Catalog site without a shopping cart

Because the price is not always fixed. If it depends on the volume, configuration, exchange rate or contract with a specific buyer, the cart forces you to put a number, which is then viewed in the conversation. A catalog with a request for calculation is more honest: the buyer sees the assortment and parameters, and receives the amount from the manager. Technically, it is also easier - there is no checkout, payment providers and order statuses, so the term is shorter.

It is possible, and often worth it: a guide price or range keeps people on the page instead of sending them to Google further. We display it as a reference with a note about the conditions, and not as the amount to be paid. At the same time, we do not put the price in the Product markup: to announce the price in structured data and not give the opportunity to buy is a difference between the markup and the page, and manual sanctions arrive for non-existent markup, not benefits.

A few thousand is fine, if the import and filter are done correctly. The B2B directory we measured held 1,848 cards and 138 categories and had a first response time of 388.1 ms with 120.2 KB of HTML. It is usually not the platform that limits it, but the filling method: several thousand lines cannot be entered manually through the admin. Separately, about the filter: it generates a combinatorial number of addresses, and if you don't cover them with canonicals and robots.txt, you will end up with thousands of almost identical pages in the index - this is a degradation, not an improvement.

Import according to the schedule. You put the file in the agreed place or give the download from the accounting system, the catalog is updated daily or hourly - depending on how fast your data moves. Manual editing is reserved for individual edits. What the exchange doesn't do: It doesn't fix the record. If the balances in 1C are inaccurate due to resort or unrepresented write-offs, the site will quickly and accurately show inaccurate data.

This is the most frequent state of the price and a separate stage of work. As long as power, size, and compatibility are inside a paragraph, they don't become a column or a filter—the machine sees a solid line. There are two options: analyze with a parser according to the template, if the description is of the same type, or compile the table by hand. What exactly will come out is visible from the file, so the price analysis is the first stage, and not after the design.

No one guarantees a position. Technically, this is what we close: each position has its own CNC address, a unique title with an article number, characteristics with text, not a picture, and a canonical for items available from several categories - it is on the last that most of the takes are built. A fair margin: the sitemap does not increase items or guarantee indexation, and the actual number of indexed pages is visible only in your Search Console, which we do not have access to by default.

Yes, if it is known at the start. Then we take the platform with the expectation of the future checkout and put in the structure the fields that the store will need: the balance, the unit of measurement, the minimum batch. It is also possible to add a basket to a catalog made without this assumption — but this is a reworking of the structure, not an improvement of the interface, and it is counted as a separate project.

We measure what is available: the number of calculation requests, which filters are applied, which requests are entered into the search, and how many of them are returned empty. The last is the most useful list in the project: it is what they want from you and what you don't have. A tip from experience: Check the counter before starting a new catalog. In the parts catalog we measured, the marking still said Universal Analytics, turned off in July 2023, and there was no GA4 at all - three years of operation without a single number. And separately about accuracy: GA4 shows a trend, not an accounting truth — blockers, consent mode, and browser restrictions on cookies eat up part of the data.