Idealo, Kelkoo and PriceRunner feeds built on margin

Start with one division: Contribution margin per order divided by the category cost per click gives the conversion rate the channel has to beat. Comparison clicks in mainstream consumer categories run from around 20 cents to 1.20 euro depending on category and engine, and higher in industrial parts. If that division says the feed must convert at 4 percent while your site converts comparison traffic at 1.6, no amount of feed work rescues it, and I will tell you in week one rather than in month six. Roughly one engagement in five ends there, with a shorter invoice. The feed is the advertisement: On these engines you have no landing page, no creative and no ad text. The row in the file is the whole ad, competing on a results page sorted by total price including delivery. So the fields that decide the outcome are the ones most feeds treat as an afterthought: delivery cost and delivery time as structured values rather than a footnote, stock state that updates faster than the engine recrawls, and a price that matches what the shopper sees at checkout. A mismatch does not cost you a sale, it costs you the listing. Matching, which is where most catalogues fail: These engines attach your row to an existing product node by GTIN or EAN. Where that identifier is missing, wrong, or shared across variants, your item lands on an orphan node with no traffic and no explanation. For manufacturers and industrial sellers the same failure arrives as inconsistent MPN and manufacturer naming, plus resellers listing your own article on a node you do not control and undercutting you on it. I audit identifier coverage article by article, produce the correction list, and specify the variant model: one row per purchasable article, never one per product family. Per engine: Idealo sorts on total price and treats delivery data as first class, and its checkout product changes the arithmetic because the sale can complete without the shopper reaching your shop, which suits some catalogues and not others. Kelkoo bids by category and rewards feed freshness more than most merchants expect. PriceRunner is unforgiving about identifier quality and node matching. Category mapping into each engine's own tree is done by hand for the categories carrying your margin and left automatic for the tail, which is the reverse of what most feed tools default to. Not part of this: I specify, your developer or your feed tool implements. I will review a Channable, Productsup or DataFeedWatch configuration and write the rules it should enforce, but I do not take write access to your shop system. No Google Shopping, deliberately, because it is a different auction and bundling them lets a weak channel hide inside a strong one for quarters at a time. No pricing strategy: I will show you the price rank a listing needs and what that costs in margin, and the decision stays with you. Wrong fit: Catalogues without GTINs and without a route to obtaining them. Single-brand shops selling only their own articles at their own price, where you already own the node and the engine simply adds a click cost to a sale you would have had. Shops that cannot set delivery cost fields in the feed independently of the on-site display, because that one limitation blocks most of the ranking work and no clever mapping compensates for it. What arrives: A field-level feed specification with example rows, the identifier correction list, per-engine category mapping for your margin categories, bid levels per category with the arithmetic behind each, and a monitoring rule that tells you when a node has been taken over or a price parity break has cost you a listing.