First, we need to think about the role of semantic search. Originally, Google functioned almost exclusively based on verbatim keywords and phrases. If you searched for a phrase like “best mountain bikes,” you would almost exclusively find results that included this exact phrase or a very close match. But starting around 2013, Google began releasing updates focused on improving semantic parsing of user search queries. Instead of merely looking at the exact phrase, Google attempted to “understand” the search query, presenting results that addressed the search query even if they contain no exact matches or close matches. With semantic search in play, “best mountain bikes” could return pages like “how to choose a mountain bike” or “which mountain bikes are suited for beginners?” Because of this, it's possible to rank for a wide variety of different search queries, even if you pay little to no attention to keywords and keyword phrases.
2. User search intent- On a related note, Google is increasingly chinese overseas america database focusing on user search intent as the core motivation for its search engine results page (SERP) rankings. Instead of looking for close matches, Google is looking for pages that are most likely to address whatever the user is trying to do. There's a great deal of subjective processing happening in the background, and since we aren't familiar with the intricacies of Google’s ranking algorithm, we can't say exactly what's happening. What we can say is that the importance of keyword data likely pales in comparison to the importance of understanding user search intent. If you can give users what they want to, and do it consistently, keywords are merely A secondary consideration.
3. Misleading data- On top of that, independent studies have confirmed that the data you view related to keywords in Google’s Keyword Planner (and multiple popular SEO tools) isn’t accurate – or might be misleading. Sometimes, the data is completely outdated. Sometimes, it's distorted because of tracking bot traffic. And sometimes, you can generate very different sets of data by using different tools, even if those tools claim to use the same primary sources.