Goldman's Nails. Part 1.

Working with GSC (Longtails)

About 15 years ago, I published a series of articles with a similar name at Lipf(dot)ru. Well, it's time to release the sequel. Today we're going to look back at a technique I've talked about before. and in the canal and in a most interesting teleconferences from Misha Shakin. And rethink it.

So, go to GSC -> Search Results and shove a Regex filter (Regular Query) of the following form:

(?i)b(who|what|where|when|why|how|which|which|whose|whom|whom|can|could|will|would|should|shall|may|might|must|do|does|did||is|are|am|was|were|be|been|have|has|had)b.

it's how we clear out all the backlog of inquiries.

This pattern helps to segment 70-80% information requests by generating ideas for FAQs, which increases E-E-A-T and ranking. Based on a single review on Misha's YouTube:

Traffic to the blog increased from 35K to 123K (without linking).

Let's keep going.

To focus on the beginning of a query, we add ^ to the beginning to filter queries that start with these words. For example, ^(?i)(who|what|...)b)

The Russian analog of the regex-pattern for identifying questions in search queries will be somewhat broader, as it includes question words, their case, gender and number forms, as well as auxiliary verbs and modal constructions often used in question sentences.

(?i)b(who|что|de|когда|po|k|k|k|koy|ka|ka|ko|ko|ko|kкие|chia|chia|choo|chi|chio|chi|chio|k|o чем|чему|чего|чем|чём|ком|быть|есть|суть|был|была|было|были|будешь|будет|будем|будете|будут|бы|буду|мочь|могу|можешь|может|можем|можете|могут|мог|могла|могло|могли|должен|должна|должно|должны|иметь|имею|имеешь|имеет|имеем|имеете|имеют|имел|имела|имело|имели|делать|делаю|делаешь|делает|делаем|делаете|делают|делал|делала|делало|делали|можно|нужно|надо|следует|возможно|стать|становиться|казаться|являться|считаться|называться|прийти|приехать|вернуться)b

Naturally, you can hack this Russian query and expand it considerably (don't forget about the 4096 character limit in the console).

For example, add «cho» (cho, che, choe, sho, shta, chta, chota, chto, what, chgo, shtago.), «who» (kta, kvo, kto, whoo, whob, ktago), «gdi» (gde, gdeta, whereto, gdizh), «kada» (koda, kod, kadda, kogdy, kadab), «pachemu» (pochmu, poche mu, pochimu, pachmui), «kag» (kk, kaak, kagto, kyak), «kokoi» (kakai, kakovo, kakta, kakava), as well as (chiy, choy, chiyo, skoka, skoka, skok, kudy, kuta, kudato, kudato, atkuda, otkud, atkudava, zachom, zach, zachim) and so on.

We also add dumb slang expressions and zelepuki like (pachiemu, pochmu, poche mu, pochimu, skoka, skolka, skok, etc.), which allows you to snag phrases like «skoka cost».

Including «cho» and the like in the regex may increase the capture of informal searches on 15-30%, especially in GSC for youth audiences or regions.

Then we shove it all into LLM and filter it by prompt

«I have a list of slang variations of Russian question words: [insert list, e.g.: cho, che, chyo, sho, shta, kto, kta, gdi, gde, kada, pochmu, kag, kokoy, skoka, kudy]. Sort them into categories: 1) By base word (what, who, where, when, why, why, how, which, etc.); 2) By type of variation (internet slang, youth slang, regional dialect, typo); 3) By perceived frequency of use (high, medium, low, based on general knowledge). Display the result in a table with columns: Variation, Base word, Type of variation, Frequency. Add a rationale for each category.»

This prompt will help to structure the data for further analysis

How to use.

The most obvious thing to do is to make an FAQ. Don't shove crooked «cho» in the text!!! You're already ranked by them (since the GSC request was found). Use normal wording and answers based on your content!

The second option is to group the promptom into clusters

«Create clusters of my semantics queries and suggest SEO applications. [attach a list of those queries].»

To be continued....

The key is to update this list regularly: Check it monthly, picking out new phrases. There are tons of them.