The search engine algorithm is confidential, and it is said that it is being adjusted every day. No one knows the details. However, Ethan believes that search engine algorithms originate from two methodologies: procedural reasoning and user judgment. Procedural reasoning is divided into two dimensions: relevance and importance. User judgment multiplies relevance and importance by a coefficient respectively. The keyword relevance is determined by the keyword density and distribution in the web page text field. Importance is also called weight. The weight of a web page is determined by its pedigree and content quality. Therefore, Ethan summarized four aspects of search engine algorithm: keyword relevance, web page pedigree, web page content quality, and user judgment, which can be called the four major factors of search engine algorithm. Thinking from these four aspects, we can often find a theoretical basis for the widely recognized seo operation and cultivate our own seo thinking. With seo thinking, we can break through the dogmatic operating rules and be flexible in seo practice.
Ranking is not for websites, but for web pages. When a user searches for a keyword, the search engine analyzes the included pages from the following four aspects and calculates a total score. The highest total score ranks first in the natural ranking.
1、 Keyword relevance
Web page text field is a concept created by Ethan. What I want to explain is that when judging the correlation between keywords and web pages, search engines not only examine the text inside the web page, but also examine the text outside the web page, mainly including the text of backlink anchor text, the text of backlink pages, and the text of domain names, All these words form the text field of this page. The so-called theme text means that the search engine will assign some text to each domain name and each page, which reflects the theme of this domain name or this page. We all know that there are keywords and descriptions in the meta tag. It can be imagined that the search engine has added some meta tags to the page, including the subject text of the domain name, the anchor text of backlink 1, the subject text of the page where backlink 1 is located, the subject text of the domain name of the page where backlink 1 is located, the anchor text of backlink 2, the subject text of the page where backlink 2 is located The subject text of the domain name of the page where backlink 2 is located. The weight of these meta tags is much higher than that of keywords and description, and some are even higher than the visible text on the page. With the concept of web page text field, we can understand many seo operations such as external chain construction.
Note that the search engine will translate the keywords searched by the user into the keywords actually queried internally according to the user's search intention. In most cases, they are the same, but sometimes they are different. For example, if you know the region where the user's IP address comes from, the relevant pages in that region will be displayed first. There is an effect that users add regional text when searching. Another example is synonym analysis. Search engines will list the search results of synonyms.
The relevance of keywords and a web page is determined by the density and distribution of keywords in the web page text field. The more relevant the keywords, the higher the ranking.
2、 Web pedigree
Web page lineage is a concept created by Ethan. What I want to explain is that in the view of search engines, a web page has a weight that comes from the url. The search engine will analyze the URL of a web page and get two important information. The first is the domain name. The domain name has weight and will be directly transferred to this page; The second is the case of backlinks. The page where each backlink is located also has weight and will be transferred to this page. The pr value we often encounter mainly reflects the weight passed through the reverse link.
Note that the weight of the domain name is not equal to the weight of the home page. Search engines will determine the weight of the domain name according to the weight of the home page, the overall situation of the website (such as the stability and gradualness of the website, whether there are many directional anchor texts on the site, etc.), and other factors (such as the life of the domain name, etc.).
The better the pedigree of a web page, the higher the weight, the higher the ranking.
3、 Web content quality
In terms of importance, web page pedigree is a transitive factor, which is simply introduced from the outside; The quality of web content is a structural factor, which is determined by various characteristics related to the web page itself. Although search engines are not intelligent enough in this respect, there are still many means. The most important one is how original and fresh the content is, especially the originality of the content. Another example is whether there is directional anchor text and reverse link anchor text in the web page. Search engines believe that if a page carefully writes directional anchor text or its reverse link anchor text is directional, it indicates that the content of the web page is of high quality. For example, it would be better to have an image on the page, a list tag on the page, and words like "first" and "second" at the beginning of a paragraph.
The better the content quality of a web page, the higher the weight, the higher the ranking.
4、 User judgment
The relevance, weight and data source are not from the user's judgment. The search engine is very clear, and the most important thing is whether the user feels good or not. Although it is extremely difficult to obtain data sources in this respect, search engines still have some methods, such as search bounce rate, such as Baidu statistics and Google analysis, such as social network data, which is the future direction of search engine algorithm development, and will play an increasingly important role in ranking. Ethan believes that search engines will spare no effort to obtain the data sources judged by users.
