Merchant Center Feed Engineering for German Retail

In a German catalogue we take over, somewhere between eight and twenty percent of items are either disapproved or approved with limitations and effectively unbuyable. Two causes account for most of it in Germany and Austria, and neither is a Google problem: missing base-price data under the Preisangabenverordnung, and missing fibre composition under the Textilkennzeichnungsverordnung. German feed law, expressed in the feed: PAngV requires a Grundpreis for anything sold by weight, volume, length or area, which covers most of food and drink, most of cosmetics, and every supplement. In the feed that means unit_pricing_measure and unit_pricing_base_measure, and getting the base unit wrong, 100 g where the category convention is 1 kg, is both a disapproval risk and a competitive problem, because Idealo and Geizhals read the same fields and will rank you against merchants who got it right. Textile items need fibre composition, and German shoppers expect the material named in the title itself. We map these from your PIM or ERP where the data exists, and we tell you plainly which fields exist nowhere and have to be created by a human. Identifiers: EAN-13 on branded resale items. Correct use of identifier_exists false for genuine own-brand goods, rather than as a blanket switch flipped once to silence a wall of errors. MPN discipline across variants. Getting this wrong pulls items out of comparison surfaces silently, with no error message to read and no obvious drop in the campaign report. Titles written for how Germans search: Brand, product noun, defining attribute, then size and colour, front-loaded inside the first seventy characters because that is what a shopper actually sees. German compounds matter here: a title carrying only Damenjacke will not match somebody typing Jacke Damen as cleanly as one carrying both forms, and there is room for both inside the limit. We rewrite titles by template against your attribute data rather than by hand, so the rule survives the next catalogue import instead of being undone by it. Where a rule is allowed to live: Anything a merchandiser will need to change monthly belongs in the source feed. Anything structural and permanent can be a feed rule or a supplemental feed. We document every rule we create, because the most expensive object in Merchant Center is a transformation nobody remembers exists, three staff changes later, silently overwriting a price. The campaign side: Priority-tier Shopping structures died with Smart Shopping and we do not rebuild them. We split by margin band into separate campaigns with separate targets, keep a lower-target catch-all for the long tail, and use the price competitiveness and best sellers reports to decide which SKUs deserve a target at all. In German categories where comparison sites hold the top of the SERP, the honest answer for some SKUs is that you cannot win them at your price. We would rather say that in week two than spend the budget proving it over a quarter. Not included: PIM or ERP implementation. Product photography. Writing German product descriptions across the catalogue, though we will name the SKUs where the description is costing you the click. Amazon, Idealo or Otto feeds; adjacent work, different specifications, and we do not pretend one file serves them all. Who this is not for: Retailers with no SKU-level cost or margin data, because every target we set would be a guess dressed as strategy. Catalogues maintained by exporting a spreadsheet by hand each week; we will help you get off that, but we will not work around it. And drop-shippers listing the same catalogue as forty other merchants, where feed quality is not your constraint and price is.