The moment we knew there was a problem.
2019, a manufacturing customer in southern Germany. 280,000 products in the shop, technical data sheets, hundreds of variants per norm. We had just installed a standard search that shone in every demo. In production: 38% of search queries led to zero results. Not because the products were missing. But because customers searched for "1.4571 m12 cylinder head" and the search understood "cylinder head" but treated "1.4571" as an order number.
We realised: there was no standard product on the market that really handled B2B search semantics. What did exist: plugins optimised for B2C. APIs that billed per search. Enterprise solutions that took 4 to 8 weeks to set up and ran on US servers.
"This is the biggest unsolved conversion problem in German B2B commerce. And nobody talks about it, because it sounds technical and is hard to fit into a pitch deck."
Three years of tinkering before we even used the word platform.
We first built a Shopware plugin. It worked, the customer was happy. Then the second customer, then the tenth. By now more than 200 shops. With every project we learned: understanding the search query is only half. The other half is data quality. And without analytics, nobody knows whether the search actually works.
The plugin became a platform with three modules: Search, Health, Analytics. Not because we planned it that way, but because our customers kept asking us: "Don't you also have …".
Why a brand of our own.
Eywora has just entered the market. We are doing this deliberately: a platform available for every e-commerce stack. Not only for Shopware shops. Not only for our existing agency customers. We believe the next three years will decide search understanding in the German mid-market. We want to be the tool that makes that happen.
What stays: our roots in practice. We have lived every pain point two or three times ourselves before we put it into the product. And we will keep building this way: in direct dialogue with customers, not in a workshop.