Numbers with a dotted underline — and the bars on most charts — open the invoices, bills or transactions behind them. Each list can be searched and downloaded as a CSV.
Month-end balances from the balance sheet. The credit line moved from City Bank to FFB in July 2026.
Payables rose from $3.4M at Dec 2025 to . Customers take about 75 days to pay.
This bridge is a fixed reconstruction, not a live figure — the eleven components below are hardcoded in the page and nothing recomputes them, so they are labelled with the period they were built for rather than inheriting the board's date. Over Jan 2024 – Sep 2026 they sum to a $0.47M rise in cash. Cash at Sep 30 2026 was actually $579,305 against $0.16M at Dec 2023, a $0.42M rise, so the reconstruction is about $47K out and the "Other" line is carrying it. Rebuilding it against the general ledger is an open item. Operating cash was about +$1.6M, and all of it came from paying vendors more slowly (+$5.5M) while customers came to owe $5.9M more and inventory and work in progress grew $1.5M. Cowboy bought $8.1M of equipment, funded mainly by the City Bank/FFB equipment loan (+$5.1M, net of other debt), the credit line ($2.25M) and $0.9M of owner contributions (mostly Fily, 2024). Related-party loans were a +$0.9M source in 2024 and a −$1.9M use in 2025–26. Depreciation is added back. "Other" is small unexplained balance-sheet movement.
Accrual basis. The charts show the last 12 months; the table below covers Jan 2024 – . Monthly revenue and EBITDA are read live from QuickBooks to , so September and October reflect costs that reached the ledger after the transaction export — the drill-downs behind them still run to . Bills arrive three to six weeks after the work, so both months can still move.
From P&L by Class, before depreciation. The chart shows the last 12 months; the margin table covers Jan 2024 – Oct 2026. Overhead is already spread across the divisions in QBO. PumpDown and Kill Truck weren't tracked as divisions in 2024.
What each division billed and what it spent, month by month over the last 12 months, with the result as a line. Sandbuster sells shaker screens and Mixing / Chemical sells chemicals and filter elements; the switch takes those sales and their cost out of both sides so you can see what the service and rental work earns on its own. Every cost that isn't the product itself — labor, trucks, rentals, fuel, overhead — sits on the service side, so the product-off view is a worst case for services.
Chemical costs run about a month behind. September and October revenue is close to final, but their costs may not be: the chemical invoices for that work land later, so the last two bars can understate costs and overstate the result. Those months are shown in grey.
Diamondback's share is falling as other customers grow.
Standard terms run 30–75 days.
Account owners from Cowboy's sales list (updated Sept 16, 2026). The same person covers every division in both regions for each customer, except BPX (Ricky Sanchez in West Texas, Teresa Franklin in South Texas) and Permian Resources' West Texas PumpDown work (Bradley Carr). Shared accounts are shown on their own row rather than split. Contribution uses division-average margins; Bayswater's PumpDown work makes the unassigned group's margin low. Known CAC (Jan 2024 – Sep 2026) is the acquisition payments identified so far, plus meals, events, gifts, sponsorships and promotions whose memos name the customer or one of its job numbers.
Estimated profit each customer contributes: its revenue in each division × that division's contribution margin (revenue minus costs that rise and fall with activity). Division margins: Sandbuster 41.5%, Roustabout 33.1%, Mixing / Chemical 29.2%, Kill Truck 23.5%, PumpDown −40.6%. Revenue and contribution run Jan 2024 – Sep 2026; per-year figures cover each customer's months since first billed. LTV assumes a 3-year customer life. 2024 history supports that for large accounts: of 31 customers billed in 2024, only 8 still bill in 2026, but those 8 were 84% of 2024 revenue. Small one-off customers rarely last past a year.
Billing at the line-item level from both operating systems — Aimsio and BlackGold — so this is the whole company's billing, not one system's view of it. Every line reconciles to its own invoice total, and the two systems together tie to QuickBooks month by month
Rolling 12 months, invoiced. Service is the work and the equipment on it; Material is chemicals, screens and filter elements. The last month is greyed because it is not complete — billing runs to and invoicing lags the work, so it reads low rather than because the work stopped.
| Item | Unit | Qty | Billed |
|---|
Quantity is in each item's own unit, so the column compares within a row, not down it. The two systems name the same service differently — Aimsio tags an item with the division that sold it, BlackGold does not — so names are matched on the item and its unit. A dot marks an item billed in both systems.
From accounting's A/R submitted-dates tracker, as of , using each customer's payment terms. Expected dates use the terms or the customer's actual 2026 pay speed, whichever is slower.
"Oldest" is days since the oldest invoice date for approvals, or days past the due date for past-due invoices.
"Actual days" is the dollar-weighted average from invoice date to payment, for invoices paid in 2026. Diamondback's 60 days start at ticket approval, so its invoice-to-cash time is longer.
From the A/P Aging Detail as of : netting .
Payroll comes on top of these bills, about $440K every two weeks at August's level.
From Payroll Details. The charts and the division table below cover complete months to ; the per-employee detail runs through the pay date. Totals jump in months with three paydays.
Labor is about 36% of revenue. Counting lodging, per diem and travel, people costs are about 41%. Contractors are the distinct payees billed to Contract Labor that month — a payee can be a firm rather than one person, so it counts contracting parties, not heads. Cowboy Elite, which has run Roustabout as a subcontractor since August, is one payee covering a whole crew. The current month always looks low: contractor invoices arrive weeks after the work.
Each employee's paycheck is assigned to the division it was coded to in QuickBooks. Mechanics and logistics staff are tracked together as "Maintenance / Logistics," and office and management staff as "Admin." "Unclassified" covers paychecks coded to Admin/Overhead or left without a division; the bookkeeper spreads those to divisions later.
Share of hours paid at overtime, 2024 → 2025 → 2026 (Jan–Aug): Mixing / Chemical 41% → 45% → 48%, Sandbuster 41% → 41% → 44%, Maintenance / Logistics 36% → 40% → 47%, Roustabout 25% → 21% → 24%, Kill Truck 34% → 33% (tracked from mid-2025), PumpDown 52% (2026 only). Company-wide: 33% → 36% → 41%. In 2026, Mixing / Chemical employees averaged about 118 overtime hours a month each, and PumpDown employees about 131.
Loaded Dice Safety bills FR clothing and PPE with the recipient's name in the line memo. These are the lines that name someone, matched to the employee roster; lines for bulk stock, closet restocks and consumables carry no name and are not traced here. A line naming several people is split evenly between them.
| Employee | Issued |
|---|
| Division | Issued |
|---|
Pay, hours, overtime, company card, accounts and contact details for anyone on the payroll or in the employee directory.
Both sources together: WEX cards — retail purchases on a card issued to a person — and yard fuel, delivered in bulk into the Midland and San Antonio tanks. Yard fuel is the larger of the two and was on neither board until now. Twelve complete months, .
| Fuel | Gallons | Share | Card cost | Card $/gal |
|---|
Yard fuel moves with tank deliveries, which arrive in lumps, while card spend tracks daily activity — so read the yard series as a trailing total rather than month by month. The window stops at August because the ledger's fuel tail is not booked: September carries one bulk delivery, and the fuel-card journal entries stopped after July. That is a posting gap, not a drop in fuel use.
| Division | Cards | Yard | Combined | Card gal | Yard gal | % of revenue |
|---|
How yard fuel is assigned here: by who pulled it at the pump, divided by every gallon pumped — so the nobody signed for lands nowhere instead of inflating the divisions that did. The QuickBooks class on the bulk purchase says something different: it puts all of it on the four revenue divisions and none on Maintenance / Logistics, which pulled the third-largest share. The gap is set out in the project notes. Divisions that bill nothing show no percentage. Fuel is not included in the division margins elsewhere on this board.
Job-level costing: invoice lines, material cost and crew hours matched to the same COS job number. It covers the work where the job number reaches all the way through, so it is a sample of the business rather than all of it — but within that sample the margins are measured, not allocated.
Revenue line by line, materials booked to the job, and every person who logged time against the job number. Narrow the list with the filters, sort by any column, then pick a job.
The two halves of the rent-versus-buy question: what owned equipment returns, and what rented equipment costs. Returns are before depreciation.
Mixing / Chemical assets include $0.80M of transfer pumps; its income excludes the preliminary August figures. PumpDown's loss is cumulative since April 2026 while it ramps up.
Rented equipment and trucks cost $7.2M over the period, about 15% of revenue.
Chemicals only, from the QuickBooks inventory register. Billed volume and consumed volume are different numbers here: many jobs carry a minimum-volume agreement, so more is billed than leaves the yard, and a reconciliation corrects the difference one to two months later.
A reconciliation is an Inventory Qty Adjust in QuickBooks. It corrects both directions — adding back volume that was billed but not used, and removing volume the physical count says has gone. Without it, neither the balance nor the month's cost of sales is final.
The operating software Cowboy is paying Blackgold to build. It replaced Aimsio at the end of October (the Aimsio subscription was not renewed) and is to replace ClockShark by the end of December.
| Item | Amount | Notes |
|---|---|---|
| Development paid to date | $186K | Since Jul 2025; $10.6K a month since May 2026 |
| Remaining (Oct–Dec 2026, est.) | $32K | Aimsio replacement complete at end of Oct; ClockShark replacement by end of Dec |
| Total project (est.) | ≈ $218K | One-time; development payments stop at completion |
| Annual subscription savings | ≈ $47K | Aimsio $34.6K (not renewed in Nov 2026, so now saved) + ClockShark ≈ $12K |
| Payback on subscriptions alone | ≈ 4.6 yrs | Faster invoicing, overtime control and job costing would shorten it. Hosting and support costs after launch would lengthen it. |
What this payback leaves out. The 4.6 years counts cancelled subscriptions only. It ignores the time savings in reducing timekeeping and job management into a single application. It also ignores 3 things that are not cost savings: the maintenance and asset management that major operators require of their vendors, the marketing opportunities of a technology different than competitors, and a platform that a potential buyer could assign value to at an exit.
The FFB Loan Agreement's tests set against the ledger, rebuilt every refresh. Internal: this section is on the Owners board only.
Where these numbers come from: the data lineage map — every source, script and file behind each figure on this page.