The document nobody reads
Every manufacturer has a version of this moment. A purchase order goes out on Monday. A supplier replies with a confirmation. The reply lands in someone's inbox, gets a quick glance or none at all, and is filed under the assumption that it says what the order said. Three weeks later a pallet arrives short, late, or priced differently, and the production schedule that depended on it is already in trouble.
The purchase order gets all the attention because you wrote it. The delivery gets attention because it is physically in your building. The order confirmation sits between them, and it is the one document where the supplier tells you in writing what they will actually do. It is also the one most often ignored.
This post is about that gap: what lives in it, why it persists even in careful shops, and what it takes to close it.
Four documents, four different owners
A single material purchase usually produces a chain of paperwork. Each link answers a different question, and in many shops each is held by a different person or system.
- Bill of materials (BOM): what the job needs. Usually owned by engineering or production planning.
- Purchase order (PO): what you asked the supplier to deliver, in what quantity, by when. Usually owned by purchasing.
- Order confirmation (OC): what the supplier agreed to deliver. It often arrives as an emailed PDF to whoever placed the order.
- Bill of lading (BOL): what the carrier actually picked up and moved. It shows up at the dock, often on paper, and is handled by receiving.
Four documents about one purchase, with at least three handoffs. The BOM-to-PO step gets scrutiny because it starts the process. The BOL-to-receiving step gets scrutiny because someone has to sign for freight. The PO-to-OC step, where a promise is made or quietly altered, tends to fall between the two.
Where the drift hides
Consider a hypothetical. You order a quantity of sheet material for a job with a firm ship date. The supplier's confirmation comes back with a ship date a week later than you asked, because one item is on backorder. Maybe the quantity is split across two shipments. Maybe a unit price changed. None of this is dishonest. Suppliers are doing their best with their own constraints, and they are telling you in the confirmation.
The problem is that nobody compared the confirmation against the order. The planner still believes the material arrives on the original date. The scheduler built the job start around it. The sales team promised the customer a delivery based on that schedule. The change is sitting in a PDF, and the people who needed to react to it never saw it.
That is why the drift is expensive. It is not a single mistake. It is a piece of information that never traveled from the person who received it to the people who depend on it. The same pattern shows up in other forms:
- A partial shipment is received and treated as complete because the dock only checked the first pallet.
- A substitution is accepted by one person and never recorded against the BOM.
- A lead time that was true last year is still assumed in every new order, even though recent confirmations have said otherwise.
Each one is small. Together they are why production schedules built on the best intentions keep meeting surprises.
A discipline that works before any software
You do not need a new system to start closing this gap. You need a habit and a clear owner. A simple version:
- Treat the confirmation as a required step, not a courtesy. A PO is not finished until a confirmation has been received and read.
- Compare three things every time: quantity, date, and price against the original PO. Anything that differs is an exception, not a footnote.
- Route exceptions to the people they affect. A moved date goes to scheduling. A moved price goes to whoever quotes or invoices the job. A short quantity goes to whoever owns the BOM.
- Check the delivery against the confirmation, not just the PO. The confirmation is the supplier's commitment, so it is the better benchmark for whether they performed.
- Keep a record of how often confirmations differ from orders. Over time, that tells you which suppliers' stated lead times you can believe.
This works on a spreadsheet and a shared inbox. It also shows why the process is fragile: it depends on people remembering to compare documents that live in different places. The more jobs you run, the more likely a step gets skipped.
Where Jelly Flow fits
Jelly Flow, the enterprise production tracking side of JellyMachine, includes a procurement flow built around exactly this chain: BOM to PO to OC to BOL. The point is not to automate your suppliers or to promise that nothing will ever slip. The point is that the documents describing one purchase are connected in one place, tied back to the production work that needs the material, instead of scattered across inboxes, folders, and the dock clipboard.
That connection matters for the gap described above. When the confirmation lives next to the PO it answers, comparing them becomes a normal part of the process instead of an act of memory. When the BOL belongs to the same chain, receiving can check what showed up against what was promised. And because procurement sits alongside the rest of Jelly Flow, including inventory and shop-floor station phases, the people planning production and the people buying material are looking at the same picture.
It does not remove the need for judgment. Someone still has to decide whether a one-week slip is acceptable or whether to call another supplier. What it changes is whether that person finds out in time to decide.
Start with one question
If you want to know how exposed your shop is, pull your last handful of purchase orders and ask: for each one, can I find the supplier's confirmation in under a minute, and did anyone compare it to the order? If the answer is no or I am not sure, you have found the gap.
Production problems that look like late deliveries are often earlier than that. They begin the moment a changed promise goes unread. Close that step, whether by habit, by a clearer owner, or by keeping the whole procurement chain in one system, and the rest of the schedule gets a lot more honest.