Semantic Core Checklist

It is better to have them. Be sure, if the site is not ranked for a branded query, the sooner you know about it — the better.

The perfect number is 3: in Cyrillic, Latin, and something else with additional information, with a chance to not be in the first place. But a few dozen branded queries are unnecessary.

The first phrase can be added as soon as you open the site and examine the first screen. The first phrase is a description of the business site as a whole, its specialization. We can immediately write phrases such as online electronics store, online farm food store, and so on.

A very important part of semantics. There are many sources, from menus to parsing competitors. The menu has a plus: the client apparently already sorted sections by business priorities. Sometimes, it's hard to get this information from a client.

Don't use product categories copied verbatim from the menu. Let's take the section «Smartphones and accessories» as an example. Nobody ever types in "smartphones and accessories." Make 2 different queries out of it.

You need to develop product categories. For example, televisions. TVs are further divided into TV brands, flat TVs, TVs of a certain size, TVs with some technology, etc.

Next, look at the bundles. For example, such as product category + brand, product category + area of application, or product category + solution to a problem. Examples: Nike sneakers, Toshiba TVs — these are product categories + brand. Bathtub linings, cottage linings - area of application. «Tablet for a child» — user segment that will use this product.

Do you know about the tag pages of online stores?

More often than not, product category + adjective queries are relevant for them. It is important not to copy everything but check for frequency. For example, we know for sure that iPhone Gold is in demand.

When you open the category page of the online store, most often, there is a filter that allows you to select products according to many parameters: supplier, color, volume, material. Take a look and try to figure out which of these filters may be frequent.

I recommend including a couple of item numbers in the semantics. For example, PPA-LX1 or ME88SUG. At least, these are indicators of the soundness of the internal optimization.

The semantic core should not include such queries as «Galaxy 5G/5K mAh, 15W FC, Samsung». This example should be replaced by "Samsung Galaxy A32 5G»

The ranking of short-tail queries is affected by the query index. It is best to boost it through long-tail queries with high relevance. But wordstat parsing and databases do not work well here. It is better to come up with your own matrix of multiplication of semantics with toponyms.