Walk into the warehouse of a bolt, nut or screw manufacturer and you will usually see the same scene: shelves labelled in different handwriting, stock cards split between a notebook and loose paper, and the production schedule living entirely inside the foreman’s head. At a mid-sized fastener plant in Bursa’s organised industrial zone, that was exactly the situation I found. The factory employed 312 people and produced thousands of tonnes of screws and studs every year, yet nobody could say with certainty how much raw material a given production batch had consumed. The owner told me plainly: ‘We need a system. Let us get on the software and sort it out.’ The intention was right. The order was wrong.ERP — which stands for enterprise resource planning — is a family of software that tries to bring all factory operations into one shared system: purchasing, production, inventory, sales and accounting. You can think of it as moving your paper notebooks, Excel sheets and telephone notes onto a single screen. But that comparison falls short. Because when you move your notebook, you also move the mess inside it. That is why the first job before any ERP installation is to lay out your processes on the table — not to open the software.Fastener manufacturing looks straightforward from the outside. Steel rod arrives, wire is drawn, press strikes, screw comes out. In practice, the material movement between production steps, several semi-finished stages and stock that splits by grade — stainless, high-tensile, standard carbon — makes the operation genuinely complex. At the Bursa plant, the M6, M8, M10 and M12 screw series each required different raw material specifications. Whether a customer order was fulfilled with the correct grade depended on the warehouse worker’s memory. Before any software is touched, the material coding structure must be settled. How many distinct product types exist? By what criteria does each one’s raw material differ? These questions must be written down on paper before the program is opened.The second major problem is production orders. In most small and mid-sized factories, a production order starts when the foreman says so or when a customer pushes hard enough. ERP wants to formalise this chain: an order arrives, the system checks whether raw material is in stock, if not it triggers a purchase request, if yes it opens a production order, work begins, work ends, stock is updated. On paper this is logical. In practice the chain breaks halfway through. At the Bursa factory I found that production would begin before the raw material was formally received into the warehouse. Whatever the software said, the warehouse reality was different. Building an installation on top of that is like raising a tall building on a shallow foundation. Before the system goes live, it must be agreed who signs the goods receipt note and when — and that agreement must be enforced, not just written down.The third area is connecting accounting to production. This is one of ERP’s genuine strengths: when a screw is produced, the raw material cost can flow automatically into the accounts. But for that link to work, every product must have a defined cost structure. Which screw is made from which steel grade? How is labour cost distributed across product lines? These questions require the accountant and the production manager to sit down together. In most factories these two departments do not speak the same language. The accountant knows line items; the production manager knows the machines. An ERP project forces these two worlds to meet. Skip that meeting and within a few months you will find the accountant running a private Excel file and the foreman keeping his own notebook — while the expensive software sits unused in the middle.So what does a factory manager do on Monday morning? First, produce a clean product list. How many distinct finished goods does the factory make? Each needs a code, a name and a raw material composition. If that list already exists, it goes into the system. If not, the list is built first — the system waits. Second, create clean supplier and customer records. The software must know who you sell to and who you buy from. If the same supplier appears under three different names in the old records, the system will generate confusion immediately. Third, carry out a physical stock count before the system goes live. Enter those counted figures as the opening balances. This step is the one most often skipped — ‘counting takes too long’ — and it is the reason so many installations start with wrong inventory data from day one. At the Bursa factory the count took two weeks. But after those two weeks the stock figure reflected reality for the first time anyone could remember.Starting with ERP in fastener manufacturing does not require a large budget. Turkish-made production tracking software with integrated accounting is available and suitable for factories at this scale. The real difficulty is not the licence cost. It is the patience to spend the first six weeks of the project defining processes clearly. During that period, three rules must be agreed and enforced by everyone: no production begins before the order is entered into the system; no raw material is deducted from stock before the goods receipt is signed; the accountant closes the private Excel file and treats the figure in the program as the authoritative number. These rules sound trivial. They are not. Embedding all three permanently inside one factory is worth more than unlocking every advanced feature the software has to offer.
This article was originally published in Turkish by Gökhan MERCANOĞLU on January 8, 2002. The English edition has been reviewed and edited by the author.