Installing Navision in a factory does not automatically standardize production. This sounds obvious, but most companies act as if it will. ‘We bought the software so it can bring order’ — this reasoning shows up in almost every troubled project. At a mid-sized machine parts manufacturer in Eskisehir — 295 employees, supplying components mostly to the automotive supply chain — this exact assumption caused the project to fall six months behind schedule. The software was installed, licenses were paid for, training sessions were held. Yet production reports were still being written by hand and placed on the plant manager’s desk each morning. The software was running. The process was not. That distinction is what this article is about.Navision is an ERP (Enterprise Resource Planning) program that Microsoft acquired from a Danish software firm. ERP connects a company’s accounting, inventory, purchasing, production, and sales into a single system. Think of a small grocery shop where the cashier keeps the register and the stock ledger separately — double work, constant mismatches. A factory with each department keeping its own records has the same problem at a larger scale. ERP solves this. But only if every person enters the same information into the same system. That ‘only if’ is where most implementations in Turkey get stuck in 2002.The economic backdrop matters here. Turkey is recovering from the 2001 financial crisis. Many mid-sized manufacturers are cautious about spending. Those who export have a currency advantage, but domestic demand is still flat. Against this background, a decision to ‘fix the system’ requires both courage and budget. Navision reaches Turkish customers through authorized local resellers, and the combination of implementation fees and annual support costs is a significant line item for a factory this size. When money is spent under pressure, the expectation is immediate results. That pressure is precisely what pushes companies to skip the preparation work and go straight to installation. It rarely ends well.The preparation work is not complicated. It is just unglamorous. Before anyone opens a laptop, the following questions need clear written answers: Who creates a production order? In what format, on what form? Who tells the warehouse to release materials — by phone, by fax, by written slip? When a defective part comes off the line, where does that information go? If the answer to any of these is ‘the foreman knows’ or ‘we handle it case by case,’ the software has nothing to work with. Navision speeds up a process that already exists. It cannot create one. In Eskisehir, the real problem was this: production orders moved verbally from one shift supervisor to the next, the warehouse kept stock cards that were rarely updated, and accounting processed invoices in monthly batches with no real-time connection to the factory floor. None of this could be entered into Navision because none of it was defined anywhere, not even on paper.The fix was methodical and low-tech. When the project restarted, the first four weeks involved no computers at all. Meetings were held with each production supervisor. A single question drove every session: ‘Walk me through exactly what happens from the moment a work order is opened.’ The answers were written down step by step. A manual stock count was done; discrepancies were recorded. The purchasing process was mapped on paper: who requests, who approves, who sends the fax to the supplier, who signs for the delivery. Once every step was on paper, the Navision screens made immediate sense. Each screen was asking for exactly the information that the paper process already defined. Implementation took five weeks after that. The previous eight months of attempted installation had produced nothing usable. The software had not changed; the clarity of the underlying process had.There is a limit worth stating plainly. Navision is not the right tool for every manufacturing situation. A small workshop — say 45 to 55 people, a single product family — will find Navision overly complex and expensive. A simpler accounting and stock program will serve well enough. At the other end, a very large operation will outgrow Navision quickly and will be looking at systems like SAP. Navision sits in the middle: it works well when multiple departments must share data, when purchasing and accounting cannot be allowed to record the same invoice differently, when production and inventory have drifted apart. That situation typically starts to appear somewhere between 250 and 400 employees. Below that, verbal coordination still holds. Above it, heavier platforms come into play. Matching the tool to the actual scale is half the decision.Three months after the Eskisehir project was completed, the production manager described his mornings differently. ‘I used to call four supervisors before I understood what happened yesterday. Now I sit down, open one screen, and I know.’ That is not a statistic. It is a shift in how a working day begins. Behind that shift, however, the numbers did matter: the same part had been entering the stock system under three different codes — that stopped. Month-end closing dropped from three days to one. These gains did not come from the software. They came from the orderly process that the software made visible. Before starting any production standardization project, ask yourself one question: when a work order is opened in my factory, does its trail exist anywhere — on paper, in a form, in a shared understanding that survives a shift change? If the honest answer is ‘not really,’ then spending one week drawing that trail is worth more than any software budget decision you will make this year.
This article was originally published in Turkish by Gökhan MERCANOĞLU on January 17, 2002. The English edition has been reviewed and edited by the author.