A catalog of three color variants of the sneaker, a filter panel with sliders and a large magnifying glass highlighting the orange product card.
Illustrative visual for article topic created with AI.

QUICK OVERVIEW

What to take away from the article

  • A separate landing page should respond to a specific need and contain suitable products.
  • Limiting crawling and banning indexing are not the same task.
  • Variants must have an understandable identity, price and availability also in the product data.

Design categories based on what the customer is looking for

The category is meant to help with selection, not just divide the database. For packaging, purpose, size and material can be important; for textile products, cut, use and printing options. A business category should have a stable assortment, a clear name and a brief explanation of the decisive differences.

Write a customer question for each intended landing page. If two pages respond the same and contain almost the same products, consider merging them or repurposing them. Don't build hundreds of pages just by changing one word in the title. We write about a wider approach in the article about content for SEO and AI search.

Filters can create unnecessarily many URLs

Google warns that filter combinations can consume a lot of capacity when browsing. If they do not have a separate search benefit, the way they are searched must be targeted. However, the specific rules must be based on the addresses and architecture of the given e-shop; blanket copying of someone else's robots.txt can also disable important content.

We recommend sorting the combinations into useful landing pages, common user filters and technical parameters such as sorting. For each group, determine how the visitor will reach it and whether it should be part of the search. Invalid and empty combinations should not form an infinite space of pages.

Example of distribution of e-shop addresses
AreaPostupWhat to check
Main categorySelection of product line.Is it stable, linked and useful?
Selected subcategoryA specific need with its own assortment.Does it have separate content and a different purpose?
A combination of filtersTemporarily narrowing the selection.Does it make sense on its own in search?
AlignmentDifferent order of the same products.Doesn't it create duplicate paths to the same content?
Resources for the chapter:Google: site browsing management with filters (new card)

Canonical, noindex and robots solve different things

Canonical indicates the preferred address between related versions. Noindex asks not to index the page and robots.txt controls crawling access. Therefore, they cannot be confused with one universal button. During the design, it is necessary to check whether the robot can even get to the instruction it is supposed to read on the page.

For this step, prepare a map of typical URLs and expected behavior. Test main category, selected subcategory, single filter, combination of filters, sorting and pagination. Document the decisions so that further expansion of the e-shop does not create conflicting rules. Follow the result after deployment; the mere existence of the canonical brand does not guarantee the selected address or placement in the results.

The product variant needs accurate data

Blue t-shirt size M may have different availability than size XL. Customer, warehouse and product data must refer to the same variant. It is not enough to change the image in the browser if the order or availability still refers to a generic product.

Google describes the labeling of variants using ProductGroup and Product, including the properties by which they differ and the relationships between variants. The specific solution depends on whether you have variants on one or more pages. Structured data must correspond to the visible content and current offer; they do not guarantee an extended result by themselves.

When choosing a variant, verify the address, name, photo, price, warehouse and put the correct item in the basket. We discuss the connection with reservations in the manual for automatic warehouse replenishment.

Resources for the chapter:Google: structured data of product variants (new card)

Products must be available through links

An important product should not be found just by typing its name into an internal search. Google recommends a clear structure from categories to products. Verify common links and access to other list items. A nice loading animation does not replace clear navigation.

Complete the category with content that helps you choose: dimensions, materials, use, compatibility and differences between options. The product detail should also answer the questions that prevent the order. Use your own useful data, verified photos and clear terms of delivery. Don't invent reviews or availability because of structured data.

Resources for the chapter:Google: e-shop structure and navigation (new card)

Inspection after deployment and when expanding the product range

Review site responses, headlines, preferred addresses, indexing rules, and sitemap. Select a sample of products and variants and compare them with the trading system. Monitor search queries, indexing and category performance in Search Console; interpret the numbers along with seasonality and availability.

SEO of the e-shop is an ongoing work. New filters, out-of-stock products and the import of a supplier's range can change an initially good structure. Therefore, include a check of addresses and product data in every major change. You can find the basics of the technical and business proposal in the article about tailor-made e-shop with ERP.

FROM READING TO IMPLEMENTATION

Let's move it to your business.

We will review your assignment and select a specific next step.

Write to us about the project