Walk into a project meeting at a seed company in Konya and you will find three people trying to solve three completely different problems at the same table. The accounting manager wants the tax books kept clean. The sales manager wants to know which customer bought which variety, in what quantity and from which lot. The general manager wants to know whether the report going to the overseas partner should be in Turkish lira or US dollars. All three answers are supposed to live inside the same program. Yet the project team usually discovers this too late. That late discovery is what this article is about.Seed trading in Turkey has its own logic. Mid-sized companies in provinces like Konya, Eskisehir and Antalya sell to local farmers and to foreign buyers at the same time. A warehouse at a firm with 280 employees might hold wheat seed, sunflower seed and several vegetable varieties side by side. Each one has a different shelf life, different storage requirements and a different price basis. On top of that, licensed variety tracking matters: which seed belongs to whose registered variety, whether a licence fee has been paid. None of this fits in a paper ledger. Standard accounting software does not fully solve it either. This is where ERP (enterprise resource planning software) becomes relevant.ERP means a program that connects multiple departments — stock, sales, purchasing, accounting — so they all see the same data. When a sale is made, inventory drops automatically. When a payment arrives, an accounting entry is generated. On paper everything is linked. But in the seed industry there is a serious obstacle in front of that picture: the season. Seed sales do not move like supermarket sales. Orders concentrate in a few months then go quiet. That rhythm creates waves in stock levels, cash flow and staffing. If you do not teach the program to understand seasonal logic, the system starts giving wrong warnings. The warehouse looks full but a customer order cannot be fulfilled. Or the opposite: stock shows zero but the physical inventory holds product reserved for the next season. I remember a Konya firm where the team spent 14 days fighting the program before realising the problem was not in the software at all — it was in how data was being entered.International reporting adds another layer. A seed company selling to buyers in the Netherlands or Germany must satisfy both Turkish accounting rules and the documentation requirements of the foreign buyer. Turkish accounting has always been tax-driven: record the transaction, file the declaration, done. A European buyer asks different questions. Which lot did this seed come from? Do you have a germination test certificate? What is the certification number? None of that fits naturally into a standard Turkish accounting entry. You can force it in, but without a systematic structure it just becomes a text note that nobody reads. ERP can handle this — but only if it is set up correctly. You need to link product information to the accounting record, carry the lot number through to the invoice and be able to pull a separate foreign-currency sales report. These are defined in separate modules, meaning separate sections of the program. Which module you open first determines whether the project succeeds or gets rebuilt three months in.Here is the central argument of this article, stated plainly. A seed company should not start an ERP project from the accounting module. Most of them do, because the accountant is the loudest voice in the room. But in seed trading the real complexity sits in commercial tracking: which customer took what, from which lot, at which price, on what delivery date, was the licence fee included, which exchange rate applied? When those questions are answered cleanly, the accounting follows naturally. When you start from the accounting side instead, the ledgers look tidy but the sales team is still writing things down on paper. I call this digitally tidy on the surface, manual underneath. Teams that skip this argument at the start of the project end up rebuilding the system after roughly six weeks of frustration.So how should a seed company actually approach this kind of project? There are five practical steps worth considering. First, separate the requirements list into three groups: commercial tracking needs, accounting compliance needs and reporting needs for overseas buyers. All three matter, but they do not all need to be solved on the same day. Second, spend one week watching how the sales team actually works. Where does an order arrive? Who writes it down and where? Usually it lands on a personal notepad or a shared paper form. The question is where that piece of paper will live inside the new program. Third, be honest about user frequency: how many times a day will the warehouse clerk open this program? If the answer is ‘we are not sure yet’, the project is not ready to go live. A system that real users do not touch every day empties out within three months. Fourth, settle the currency question before anything else is configured. If you sell abroad in dollars or euros, decide how exchange rates will be updated — manually entered, taken from the bank, and at what frequency. This looks like a small detail. It is the most common reason why monthly reports from the sales side and the accounting side do not match each other. Fifth, run a season simulation before going live. What happens to the system when orders pile up in the busy months? Where does it slow down or produce errors?Seed companies look modest in scale. In terms of data complexity and operational variety they are not. A mid-sized firm in Konya processes thousands of transactions a year, manages perhaps twenty different crop varieties, exports to several countries and tracks licences that carry legal and financial obligations. When the right program is configured correctly, tax compliance, commercial tracking and international reporting all come from the same screen. When it is configured the wrong way, the accounting looks clean and the sales team is still sending faxes. The difference between those two outcomes is not technical. It is whether the right questions were asked at the very beginning of the project. Skip those questions and years later someone in that office will still tell you: ‘we keep that part in a separate spreadsheet.’
This article was originally published in Turkish by Gökhan MERCANOĞLU on January 2, 2003. The English edition has been reviewed and edited by the author.