Online pet store: feed, scales, repeat orders

Usually one product with a choice of weight: the buyer compares the formula, not the packaging, and it is more convenient for him to switch packaging directly in the card. Separate cards are justified when a specific weight has its own article number and is searched for by a separate request. Our recommendation: as long as there is no such request, do not produce cards - three almost identical pages compete with each other and blur the site navigation. We fix the solution in the TK before the assembly, because the number of addresses and the work with duplicates depend on it.

The technical part is simple: order history, a repeat button in one step and reminders by means of the postal service. We do mechanics. Whether it will work depends on the assortment and price, and we will not promise a specific increase in repeat purchases - we do not have our own measurement of this indicator. What can be measured immediately is the share of repeat orders, provided that the site has analytics. In two of the three stores we measured, she was not standing at all, so a similar conversation there would be based on words.

Part of it is yes, part of it is no, and this question is not related to the site, but to the rules of circulation of a specific group of drugs. Technically, we display the required fields, card warnings and order limits, but you are responsible for what exactly you are allowed to sell online. Let's talk about it at the start: it is more expensive to transfer a part of the catalog to another branch after launch than to separate product groups at the structure stage. One of the stores we measured solves a similar problem with a legal block right in the card — the buyer reads the restrictions before the cart, not after.

Because it can return zero bytes. In two of the three measured stores, /sitemap.xml returned an HTTP 200 with an empty body, the sitemap feed returned a 404, and the Sitemap line in robots.txt was commented out from the default template. In the browser, everything looks normal, in fact, the search engine does not receive any address. We check this first and measure it in response bytes. A fair limit: the site map itself does not raise positions and does not guarantee indexing - Google writes about it directly, and the real number of indexed pages is visible only in your Search Console.

As much as is actually in the data. Animal type, age, animal weight and special needs close most of the selection scenarios; then the profit growth falls rapidly. Two limits to be aware of in advance. The first: on a catalog of several hundred items, complex filtering does not give much - there a clear structure of categories works better, and a measured irrigation store with 300 items on four nesting levels is just such a case. The second: filters generate a combinatorial number of addresses, and if they are not closed by canonicals and robots, thousands of almost identical pages will go to the index.

No, there are enough buyer groups: the breeder or clinic sees its price and multiple after authorization, the rest of the catalog is the same. A separate entrance makes sense when wholesale is your main direction and retail lives next to it according to the residual principle - then it is better to build the frame around wholesale, and this is already a neighboring page. We don't recommend the intermediate option of "a little wholesale on the side": it usually ends up with two catalogs that someone has to keep in sync with their hands.

Most often, the theme and the number of items on the page, and not the engine, are to blame. In the measured irrigation store, the main one is given in 235 ms and weighs 143.6 KB, and the category page with 100 items is already 554 KB; it's the same site. Therefore, we fix the weight at the design stage separately for the main and separately for the heaviest category, and do not measure after complaints. The line of fairness here is simple: speed removes the technical loss, but does not make the offer attractive. A page that loads in 200ms will simply give a person a reason to leave faster.

The ones that your buyer actually uses, and nothing more. Measured benchmark: in the irrigation store, there are six ways to pay in the checkout from two carriers — a branch, a courier, and a Nova Poshta post office plus three Ukrposhta options. Choosing a branch right at the checkout removes the need for the manager to clarify the address on each order. What it doesn't do: it doesn't make the delivery itself cheaper, it doesn't speed up the carrier, and it doesn't save the buyer who chose the wrong branch. And the freed hour of the manager becomes money only when it is really redistributed to something.

25–45 working days from approval of technical specifications, six stages. The catalog in this niche is usually medium-sized, so the main time is not spent on imports, but on packaging, filter by animal and age, and replay mechanics. If the price is structured and the weight in it is a separate field, the term is at the lower limit. If the weight has to be extracted from the names of the goods by the rules and checked by hand, use the upper one.