fabricationcash-flowdsoindustry-datametrics

What a $500K Fabrication Shop's Cash Cycle Actually Looks Like

Andrew Jacob · July 24, 2026

A fabrication shop can be profitable on every job and still spend Thursday nights wondering how to make payroll. That contradiction confuses a lot of owners, because profit and cash feel like they should be the same thing. They aren't. Profit is what the jobs earned. Cash is when that money actually lands — and the gap between the two is the cash cycle, where a healthy shop can quietly starve.

The scale of the gap is the whole story. A US Bank study found 82% of small business failures trace back to cash flow, not profit — most of those shops were making money on paper right up until they couldn't cover the next bill. Per QuickBooks aggregate data, the average small business has roughly $84,000 tied up in unpaid invoices at any given moment. For a $500K/yr fabrication shop, that's not an abstraction. That's real steel, real labor, and real hours already spent, sitting as somebody else's IOU.

Let me walk one job through the full cycle, day by day, so you can see exactly where the money gets stuck — and put a number on each stuck day.

The job: from RFQ to cash, in real days

Take a representative $10,000 fabrication job at our $500K/yr shop. Here's the honest timeline most shops actually run, not the one they'd describe:

Day 0 — RFQ arrives. A customer emails a drawing and asks for a quote.
Day 2 — quote sent. The owner was on the floor; the quote gets built and sent two days later.
Day 5 — customer says yes. Three days of silence, then a "looks good, go ahead."
Day 5–19 — the work. Material, cut, weld, finish. Fourteen days on the floor. This part, the shop is great at.
Day 22 — invoice sent. The job finished on day 19, but the invoice waited three days for a spare hour at the desk.
Day 22–49 — the wait. Invoice says net-30, customer pays at 27.5 days past invoice by the Sage average.
Day 49 — cash lands. Forty-nine days after the RFQ, thirty after the job was done, the money is finally in the account.

Forty-nine days. The customer thinks they paid on time — they paid net-30 from the invoice date. But the shop financed the whole thing from day 0.

Where the money actually gets stuck

Break that 49-day cycle into its segments and the surprise is where the delay lives. It is not in the fourteen days of fabrication — that's the productive part. It's in the handoffs around it:

Quote lag (day 0→2): 2 days. The first credible responder wins roughly half of competitive bids, so this delay costs win rate, not just time — some jobs you never book because you were second.
Decision gap (day 2→5): 3 days. Partly the customer, partly a quote that arrived without enough clarity to say yes fast.
Invoice lag (day 19→22): 3 days. Pure dead time. The work is done; the money is just unbilled because the invoice got retyped late.
Payment wait (day 22→49): 27.5 days. The biggest single segment, and largely a function of how you asked to be paid.

Of the 49 days, only 14 were the actual work. The other 35 are cycle overhead — and unlike the fabrication, most of it is controllable. The full DSO math shows how these days convert straight into a working-capital number.

What each stuck day costs

Here's the part that makes it real. At a $500K/yr run rate, the shop turns over roughly $1,370 of revenue a day. Every day of cycle overhead is, on average, a day's revenue sitting as a receivable instead of cash in the account. Trim the controllable overhead — say the 3-day invoice lag and 10 days off the payment wait — and you've pulled 13 days out of the cycle. On this shop's volume, that's on the order of $18,000 of working capital freed up, permanently, without booking a single additional job.

That's the number owners miss because it never shows up as a loss. Nobody writes down "financed the customer for 35 days." It just shows up as a tight Thursday and a line of credit you didn't think you'd need. The Cash Cycle Scorecard turns each of these gaps into a dimension you can score, so the overhead stops being invisible.

Which gap to fix first

You don't attack all 35 days at once. Two segments give almost all the return for almost none of the effort:

The invoice lag (3 days → same day). The cheapest fix on the board. The job's done; the money's earned; the only thing between you and billing is a retype. Kill the retype and this gap goes to zero. A 30-day same-day-invoicing plan is the concrete path.
The payment wait (27.5 days → under 10). Put a one-click pay link in the invoice. Customers pay slowly because mailing a check is the path of least resistance; change the default and the wait collapses.

Fix those two and our example job goes from 49 days to roughly 33 — a third of the cycle gone, on the two easiest levers, with the quote-speed win rate still sitting there as upside. The fabrication was never the problem. The shop is excellent at the fourteen days it gets paid for. The money was stuck in the thirty-five it doesn't. If you'd rather the invoice fire the day the job closes and the pay link ride along automatically — with QuickBooks synced and the cycle overhead squeezed out — Setell runs the quote-to-cash loop so the cash lands sooner. Free tier is 3 AI quotes a month; paid plans from $49/mo. Start free.

Ready to quote faster?

Start free — no credit card. See your first AI-drafted quote in minutes.