How to run a manufacturing business on Odoo.

Here is how a factory in Dubai, Karachi, or anywhere between loses money while every line looks busy: the bill of materials in the system says a unit takes four kilograms of raw material and forty minutes of machine time. The floor actually uses four point six kilograms and fifty-five minutes, because that is what the process has quietly become since anyone last checked it — a supplier substitution here, a rework step there, an operator's own shortcut nobody wrote down. Whether it is food, plastics, metal, furniture, or an assembly line, the pattern repeats: nobody owns updating the BOM, so the gap between what the system says and what the floor actually consumes just sits there, quietly wrong, feeding every cost calculation built on top of it.

The rest of the losses follow the same pattern. Raw material gets bought on gut feel — a purchasing officer's sense of what usually runs low, not a number tied to actual consumption or a real sales forecast. Work in progress is invisible: half-finished goods sitting on the shop floor represent real money, and no report shows where it is or how long it has been sitting there. Scrap gets swept up and forgotten, never logged against the job that produced it, so nobody sees which product or which machine is quietly eating margin. The price quoted to a customer is still built on last year's cost sheet, because nobody has recalculated it since raw material prices moved. And when a machine breaks down with no warning, the whole line stops, and the only maintenance record that exists is whatever a technician remembers doing last time. None of this looks like one dramatic failure. It looks like a factory working hard and still unable to say what anything it makes actually costs.

What standard Odoo gives you — and where it needs shaping

Odoo Manufacturing is genuinely deep out of the box. Multi-level bills of materials, routings across several work centres, work orders, and both standard and actual costing all come natively, and for plenty of ERPs this is the module sold separately as an expensive add-on rather than shipped as the core. What standard Odoo will not do on its own is keep any of it honest. It will not stop a bill of materials drifting away from what the floor actually consumes once a supplier substitution or a process tweak quietly becomes the new normal — that takes a review someone actually owns. It will not choose between standard costing and actual costing for you; that is a decision about how your business wants variance to show up, and the wrong choice buries the truth instead of revealing it. And it will not put a barcode scanner in an operator's hand or make a technician log a stoppage reason instead of shrugging — adoption on the shop floor is implementation work, not software. That shaping — BOM discipline, the right costing method, and a floor that actually uses the system — is exactly the layer we build around every manufacturing business we work with across the UAE, GCC, and Pakistan.

The workflow: from sales order to finished goods

• 1. Demand triggers MRP, not a gut feeling. A confirmed sales order, a sales forecast, or a reorder rule runs through Material Requirements Planning, and Odoo works out what has to be built, bought, and by when — not a planner guessing from what looks low on the shelf.
• 2. Purchase suggestions carry real supplier lead times. Raw material shortfalls become purchase suggestions automatically, timed against each supplier's actual lead time, so an order lands before the line needs it, not after a machine has already gone idle waiting on stock.
• 3. Work orders route through the right work centre. Each manufacturing order splits into work orders that follow the product's routing — cutting, moulding, assembly, finishing, whichever sequence is real — scheduled against the capacity of the work centre that actually does the job, not a generic calendar.
• 4. The floor runs off a barcode terminal, not paper. Operators start, pause, and close work orders from a barcode-driven shop floor terminal at the machine, logging scrap and downtime reasons as they happen, so a stoppage is a timestamped record, not something reconstructed from memory at the end of the shift.
• 5. Material consumption posts against the BOM — and variances get flagged. What the floor actually pulls from stock is recorded against the bill of materials for that order, and when actual consumption diverges from what the BOM specifies, the variance is visible immediately, not discovered when someone finally audits the BOM a year later.
• 6. Quality checks sit at the points that matter, not just at the end. Inspection points are configured into the routing itself — incoming raw material, mid-process, or final inspection — so a defect is caught and tied to the batch and the work order that produced it, not found by a customer three weeks later.
• 7. True cost per unit rolls up on its own. Material cost, labour time costed by work centre, and overhead all attach to the manufacturing order as it completes, so the cost of a finished unit is what it actually took to build — materials plus labour plus overhead — not a number carried forward from last year's estimate.
• 8. Finished goods carry lot or serial traceability from day one. Every finished unit is traceable back to the batch of raw material and the work orders that made it, so a recall, a warranty claim, or a customer query is a lookup, not a forensic exercise through old paperwork.

What management finally sees

Because every material issue, work order, and quality check sits on one system, true margin per product stops being an assumption made at quoting stage and becomes a number pulled from what the floor actually consumed. Work in progress has a value management can actually see — the money sitting half-finished on the shop floor is a figure on a report, not an unknown. Scrap rates surface by product and by work centre, so a quietly expensive process gets found before it has eaten a year's margin, not after. Work-centre load shows where capacity is tight and where a machine sits idle, and unplanned downtime shrinks once a stoppage automatically raises a maintenance request against the work centre that caused it. On-time delivery becomes a tracked number instead of a promise repeated every time a customer calls asking where their order is. And with our AI Connector, a factory manager can just ask: "which products cost more to make than we quoted?" — and get the answer from live production data, not a costing exercise somebody has to build specially to answer it.

How we implement it

We start with your bills of materials, your routings, and however costing is done today — usually a spreadsheet built by someone who has since moved on, and that is fine, it is exactly what we are here to replace. We phase the rollout deliberately: inventory and purchasing go live first, so stock and procurement are accurate before production leans on them; work orders, routings, and the shop floor terminal follow once that foundation is solid; costing and reporting close out the project once real consumption data exists to cost against. Most manufacturers go live in eight to twelve weeks. And it comes with our guarantee: you see the system running on your own bills of materials and your own products before you pay, and if we miss the date we agreed, you walk away.