One Feed, Three Jobs
A jobber's office runs on other people's data. The supplier's rack prices set what you charge. The supplier's invoices say what you owe. The card batches say what your stations sold and what is coming back to them.
Three streams of numbers, arriving all day, every day — and in most offices, all three still arrive the same way: somebody logs in somewhere, downloads a file, opens it, and types what they see into another system.
That is remarkable, when you think about it, because the data is already digital at the source. If your company has access to its daily prices, supplier invoices, and card batches from its suppliers — through a data service like DTN, which delivers them as files to an FTP drop, or by email straight to your inbox — then every one of those numbers already exists as a file your software can read. The only manual step in the whole chain is the one the jobber's office added: a person in the middle, downloading files and re-typing what is in them.
This article is about removing that person from the middle — not from the business, just from the data entry. One connection, set up once, polled every half hour, feeding three parts of the office at once.
Prices. Suppliers publish rack prices daily — through the data service's files, and in price emails — and again whenever a price moves. A price file holds every supplier, terminal, and product — of which your slice is the few you actually buy. Pulled automatically, that slice arrives as an incoming price update — staged by terminal, with an accept, modify, or reject on each one, so nothing posts to your price book until a person has seen it. Accepted, it lands in the price book with its effective time attached, and the morning price sheet, the quote, and the invoice for Tuesday's load all read from the same numbers.
Supplier invoices. The same access carries the supplier's invoices for the fuel you lifted. Pulled automatically, an invoice does not land as a PDF to be re-keyed — it arrives parsed, as a staged supplier invoice with its loads, gallons, and prices already read. Then the checking can start: each load on the invoice compared against the rack or delivered price it should have carried, so a wrong price is flagged before the invoice is approved, not discovered after it is paid.
EFT and credit card batches. The third stream is the card money: the batch data behind station card sales and settlements. Pulled automatically, batches arrive as incoming batches, staged per store for the same one-click review — accept, modify, or reject — then move into reconciliation, matched against bank deposits with gaps flagged, instead of a report somebody downloads when they get around to it and keys into a spreadsheet.
Three different jobs in the office — pricing, payables, card reconciliation — all fed from the same access you already have.
The setup happens once. The daily operation needs nobody.
1. Bring the access you already have. If your company gets its data through a service like DTN, the service issues your company its own FTP credentials for the file drop — the same credentials a person would use to log in and download files by hand. Software can do that logging in instead, using your access, on your behalf. In FuelJobberX, credentials you provide are stored in a secure credential vault, not in the program and not on a sticky note, and each company's connection is its own: one jobber's prices and invoices never mix with another's. And if your suppliers reach you by email instead of by FTP, the same idea applies to your inbox — the files are read where they already land.
2. Map the files once. Feed files are generic; your business is not. The mapping step teaches the system your slice of each file: which suppliers and terminals are yours, how their product codes match your products, and how invoice fields line up with your loads. Prices and invoices each get their mapping, set up once and adjusted only when a supplier changes their format.
3. Poll every 30 minutes, around the clock. Prices do not wait for office hours, and neither does the pulling. Your file drop is checked every 30 minutes; when a new file has landed, it is downloaded and held complete — a file caught mid-upload is not parsed half-way — and then staged for processing.
4. Parse into the right desk, not into a folder. A downloaded file that sits in a folder has only moved the data entry, not removed it. So each feed parses into the place its numbers belong: prices into the price book, supplier invoices into staged invoices ready for price checks, card batches into the reconciliation queue. If a file ever needs a second look, it can be re-parsed from the stored original — the source file is kept, not consumed.
5. Work the exceptions, not the files. Automation should not mean blind trust. What a human needs to see is the unusual: a price that moved far more than the market, a supplier invoice line that does not match the price it should carry, a card batch with no matching deposit, a feed that did not update when it should have. Flag those, let the routine flow through, and a morning of downloads becomes a two-minute review.
The morning opens differently. The price sheet is built from prices that arrived overnight, not from a file somebody downloads with their first coffee. A 2 p.m. price move is in the system by 2:30.
Supplier invoices check themselves. The invoice is verified against the same price data the feed already delivered — the two streams cross-check each other. Overcharges and wrong-price lines surface at approval time, while the load and the BOL are still in front of you.
Card money stops waiting on a person. Batches are in the system the day they are issued, so reconciliation runs on current data and the available CC net is a number from this week, not from whenever the last download happened.
Every number has a history. Each pull is stored with its file and its effective time. "What was the rack on the 14th?" and "what did the supplier actually invoice on that load?" are lookups, not arguments — and the source file behind any number can be re-opened and re-parsed.
Errors move to where they are cheap. A transposed digit happens at a keyboard, not in a parser. The mistakes that remain are the interesting ones — a supplier's wrong price, a missing batch — which is exactly the work an office should be spending its attention on.
Does one access really carry all three? Yes. Prices, supplier invoices, and EFT/credit-card data are separate files, but when a company gets its supplier data through a service like DTN they come through the same file drop, so one credential set and one polling routine serve all three. Each file type is configured and mapped separately, because each lands in a different part of the office.
What if my suppliers don't come through a data service? FTP is the main road, not the only one. If you get your prices, invoices, and batches by email instead, the same system reads them from your inbox: supplier invoices arrive as emailed PDFs, price sheets as emailed files, card reports the same way — each parsed by format, the same discipline applied to a different doorway. A supplier who only sends paper or uses a portal login with no file at all stays manual, and in most offices that list is short and shrinking.
What if a file is incomplete or a feed goes quiet? A file caught mid-transfer is held, not parsed — partial data never posts. And a feed that stops updating is itself an exception worth flagging: you keep working from the last confirmed data, and you know that is what you are doing.
Who can see the feed credentials? Nobody, in daily use. They live in a secure vault and are used by the polling routine; the office works with the prices, invoices, and batches that come out the other side. A connection test confirms the login works without anyone reading the password back.
Does this replace checking supplier invoices? It replaces re-typing them. The checking gets better, not smaller: every line arrives already compared against the price it should carry, so the human review starts at the discrepancies instead of at a blank screen.
None of this is exotic — suppliers have published prices, invoices, and card data as files for years, and most jobbers already have access to them in one form or another. The waste was never in the data; it was in the hand-off, offices re-typing digital numbers off a screen as if they had arrived by fax. Bring the access you have, let the files be picked up every half hour, parse each stream into the desk it belongs to, and the office's mornings go back to the work that actually needs a person.
FuelJobberX works with the access you already have. If you get your daily prices, invoices, and card batches from your suppliers through a service like DTN, FuelJobberX can pick up all the files from your FTP drop — polled every 30 minutes — and input the data: rack prices into the price book, supplier invoices staged and price-checked before approval, card batches straight into reconciliation. If your suppliers send them by email instead, it reads them from your inbox the same way. If you want to see what your office looks like with the download-and-type step removed, you can book a demo at fueljobberx.com — a walkthrough with real data in under 30 minutes, no sales pressure.
We'll walk you through the full platform in under 30 minutes — with real data, no sales pressure.
Book a Demo →