Online store of garden equipment and spare parts

It will drag on, but this is already a large catalog with its own rules. The equipment and spare parts store we measured gives 55,321 product addresses and 3,868 category addresses — a total of 59,189, collected from 18 subcards of 3,000 addresses plus 1,321 in the nineteenth. The first server response there is 840.5ms, the slowest of the three niche stores we measured, and that's a direct result of volume. On such catalogs, console import works in batches: the admin on tens of thousands of items falls off after a timeout.

Because of the attachment to the model: comes from his lawnmower - sees compatible knives, filters, candles. If there is no compatibility table, we collect connections according to the manufacturer's articles, and this is a longer job. The second way is the depth of categories: in a measured store of equipment and spare parts, the catalog is expanded to four levels, because fittings and knives differ by a trifle such as thread or mounting size. The limit is fair: neither binding nor filter will save if there are no characteristics in the cards — the data must first be put in order.

30-50 working days before the peak, and this is the lower estimate: the term is calculated from the approved TOR, not from the first conversation. On top of that, add the time it takes for a search engine to crawl the directory — a sitemap does not guarantee indexing or positions, and no one speeds up this part. If there are less than two months left until the peak, it is fairer to close the bottleneck of the current store - speed, site map or checkout - and make a complete replacement after the season.

Check that it actually gives addresses. In one of the three niche stores we measured, /sitemap.xml returns an HTTP 200 with a body of zero bytes, and the fallback feed is 404: the file formally exists, the address is zero, and the directory size from the outside is not counted at all. The opposite extreme in the same niche is 33,493 card entries for 8,690 unique addresses, meaning that each card is repeated almost four times under different category paths. Both conditions are visible on the audit in half an hour.

Yes, an XML feed from the same catalog: prices and availability are pulled from the same source, so there is no difference between the site and the platform. In all three measured niche stores, the platform is connected — both as a sales channel and as a delivery method at the checkout. The limit: the feed opens the channel, but does not make your price competitive - in commodity distribution you are compared to everyone at once.

Two of the three measured stores have eight niches — from card and installment purchases to cashless with VAT and cash on delivery; in the third five, and the shop is open. It is not logos that should be counted, but segments: a private gardener, a private gardener, a VAT company. Each additional gateway is a fee and an additional line in the reconciliation, so we determine the list during the audit.

Scheduled import — daily or hourly, depending on how fast the warehouse moves; if there is an exchange with the accounting system, we draw prices and balances from there. Be honest about the limit: the exchange does not fix the account. If there is a re-sort or unclosed write-offs, the site will quickly and accurately show inaccurate data, and between downloads there is always a window in which the last unit can be bought twice.

Because the estimate in this niche depends on one number - how many addresses the catalog together with spare parts will give, and it is impossible to guess it from the outside. The difference between the stores we measured is almost 100%: 8,690 unique addresses versus 59,189. Therefore, first, a free audit in 2-4 working days, and already according to its figures, the amount and term by stages. The document remains yours regardless of who will make the store.