Why I’m building Catalogica

Before I started building Catalogica, I managed multiple ecommerce stores. Preparing product listings was part of the work, and I knew how much time it could take.

Getting the products from a supplier was only the beginning. You still needed enough information to sell them: a useful title, a description, the right specifications, suitable images. Depending on the product, you also needed a manual, a datasheet, a video or content in another language.

Some of that information arrived with the product. Some had to be found elsewhere. Then it had to be checked, rewritten and fitted into the store’s catalog.

I started building Catalogica in March 2025 because I had dealt with that work myself and wanted to make it easier.

Working with Ozone changed the product

My experience managing stores gave me a starting point. Having Ozone.bg as a client changed how much I understood about the problem.

Ozone is a retailer in Bulgaria that works with many suppliers. Building software for their catalog gave us a much closer view of what happens when this work runs at scale: the variety of incoming data, the differences between product categories, and the decisions editors have to make before information is useful.

It also challenged assumptions I brought from my own experience. Knowing the work firsthand helped me start, but it didn’t mean I already understood every version of it.

Knowing the work firsthand helped me start, but it didn’t mean I already understood every version of it.

More than 70,000 products have now been enriched through Catalogica at Ozone. The number gives some context for the work. The relationship has mattered much more: it has given us enormous insight into the actual problems a catalog team faces and shaped what we build.

A product listing contains a lot of decisions

Consider a simple example: a supplier gives you a model number, a short name and an image.

You find a manufacturer’s page with a longer description. There is also a PDF covering several models, a video showing a similar version and specifications expressed differently from the ones your store uses.

Someone has to establish which information applies to the exact product. They need to distinguish the product’s dimensions from its packaging dimensions, check whether an accessory is included, and decide how to present the result.

Then there are the store’s own conventions. Which details belong in the title? Which specifications should be separate fields that customers can filter by? What should stay unchanged when the description is translated?

Those decisions affect whether the listing helps someone choose the right product. Generating readable text is one part of that work.

That is the problem we are working on with Catalogica.

It takes supplier feeds, spreadsheets or a starting point as small as a barcode and researches and prepares product information around the retailer’s catalog rules. That includes titles, descriptions, structured specifications, images, PDF documents, videos, translations and SEO fields.

We handle feed mapping, implementation and rule configuration with the customer. Their team reviews the results and makes corrections, using our editor or working in their own system through the API.

Those corrections matter. They give us something concrete to examine: where the information was wrong, where a rule was unclear, or where the output didn’t fit the team’s needs.

Why we’re talking about it now

Catalogica has been in development since March 2025, but we are only now starting to introduce it more widely.

Much of the work so far has happened through building and operating it with Ozone. We now have a working product and experience using it in a real catalog operation.

We also have more to learn. Other retailers have different suppliers, conventions and ways of working. I want to understand which problems we share and where our approach needs to change.

That is why we are starting those conversations now.

If you manage an ecommerce catalog, I’d be interested to hear where your team spends the most time preparing products for sale. We can start there and, if Catalogica looks useful, explore a pilot on one supplier feed.