How does the website manage a large number of long tail keywords when doing SEO?

For example, 200,000 long-tail keywords, how to classify, manage, create related pages, rank tracking and other related work? This question is really a very, very good question.

The management of 200,000 or even millions of words will be a disaster without better management measures.

It is said that it is better to put time on keyword management than on content creation, which is his answer; With this energy, it is better to focus on new content mining!

Since it is a long tail word, I don't know what industry you are in, but it may be time-sensitive! It is also possible that there is no such defect! However, it is difficult for you to solve the ranking of each long tail word, which involves whether the layout structure is considered when designing the website structure. If there is a system structure, then communicate with the program and make corresponding tools! For example, export the keyword list according to the log, and then import the query ranking! If there is no system architecture with 200,000 times and 200,000 pages, then even monitoring is a huge project and makes sense.

The premise of this truth is that you can't manage these thesaurus. I can't manage keywords, and I hope to pile up content to cover up the problem. Ok, now you have two confusing questions-keywords and content.

This problem is worth writing a few articles to talk about. I wrote an article about establishing an effective page database in SEO: purpose, definition, process and application. The core logic can be learned from the management of massive keywords-building a keyword database. Here is a simple idea.

Please note: the idea is the key, not the technology. Mastered the train of thought, even if you use Excle, you can do it. Of course, upgrading skills is the best. Excel is a bit slow to process data above100000.

It has been assumed that your words themselves don't need to be cleaned up, and I don't care how you come from these words (of course, the source pattern has an influence on the subsequent sorting method). Anyway, there are already 200 thousand words, which is very good for keyword management practice, and it feels a little less without it.

First, create a data table.

First of all, we should establish the concept that any word is a record. Similar to any page is a record. In fact, any page has a corresponding relationship with words (not one-to-one correspondence).

Because each word is a record, the record has corresponding fields. Records+fields form a data table, and multiple tables form a database. The core of this is to build the data table.

The detailed logic will be a bit circuitous, as shown in the figure.

Chart 1:

It looks very regular.

Chart 2:

It still looks very regular.

Chart 3:

It still looks very regular.

Chart 4:

It still looks very regular.

Chart 5:

Anyway, it's very regular.

The completion of the above table construction far exceeds the small goal of how to classify, which is simply a big break compared with classification.

How did these things evolve? Combine these five charts. It seems that I can't find a lot of information about keyword management on the internet, and I have figured it out in actual demand. There must be many people in this industry who have various routines to manage these massive words, but they have not seen sharing.

After proficiency, the key word of the 100,000-level is to plug teeth, and the million-level is to drink soup. But if you want to do detail management, you need to have a further understanding of the business, otherwise it is easy to make jokes.

Secondly, create related pages.

Pages include templates and content. In fact, from the previous table, you can clearly design the page template.

The next step is to determine the content. Consider the page type before determining the content. There are two basic types: detailed pages and aggregated pages. The two can even be transformed into each other, but it is recommended to distinguish these two basics, or even be more detailed. There can be a division between detailed page A and detailed page B, and so on.

Besides PGC, content is understood as data organization, which will significantly improve efficiency and quality. The content is basically completed by several basic actions such as copying, combining, simplifying and associating.

The entry is the landing page, and the subdivision of the entry and its relationship with other entries (this relationship needs to be done in another table) have decided what the page content is. Suppose two people produce content under the guidance of thesaurus. The specific words are different, but the basic points cannot be deviated. Deviation is the problem.

Third, ranking tracking.

There are two kinds of keyword ranking tracking: one is uncertain page ranking tracking corresponding to keywords, and the other is certain page ranking tracking corresponding to keywords. Do both better.

The essence of ranking tracking is statistical analysis of data. Then you have to follow the rules of data analysis. Someone mentioned grouping sampling earlier, which is correct. He also mentioned random sampling. Although random sampling is also used, there are still some preconditions. The sample should have typical significance. Just like 10000 brand words, we generally can't do random sampling of brand words, because 80% of these brand words may not be very important, while 18% is generally important and 2% is very important. Pay attention to the application of this 2/8 rule in data analysis.

There are other commonly used thinking methods in data analysis, such as quadrant method, contrast method and multidimensional method. And can be flexibly applied together.

Finally, once again, the ready-made tools on the market can satisfy these actions, but many times it is very complicated, there is no clear idea, and we are still wondering how to traverse 200,000 or even millions of words. In the programmer's words, there is something wrong with this algorithm. So many times, optimizing thinking is the shortcut and the right way.