When I opened the product list at a mid-sized plastic pipe and profile manufacturer in Adana last year, the first number that stopped me was 1,340 — the count of active product codes in a plant of 284 people. When I asked how this had happened, the answer was simple: ‘Customers asked, we opened them.’ Digging deeper, roughly one third of those codes had seen zero shipments in the previous two years. A significant portion had split a single-dimension item into two or three separate codes purely by colour variation. What the plant manager called a ‘stock problem’ was, in fact, a coding discipline and decision-making problem wearing a stock problem’s clothes.In plastic pipe, profile and fittings production, variant proliferation can feel like a natural law. Diameter options, wall thickness combinations, colour requests and raw material recipe variables mean that almost every customer order carries the potential to generate a new product code. The question that lands on the production planner’s desk is always the same: do we open a new code, or do we load this order onto an existing one? In most plants there is no written criterion for this decision, so each planner answers it according to habit. Over time the catalogue grows, stock monitoring becomes difficult, and the MRP system struggles to read real demand through a fog of near-identical codes.My core argument — and it is a deliberately contentious one — is this: cutting the number of variants on its own delivers nothing lasting. The complexity is not produced by how many variants exist; it is produced by the absence of logic governing when and why a variant is created. A plant can reduce its product codes from 400 to 200, but if sales, production and management have not agreed on the criteria behind that reduction, new codes will start accumulating within six months and the same point will be reached again. Sustainable simplification means changing the decision mechanism, not the catalogue.In the Adana project, the first practical step was to pull eighteen months of shipment data across all active codes from the sales team. The analysis showed that roughly 63 percent of total revenue flowed through just 180 product codes. The remaining 1,160 codes generated modest volumes, typically a few small shipments a year. The second step was the colour audit: the plant was maintaining 34 colour codes for its pipe lines. Twelve of those had been created not because any customer standard required them, but because someone had decided internally to distinguish them at some earlier point. When we calculated the dedicated storage space, the separate production instruction sheets and the additional raw material lot separation those 12 codes demanded, the plant manager’s expression shifted. A visible stock cost had been hiding a far less visible operational burden.The most concrete technical obstacle when we moved this picture into the MRP configuration was a unit-of-measure conflict. A portion of the existing product codes was defined by weight, another portion by length and a third by piece count. When unit definitions are inconsistent, recipe calculations and stock conversion factors collide, and planning figures lose meaning. This type of unit confusion is, in my direct experience, a recurring issue specific to the plastics sector: the diameter-to-wall-thickness combination is meaningful on a weight-based recipe, the customer is invoiced per metre, and the warehouse counts pieces. Who enters the conversion factor between these three measures, who updates it when a specification changes, and who checks it when a shipment discrepancy appears — if these questions have no clear owner, the MRP system will generate unreliable output regardless of how many codes you remove. Reducing variant count before fixing this is rearranging furniture in a room with a structural crack.What can actually be done? During the first eight weeks of the Adana project, three steps were implemented. First, the authority to create a new product code was tied to a single approval checkpoint: a request originating in sales could not generate a system record until production planning confirmed it. Second, a standard matrix was built for colour and diameter combinations; any request outside that matrix was accepted as a ‘special order’ category, carrying a different price and a longer lead time. Third, a simple annual review process was defined: any code with no shipment in twelve months would be frozen, and after eighteen months it would enter the deletion queue. None of these steps required a technology investment. They only required writing down who decides what, when, and by what rule.The hardest obstacle in this journey is not technical but cultural. Sales resists with ‘the customer asked for it, we will lose the order.’ Production planning asks ‘opening a special code is easy, why add a gate?’ What breaks that resistance is not an analytical presentation but a concrete calculation: show how each additional colour code raises the average value of raw material sitting immobile in stock, and the tone of the conversation changes quickly. Fourteen months after this work was completed in Adana, active product codes had fallen from 1,340 to 690, and stock turnover had improved in a measurable, auditable way. Simplifying the logic behind variant creation produces exactly this kind of tangible business result — and it does so without requiring anyone to sacrifice a genuine customer need.
This article was originally published in Turkish by Gökhan MERCANOĞLU on January 29, 2005. The English edition has been reviewed and edited by the author.