: what was done, where, by whom, how long it took, what materials went into it, and who signed off. Until that description exists in a usable form, your accounting team is not slow. They are blocked.
The real clock starts the moment a crew finishes the job. Everything between that moment and the invoice going out is billing lag, and most of it happens nowhere near accounting.
The industry-wide payment lag in construction averages 83 days, with subcontractors waiting roughly 56 days longer downstream. Nearly three-quarters of construction companies report cash flow challenges. Those are conditions of the market — payment terms you didn’t write and mostly can’t renegotiate.
But layered on top of that industry number is a stack of days that belongs entirely to you. Ryan Clayton, who manages corporate systems at pipeline contractor Atlantic Pipe Services, described exactly what that looked like:
“We were consistently about two weeks behind on invoicing. If we wait 15 days after the work is done to send the invoice, we’ve already delayed when we’re going to get paid.”
Fifteen days. Not imposed by a customer. Not negotiated into a contract. Added on top of terms the company already considered too long.
Follow one form through a Tuesday
Pick a single work order and trace it. Not the process diagram version — the actual route it takes.
A tech finishes a job at 4:40 in the afternoon. The hard part is already behind them: they climbed the structure, made the judgment call, caught the thing that wasn’t on the checklist, got the customer’s signature. The form is filled out and correct.
Now the process asks them to also be its courier. They have one more stop, then a fifty-mile drive back, in the rain, at the end of a ten-hour day. The form goes on the passenger seat, then into the glove box. It is not lost. It is not forgotten. It got outranked by actual work, which is the correct priority for a field technician to have.
Thursday, the paperwork comes in. Friday, someone in the office retypes it into the system, which produces no new information and one new opportunity for a digit to move. Monday, a PM notices a blank field and starts trying to reach a tech who is now two jobs away and reconstructing Tuesday from memory. Wednesday, the customer asks how long the crew was actually on site. The record has an address, which is where the crew was sent — not proof of where the work happened or when.
Eight days, and the invoice still hasn’t gone out. Ask who’s responsible and there’s no satisfying answer, because at no point did anyone make a mistake. Every person in that chain did the reasonable thing given what the process handed them.
That’s the part worth sitting with. A form in a filing cabinet, or a scanned PDF in a shared drive, hasn’t disappeared — it just can’t answer a question anymore. And a record that can’t answer a question can’t support an invoice, which is why “we have the documentation” and “we can bill for it” turn out to be very different statements.
Where the days actually go
Break the interval down and the missing time is rarely in one place. It’s distributed across handoffs, and each one looks small enough to ignore:
- Transit. The form has to physically travel from the site to someone who can process it. On a distributed crew, that’s measured in days, not hours.
- Re-keying. Someone in the office types up what someone in the field already wrote down. The day nets to zero in terms of new information created, and it introduces a fresh chance to get it wrong — 47% of procurement errors trace back to manual data entry.
- Clarification. A field is blank, a number is ambiguous, handwriting is unreadable. Now a PM has to reach a technician who is already on a different job to reconstruct details from memory.
- Approval chasing. A supervisor signature exists in principle. Locating it, and proving it happened before work started rather than after, is a separate errand.
- Dispute. A customer questions whether the crew was on site, or for how long. The record has an address, which is where the crew was sent — not verifiable proof of where and when the work occurred. That argument alone can hold an invoice for weeks.
Not one of those is an accounting failure. Every one of them lands on the AR report.
Let’s be clear about where the problem isn’t
When documentation comes back late or incomplete, the reflex in a lot of organizations is to look at the crew. More training. A stricter policy. A reminder email before the weekend.
That reflex is aimed at the wrong target, and everyone in the field knows it.
Your crews are on site with all the context in the world. What the process asks of them is genuinely unnatural: stop the task, switch to a maps app, copy coordinates, switch back, transcribe them into a form, then hand-carry paper through a workday that has no slack in it. Every one of those steps is friction, and friction is where data dies. A coordinate typed in later from memory isn’t evidence — it’s a guess wearing the costume of a record.
That’s not a discipline failure. It’s an architecture failure, and architecture failures have architecture fixes. You don’t close billing lag by asking the field to try harder. You close it by shortening the distance between the work and the record until there’s no room left for days to hide.
What it looks like when the distance closes
The mechanism is unglamorous: capture the data in structured form at the point of work, and let it move to the systems that need it without a human retyping anything.
When the form is digital, required fields mean it can’t be submitted incomplete, so the clarification loop never starts. Approval routing means authorization is on record before work begins rather than reconstructed afterward. GPS and timestamps are captured automatically on submission, on photos, on signatures — the field user does nothing differently, and the dispute that used to cost three weeks gets settled in thirty seconds. Completed forms flow straight into the ERP or accounting system, so the office isn’t a bottleneck; it’s a destination.
The results show up fast, because the constraint being removed was never a hard one:
- ServiceMaster by Cornerstone tracked average days-to-bill from just over 20 in April down to 8.6 by July after field documentation went digital.
- Atlantic Pipe Services got back the two weeks it had been adding on top of its customers’ terms — field notes now come in immediately.
- Elecnor, a global operation with more than 22,000 employees, cut field record processing from three to four weeks down to same day.
Read those three again and notice what none of them involved. Nobody renegotiated a payment term. Nobody added collections staff. Nobody ran a campaign to make crews more conscientious. ServiceMaster’s VP of Operations described the rollout this way: “There wasn’t any pushback when we introduced GoFormz.” The forms looked like the forms crews already used, so there was nothing to push back against.
The metric worth adopting
If you take one operational change from this, make it this one: start tracking days-to-bill — the interval from work completion to invoice issued — and report it next to DSO.
DSO tells you how long customers take to pay. Days-to-bill tells you how long you take to ask. The first one is largely a market condition. The second is entirely yours, it’s usually invisible, and it is almost always bigger than the people running the cash conversion meeting assume.
It also makes the ownership question obvious. The moment days-to-bill has a number attached, it stops being an accounting problem and becomes what it always was: a field execution problem that shows up on a finance report.
At the net margins this industry runs on — 0.38% to 2% after overhead — a 1–2% efficiency gain translates to a 20–40% improvement in net profit. Faster cash conversion isn’t a tidiness benefit. It’s leverage, on the one input you fully control.
Ready to find out how many days are hiding in your own cycle? Schedule a demo and bring your slowest-billing workflow.
Related reading:
