How can I build a website with tens of thousands of daily traffic?

This is the result of traffic statistics in May 31. The website has been launched since the end of March for almost two months. In middle May, the independent IP address of the website remained around 7 kB. It wasn't until around 25 that we broke the W line, but it was a bit turbulent in those days. Now it is basically stable at around daily IP addresses and around daily PV. Most of the sources of these traffic are long tails. Because of the popularity of the website, some vertical portal websites are on the Baidu homepage. You cannot compete with them in terms of resources. However, we can make a long tail, and a page can only make one topic, but we can make many long-tail words, which is also very good at doing it.

This is the website access statistics in various periods. From the perspective of the review rate, the sorting is good, because we have been trying to improve the website stickiness for a long time. Later, with the guidance of an elder in the group, some adjustments were made in content relevance and extended reading, making the entire platform more responsive to the needs of netizens and friends.

Through these two months of practice, we found that seo is not as difficult as we thought, but the key is the idea. However, SE itself can mine deeper things. For example, the concept of probability theory, spatial vector model, the weight of tf-idf analysis keywords, the role of Bayesian nodes, and the concept of classification weight. It is good to know more about these basic and theoretical things. At least you can know how the search engine tutorial uses a dark box and how the weights are obtained. I did not know how the rankings of some previous sites were made. When I fell down, I tried my best to add external links. Sometimes it was often counterproductive.