Data analysis is a very important thing for seos, many of whom simply don't analyze the data on the website, and who do things on their own, such as changing keywords, have to be justified. The above figure is a map from a group of friends of the seo research association network (182025355) and, as can be seen from the above, the search engine, pv, uv and ip numbers, since there are no other reference data, the analysis below will vary slightly。
First look at ip versus uv, uv means "unique visitor." what difference does it make with ip? What's the relationship? For example, you can see that in the same office, everyone's ip is the same (not the lan ip), and the difference is uv, which is an independent access to a natural person. The figure above shows a total of 554 ips and 557 uvs, i. E. Three people in the same section visited the site。
So we can draw a rough conclusion that these three are employees of the company, and then we look separately at the relationship between ip and uv. From 100 degrees of traffic, the difference between ip and uv is two, so it can be understood that two people from the same ip come in 100 degrees, and that people from the same ip (on the internet) are less likely to enter the same site, for example, people from one apartment may be the same ip, but the chances of getting into the same site are quite small, so we have every reason to believe that it is staff traffic。
Now, looking at the search from 360, the difference between ip and uv four, a careful friend may have found a problem. The ip here is bigger than uv, and it's usually less than or equal to uv. In fact, the phenomenon is normal, and there are probably several cases:
Ip addresses are absolute, taken from the tcp chain, real and not unique。
2, uv settingscLookie, randomly set, repeatable, but only low enough。
3. The replacement ip of mobile notebooks from time to time can cause such problems。
4 client disabledcLookie or client security level will result in cLookie can't set it up. There's a problem。
5. If a photo count is used, it is not availablecLookie's gonna have this problem。
From the above analysis, we can infer that the same person may have visited the site in different ips, because this is the image that we are looking at to analyze, so we don't know what type of site it is, whether it has a moving end or not, what the user community is, and we don't know, and we can only extrapolate the analysis through this data. So we can understand that maybe employee a visited the website on the company's mobile phone and then looked at it in the restaurant at lunchtime, because the company's office ip is different from the restaurant, so there's more ip than uv。
There is also a situation where user a has entered a 100-degree account or something, visited the site and then came home and visited the site, so there are cases where uvs are smaller than ips。
Google, search dogs and bing's ip and uv profiles are similar to those above, allowing for a simple analysis of the contours。
Then we look at the relationship between pv and uv, and look at the flow of 100-degree searches (here we do not know what the word is), the average page per person = 406/264 = 1. 68, so the average traffic per person from the 100 years passed is 1. 68 pages, followed by 360 searches: 2. 43, google: 2. 46, dog searches: 1. 3, by which we can draw the following conclusions:
1. The flow from google is relatively precise or most valuable。
Next to 2, 3, 360 searches。
Three and a hundred degrees down。
Four, then the dog traffic。
As we can see from this page, the general direction of the development and dissemination of our website in the future is that google traffic is relatively accurate, although it is less than 100 degrees, but if it is promoted, it is believed that google traffic is much higher。
The overall number of page views is about two, which is still low, so we should strengthen the construction of content and the layout of internal links。
Summary: these are just things that are analysed on this map, and they are not very informative, because the correct analysis of a website can be done, but the methods and ideas of analysis can be used, so we should get as much data as possible for analysis of the website as we can, if we are able to make a positive analysis of the website uv and pv through scientific analysis。
