Longevity Holistic.

I prepared the source catalog from Shoper for import into Shopify. I separated products and variants and organized SKUs, images, manufacturers, tags and categories. Preserving the relationships was essential: the correct color, image and inventory position still had to represent the same variant.

I first checked one product with its category. Only after confirming the correct assignment did I run the import of the main file: 639 products, 903 SKUs and 3,469 images. I also verified the variants in the Shopify dashboard. This was a completed stage of catalog work; it does not mean that the current public website was migrated to Shopify.

The catalog contained 57 products with multiple variants. I checked repeated option combinations and, where color alone did not distinguish two SKUs, added a second option. Instead of removing a duplicate, I preserved the real difference between the items. On one of the control pages, I checked 23 variants together with their images and inventory levels.

I created a structure of 157 automated collections: 20 brands and 137 brand-and-category combinations. Tag-based rules assign products without each collection having to be completed manually. I checked selected product groups and their presence in the menu.

The next part of the work concerned the public store and its PL, EN and CS versions. I completed and corrected product descriptions, ingredients, directions for use, tab names and SEO fields. The translation had to preserve product information, units and variant labels; a short description did not replace the full source content.

Not all information belongs in one description field. Some of it was stored in additional modules. I separately checked how it was saved and what the public product page actually displayed. I corrected missing and mixed-language content in ingredients and directions for use. Checking the dashboard was the beginning of verification, not the end.

I made changes in stages, with a data backup and consistency checks before saving. After saving, I checked both the content and the public pages again. Materials for further category localization were prepared separately; I do not present those proposals here as already published content.

I reviewed URLs and links, recording the source page, target and type of problem. The report distinguished a 404 response from a DNS error, a block and a redirect loop. During the final correction, I removed false positives involving email links. The result was a spreadsheet for subsequent fixes, rather than a promise that scanning alone had repaired every URL.