Every garments ERP demo looks the same. There is a dashboard with a production chart, a few screens of master data, and a costing report that fills in instantly because the demo database was seeded for it. Nobody in the room has yet asked the question that decides whether the system will survive contact with your floor: when does a consumption entry get made, and by whom?
Ask that first. The answer tells you more than an afternoon of screens.
What actually breaks
In most factories we see, style-wise costing is not wrong because the formula is wrong. It is wrong because the numbers arrive too late to be checked.
The order is booked. The bill of materials exists, usually in a spreadsheet held by one person in merchandising. Fabric and trims are issued against a paper requisition. The floor writes consumption in a register because stopping to enter it would stop the line. A week later, sometimes two, a data entry operator keys in a page of handwriting.
By then three things have happened. The stock figure has been wrong for a week, and every reorder decision taken in that week was taken against it. The gap between what was issued and what the register says was used has become "wastage", because there is no longer any way to find out what it actually was. And the costing sheet for the order is finished after the shipment has left, which means it is a record of a decision rather than an input to one.
The next order is already cut by then, against the same assumptions.
Why demos avoid it
A demo shows the costing report. The costing report is the easy part — it is arithmetic over entries that already exist. What a demo rarely shows is the screen the storekeeper uses at eleven at night when a line is waiting, on a machine with a cracked display, on a connection that drops when it rains.
That screen is the whole system. If issuing material against a production order takes longer than writing it in the register, the register wins, and you have bought a reporting tool that runs a week behind your factory.
So when a vendor opens with the dashboard, ask to see the issue screen instead. Ask how many taps it takes to issue five items against an order. Ask what happens to that entry when the connection drops halfway through.
The questions worth asking
Where does the bill of materials live, and who can change it? If it lives in merchandising's spreadsheet and is copied into the system per order, you will have two versions within a month. If it lives in the system and merchandising works from it, the costing has a chance.
Can the floor record against a production order, or only against a date? Recording against a date gives you a daily consumption total. It does not give you cost per style, which is the number you actually need, and no amount of reporting on top will recover it.
How does the system handle a batch that is split across two lines? Real production splits, merges and reruns. A system that assumes one order maps to one continuous run will be worked around within a week, and the workaround will be a spreadsheet.
What happens to wastage that is recovered and reused? Cut waste that goes back into a smaller component is normal in this industry. If the system can only write it off, your material cost will be systematically overstated and nobody will trust the sheet.
Can it produce the Mushak 6.3 challan from the same entries? If VAT documentation is a separate exercise in a different system, you have not reduced the paperwork. You have added a second place for it to disagree.
Who answers the phone in month four, and in what language? The floor supervisor who needs help at the end of a shift is not going to raise a ticket in English and wait two days. This is not a soft question — it decides whether the system is used after the consultants leave.
What a first release should contain
Do not start with everything. A factory-wide rollout that begins with master data for every module ends with six months of configuration and nothing running.
Start with the chain that produces the number you cannot currently trust: production order, material issue, consumption and wastage against that order. That is a small system. It can be live in weeks rather than quarters, and it produces a costing sheet that is available while the order is still on the floor.
Procurement, payroll and the ledger attach to it afterwards, in the order your operation needs them. Each one is easier to add once the floor is already recording as it works, because the hard part — changing what people do at the moment they do it — has already been done.
The honest version of the timeline question
A vendor who answers "three months" before seeing your floor is quoting a standard project length, not your project. The things that actually decide it are how clean your existing records are, how many systems the new one has to talk to, and how many people have to change what they do on day one.
Ask for a scope that is cut to a date you choose — the audit, the buyer visit, the start of the season — rather than a date that comes back from a template. A shorter deadline should mean fewer modules in the first release, not a thinner version of each.
If you want to work through what that first release would contain for your factory, tell us what breaks first and we will go through it with you.