What are the ways to analyze website operation data?

CEC's information technology operation experts believe that the data to be analyzed generally include: what changes can attract more website browsing (such as clicking on network ads to enter); which web pages have the most clicks; the source of website visitors entering the website; through what keywords they enter; how long website visitors stay on various web pages, and so on. Of course, the most critical metrics are how high the sales conversion rate (conversion rate of intended customers) is, how many people register on the site, what the cost per customer is, and so on. In addition, companies want to know whether new initiatives (e.g., customizing new same-day delivery pricing, running promotions on the site to increase sales, etc.) are working. The data needed to analyze the operation of a website can be obtained in a number of ways: the server logs of a company's website record the IP address of the user, what browser the user is using, where the user was before entering the website, the exact time of browsing, and the user's registration information. The IP address allows the organization to know the location of the user, for example, jp means the user lives in Japan. Web tracking files are files that are automatically generated on the user's hard disk when he visits a website. These files come into play when a customer enters the site and performs an action, such as using a shopping cart. When a customer revisits the site, the data in these files can be retrieved to learn information such as the number of times the customer has visited the site. Amazon.com uses web tracker files to automatically generate user names on its home page. Page tags are actually pixels on a page that are invisible to the user. Page tags are used to activate a piece of information on a page as the user navigates through it, such as when to remove an item from the shopping cart. Web tracker files on the user's computer hard drive can also be used to activate tags that show when the user returned to the site and what actions they took on the site.

With web parsing software, companies can analyze server logins and, in turn, parse user behavior patterns.

CEC's information technology operation expert reminds: at present, Google Analytics and Baidu statistics are intelligent, very powerful professional tools used to count the operation data of enterprise websites, with many users and highly respected.

B2C website operations weekly data analysis of those indicators

Users order and payment will not necessarily be completed on the same day, but the weekly data is relatively accurate, so we take the weekly data as a reference object compared to the main purpose is to compare the last week with the last week of the difference between the data, the operation of a certain aspect of the work, the product made some kind of adjustment, the corresponding data will have a certain change, if there is no improvement. The main purpose of the reference object is to compare the difference between the last week's data and the previous week's data.

1. Site utilization: IP, PV, average number of pages viewed, online time, bounce rate, return visitor ratio, access depth ratio, access time ratio.

This is the most basic, each of the data to improve is not easy, which means that it is necessary to constantly improve each of the details found in the problem, and constantly go to improve the shopping experience. To illustrate the important data indicators:

1.1 Bounce rate: a high bounce rate is never good, but the problem of where the bounce is is the key. In my experience, the bounce rate is very high in some promotional activities or the placement of large media ads. A high bounce rate may mean that the crowd is not accurate, or there is a huge difference between the advertisement demand and the content of the visit, or there is a problem with the visit page itself. Regular bounce rate I note in the login, registration, order process 1-3 steps, user center and other basic pages, if the bounce rate is higher than 20%, I think there are a lot of problems, but also based on the bounce rate to improve the shopping process and user experience.

1.2 Return rate = 2 times in a week back to the visitor / total visitors, means that the site attraction, and member loyalty, if the traffic is stable, this data is relatively high will be relatively high, too high means that the development of new users is too little, too low means that the user's loyalty is too poor, the repurchase rate will not be high.

1.3 Access depth ratio = access to more than 11 pages of users / the total number of visits, access time ratio = access time of more than 10 minutes of the number of users / the total number of users, these two indicators represent the attractiveness of the content of the site, the higher the rate of data the better.

2. operation data: total orders, valid orders, order efficiency, total sales, customer unit price, gross profit, gross margin, order conversion rate, payment conversion rate, return rate;

daily data summary, weekly data must be stable, mainly compared to the last week's data, focusing on guiding the operation of internal work, such as product guidance, pricing strategy, promotional strategy, postage

2.1 Compare the data, why the number of orders decreased? But sales have increased? Is this a good thing?

2.2 Compare the data, why did the unit price increase? But margins are down? Is this a good thing?

2.3 Comparing the data, is it possible: sales increase, margins increase, orders increase? It is not impossible.

All questions can be answered in operational data.

How to quickly get started with website data analysis and operations

First, how to get started with Internet data analysis

1, website analysis is a kind of ability

For most of the people in the Internet, website analysis is a kind of ability, because based on the conclusions of the website analysis on top of the colleagues can guide the work of operations, product, design, and technology.

2, website analysis to solve the problem

Who are the users (target users),

from where (where the traffic comes from, the value of the traffic, etc.),

to where to go (why leave, how to reduce the loss of users)

3, for the product OR operation, website analysis can do

Product revision is reasonable?

The user's behavior is not the same as the user's behavior.

How about user feedback?

What features are problematic?

How often are features used?

Is the conversion path reliable?

For operations:

Where do users come from?

How active are users?

How do you allocate your advertising budget

Is the content effective?

How to break down KPIs?

4. Why Web Analytics

5. The Core of Web Analytics

2. The Process of Web Analytics

Define the Problem - Measure - Analyze - Improve - Maintain

3. Define the Problem

How you already know how to effectively describe a problem, then you're already halfway there because you know the problem and you know how to ask it.

The job is not to test coupons designed to ask you questions, first you have to find the problem yourself.

For example, a decrease in the conversion rate of registration is positively correlated with a very large number of problems.

Is there enough product support?

Avatar upload

Email verification

Required information

Is marketing on point?

How is the ratio of new and old visitors

How is the external word of mouth

Elements of the problem: essence, phenomenon, characteristics, quantification

Define a problem: that is, to the entire team to confirm the direction of a downward decomposition of this goal around the development of the plan, the process of the specific implementation of the plan to find a certain problem, and then come back to the specific analysis of the.

So as a web analyst, you should start from the company's strategy, understand the product, operation, technology, business logic and other levels of knowledge, and provide a lot of advice to the company's development.

The first time I saw this book, I was able to get a good understanding of what it was like to be a product manager, and I was able to get a good understanding of what it was like to be a product manager, and I was able to get a good understanding of how it worked. The number of recommendations in the Douban list is 986, and the number of favorites is 7774. I'm ashamed to say that I've only read 20% of the books in the Douban list.

Dang, just by reading the book is unable to cultivate the industry pattern, but also need to be good at asking people for advice, make good use of network resources, their own experience, practice and so on.

Job search Internet data analysis, how to prepare for industry knowledge?

Fourth, measurement

Collect data.

Currently commonly used data traffic monitoring work:

Google AnalyticsGoogle website analysis tools

Omniture Omniture SiteCatalys

Baidu Statistics Baidu statistics tools Tencent analysis is mainly for the forums

etc. 。。。。

For example, the data of the education industry can be found from some industry data collection sites

In addition, as a product OR operation will not write the program, you can only get the data through the third-party tools or platforms, or to the technical students to raise the demand.

Technology is the first productive force. If you know some SQL or Python, the data you get is too much to be too wonderful ......

Recommended books: what are the books you have to read to do data analysis?

The books recommended below this Q&A are basically about data mining or acquisition.

V. Analyze, Improve, and Maintain

For example, the player industry trajectory of a game is like this

So when analyzing it, we decided to focus on the churn of new users

Analysis of the types of tasks that are churned:

Complicated operation

Tasks that are not smooth and fluid

Slow upgrades

There are group tasks or other interactive tasks

Then there is a continuous cycle of optimization. Problems are analyzed, user needs are identified, the product is improved, further statistics are kept and the results of the enhancements are maintained.

The process of analyzing the methodology is roughly the same, and it is relatively easy to grasp, but when it comes to the specifics of the work, it is far more than what can be explained in a few words, so take your time to practice and grow.

1. Lean data analysis

2. Conversion: techniques to improve website traffic and conversion rates

3. Data analysis : the enterprise's wise help

4. Optimization and operation

5. Everyone is a web analyst: from the analyst's point of view to understand the site and interpret the data

6. Graphical website analysis of the 36 big data

Through the website data analysis, on the network marketing to play what role?

Website data is the data that directly reflects the effect of online marketing.

The effect of online marketing is to need website data to reflect.

If there is no website data statistics is difficult to quantify the results of network marketing, there is no quantitative data statistics, you can not have a holistic analysis of the effect of network marketing, only a systematic analysis will have a good network marketing results.

Kang Naresh digital marketing, long-term engaged in network marketing planning and promotion work.

How can website operation lack data statistics? How to analyze data most effectively

Simple statistical code or tracking traffic sources ...... or according to the purpose of website operation to achieve a particular demand for effective analysis. Early in the construction and operation of the site, when there is very little data, the site focuses more on traffic and channels, while the site operation is relatively mature, data analysis should be more biased in the direction of the functional development of the site, for example, a corporate marketing website, the user purchase rate is particularly sensitive, then the data analysis should be used as the core of the analysis; and for the display of type of website, for the user retention rate is particularly interested in

What is the use of website data analysis

What do you want to achieve, data analysis can play an auxiliary role in supporting decision-making

WeChat operation data analysis how to do?

Starting from the user side of the analysis includes the following points: (micro ask data)

1, including user growth statistics and user attribute statistics.

2, user growth statistics, is by day statistics, there are 4 dimensions:

3, new attention, unfollowed, net increase in attention, cumulative attention.

How about website data analysis

quite professional, webmasters must be ah ~ ~ also go out and care about the site traffic with a cell phone, to not be able to go online at the computer, directly after the web page GA, but also hard to get 3G to see how troublesome.

How to solve the refinement of the operation through data analysis

This question asks a wider range of what you need. How to understand some of the data of their own stores, and then based on the combination of data, to change the entire store Buji.

How to do a good job of APP data analysis and operation through statistical analysis tools

1

Industry data

Industry data is crucial for an APP. Understanding industry data, you can know the level of your APP in the whole industry, and you can compare the difference between your product and the industry average and the ranking of your product's corresponding index in the whole industry from multiple dimensions such as new users, active users, startups, and hours of use, so that you can know the shortcomings of your product. This vertical comparison will make their product positioning and development direction clearer.

2

Evaluating channel effects

In China, there are a lot of channels to acquire users, such as microblogging, WeChat, carrier stores, operating system stores, app stores, pre-installed by cell phone manufacturers, CPA ads, cross-promotions, time-limited freebies, and so on. When looking at the data of an APP, the first thing you need to know is where the users come from and where the quality of the users is the highest, so that the developers will face a difficult problem of choosing and evaluating the channels. But through statistical analysis tools, developers can compare the effect of different channels from multiple dimensions of data, such as from the perspective of new users, active users, next-day retention rate, single-use duration, etc. Comparison of different sources of users, so that you can find the most suitable channels for their own data, so as to get the best promotional results.

3

User analysis

After a product attracts users to download and use it, the first thing we need to know is who the users are. Therefore, we need to exhaustively understand the user's device terminal type, network and operator, geographical distribution characteristics. These data can help understand the attributes of the user, in product improvement and product promotion, you can make full use of these data to develop accurate strategies.

4

User behavior analysis

After focusing on user attributes, we also need to pay close attention to user behaviors within the app, as these behaviors ultimately determine the value that the product can bring. Developers can set up custom events and funnels to focus on the conversion rate at each step within the app, and the impact of the conversion rate on revenue levels. By analyzing event and funnel data, you can target and optimize steps with low conversion rates to improve overall conversion levels.

5

Product popularity

After understanding user behavior, we should look at whether our product is popular enough, which is fundamental to keeping an app alive. Developers can evaluate user viscosity in terms of retained users, user engagement (length of use, frequency of use, pages visited, intervals of use) and other dimensions. When comparing and analyzing data, it is important to make full use of the time control and channel control, so that you can compare the user viscosity of different channels at different times, and understand the effect of operation and promotion methods on different channels.