First, traditional website data analysis
1. Visit
A visit is when a person comes to a website, looks at some content and then leaves. This process is also called a session, or session. Note that the session will not be automatically destroyed after the browser is closed, because the session on the server usually has a default expiration time of 20 minutes. If you open the session after closing the browser and access the same URL, as long as the browser’s cookie still exists (that is: If cookies are not set to auto-destroy when the browser is closed, the server will still assume that they are logged in, that is, belong to the same session. Conversely, if the browser is not closed but no action is taken, the session is destroyed after 30 minutes. Subsequent accesses belong to a new session
2, traffic visits
The number of visits during a period of time is the number of sessions during that period
3, number of visitors UV
Unique Visitor, the number of people who visit a website. How do YOU identify a user? In the website analysis system, users will be assigned a number according to their browser, device model and other information, and this number becomes cookie.
The number of visitors is the number of cookies that visit the site. If the same person visits the site with a different browser or device, his cookie changes, that is, a different user.
4. Page views
Often referred to as PV (Page View), is the number of faces viewed. Visists, UV and PV are the popularity indicators of a website
5. Page stay time
The length of time a visitor spends on a page is equal to the total length of time on the page divided by the number of visits to the page
6. Website stay time
The length of time a visitor spends in a session is equal to the total time spent on the site divided by the number of visits
7. Jump rate
The common algorithm is that of all the sessions on a site, the percentage of sessions that arrive at the site and leave without doing anything is equal to the number of visits to the landing page divided by the total number of visits
8. Exit rate
This measure measures the percentage of web site exits from a page and is equal to the number of web site exits from a page divided by the number of visits
9. The difference between page bounce rate and exit rate?
Exit rate is the percentage of users who end up exiting from any page they enter;
Bounce rate is the percentage of people who jump from this page to this site without doing anything
10. Conversion rate
The number of visits to achieve a goal divided by the total number of visits, or the number of visits to achieve a goal divided by the total number of visitors
For example: Place an order
Is the denominator the number of visits or the number of visitors?
Using visits as the denominator means that every visit is considered an opportunity to order or buy;
If the number of visitors is used as the denominator, it is considered normal for a visitor to visit multiple times before placing an order.
How to conduct macro analysis of the website, not too entangled in the details?
(1) How many visitors visit the site and how deep is the visit?
Through audience visitors
(2) Where do these visitors come from and how effective are they?
(3) What do visitors do on the site?
A) View the landing page with the most traffic
B) View the page with the most traffic
C) Click the heat map on the page
D) The transformation funnel of important processes, and analyze which process points have the highest loss rate
Second, mobile application data analysis
User acquisition phase
1. Downloads
The number of users who have downloaded the app, as well as app Store rankings and ratings. High download rankings and high ratings help attract more users to download.
2. Install activations
Number of devices that have the app installed and open
3. Activation rate
Activated devices/Installed devices
4. Add users
If the device is the first active application, it is new. The number of users of a mobile application is usually a unique device, so the number of new users is the number of new devices
5. User acquisition costs
The cost per user acquired
User activity and participation
6. Quantity indicator: Daily active users and monthly active users
Number of devices that have started applications in a period of time, indicating the number of users. The number of daily active users can reflect the effect of traffic introduction on the day, but fluctuates greatly. The number of monthly active users is relatively stable. The size of an application is usually expressed in terms of monthly active users.
7, quality indicators: activity coefficient, average duration of use, function utilization rate
Active coefficient = Daily active Users/Monthly Active Users
Average duration of use: The amount of time an average user spends on an app in a day
Feature usage: The percentage of active users who use a feature. The higher the usage, the more popular the feature.
Retention phase
8. Next-day retention, 7-day retention, 30-day retention
Definition: N-day retention calculation refers to the percentage of new users or active users that come back on the NTH day. Look at industry values to see if your app’s retention is healthy.
General interpretation:
If the next day retention rate is low, it means people are not interested in our app.
Seven-day retention means it’s not playable, it’s not fun;
The 30-day retention may be that our version was poorly iterated and didn’t deliver content in time
9. Function utilization rate: Monitor the proportion of the number of users of a function;
10. Continued use of feature: The percentage of users who used feature A this week and continue to use feature A next week represents the popularity of feature A
Use of new function – Promotion effect on core function: proportion of listeners who have used function A (such as adding the latest song function) – (proportion of listeners who have not used function A)
If the value is 0, there is no contribution, if it is greater than 0, there is contribution, and if it is less than 0, it is negative
The 7-day retention rate of cloud music tourists (non-login users) and login users was compared before and after the revision, and it was found that the overall retention data after the revision was improved, indicating that the effect of the new iteration version of the mobile terminal version was better.
Then compare the proportion of new users listening to songs within 14 days before the release of the new version
Year-over-year: 14 days after the old release; Sequential: 14 days prior to the release of the new version
Found new users listening proportion has increased obviously (the number of songs (new users rate – the number of tourists and listening to music (ratio of new users – the logged in user two classification index), the core function – music utilization rate has increased significantly, redesign is successful, the new has good effect to guide users to the proportion of listening to music!!
It can be seen from here that in order to obtain these data users to support the later operation, we must put forward the data statistics related product requirements in the product design stage, and realize!!
We should note that in the evaluation of the effect of the new version should be used to measure the new users, because the old users themselves are quality users, do not eliminate the old users, it is difficult to draw obvious conclusions