Data thinking Second article: Data index System

Data analysis is inseparable from the tracking of key indicators, indicators are a measure of the quality of an index. There are many data indicators, however, not all the indicators are good, the vanity indicator (Vanity Metrics) looks beautiful, makes you feel good, but such indicators are superficial, and even fraudulent, for example, you will focus on the site's traffic, but ignore the jump rate, the user's browsing time, etc. It's easy for you to fall into the trap of vanity indicators. Effective indicators are those that really bring benefits, such as conversion rates, retention rates, daily activity ratios, and so on, where you can better insight into the real trends in product and user behavior.

What is a good indicator?

Good data indicators must be comparative, it is usually a ratio, and the ratio is the best data indicator, because the ratio is inherently comparative. As an example of driving, miles reveal distance information, and speed (distance/time) tells you whether you are accelerating or slowing down.

The object of the indicator comparison is usually the "ego" contrast at different times, the contrast between different objects of the same period, the former used to compare the trend of things growth, the latter used to compare the growth of different things.

Although the ratio is the best indicator, but, aside from the base, the significance of the ratio will be greatly discounted, citing the "benefit data analysis" in the example, "your product just on the line, from the strict sense, your father registered an account, you can also make your users doubled." Therefore, to gain insight into the story of the indicator description, it must be analyzed together with relevant indicators. Fortunately, an indicator and other indicators are usually associated, for example, the conversion rate is a good indicator, it is usually associated with time, the number of new registrations.

Typically, user behavior analysis consists of two types of important ratios (or one): time-related metrics, called "xxx speed", for example, new user growth (number of new users per time), and number-related ratios, called "XXX rate/ratio", for example, the active user ratio (active users account for the total number of users).

Restore a scene

Data analysis should consider the different applications and scenarios, in the analysis of data, should be combined with the actual business, restore to specific scenarios, design key indicators.

In analyzing the user's behavior, the user's stickiness (stickiness) of the service or product is an effective indicator, and stickiness refers to the degree to which a customer is willing to buy or recommend a product or service to another person. The viscosity is reduced to specific scenes, can be measured from the retention rate, jump-off rate, conversion rate, and other indicators, these three indicators are similar to the meaning of expression, but the use of different scenarios:

• Bounce Rate: For content products, when a visitor clicks on your page, how many people will close immediately? What percent is it?
• Retention (Retention rate): For new users, the percentage of users who have been using the product for a period of time to test the new user's use of the product for a certain amount of time.
• Conversion (Conversion rate): When the product is promoted, the conversion rate is the proportion of the number of registered people and visitors, indicating how many visitors really become registered users of the site.

Activity is also an effective measure, and for social networking sites, the percentage of daily active users is a key indicator.

Note: Although the ratio is the best data indicator, but also need to assist other qualitative and quantitative indicators, the indicator has never existed alone, but should be from a number of indicators to synthesize evaluation, it is necessary to understand the coupling between indicators, design indicator system, from all angles to reveal the hidden facts of the data.

Case description

Case: In a Cloud product presentation meeting (activity), the organizer of the product has made a full demonstration of the major updates, and hope that through offline interaction and online live, to attract more users to use the product.

How to evaluate the effectiveness of activities?

How do you evaluate the effectiveness of this activity as a data analyst? For this case, it is possible to use cohort analysis to subdivide participants into unregistered (potential users) and registered (old-user) personnel in accordance with the participants ' registration, and some people register as new users after the activity is completed.

There may be a variety of sources for new users, and the most common source is a free trial, which allows users to track the behavior of their products using the source.

The old user is a product of the fans, is paid users, can bring benefits to the company, paid users may be lost, but also may be more dependent on the product, so that more frequent use of the product.

When analyzing user behavior, you should be aware that some users may not give you a direct benefit, but he can bring a lot of users, these users use the product, thereby indirectly to the company to bring benefits.

Design indicators

Thinking: According to the human behavior as the starting point to design indicators, from the quantitative data and qualitative data from the perspective of design indicators.

To analyze the benefits of the activity, the following key indicators can be developed:

1, participation degree

Participation measures how many people participate in the event to evaluate the impact of the event:

• Total number of people, number of old users, number of potential users
• How many users have registered (free or trial) after participating in the event?

2, indicators for observing the behavior of new users

Attendance of unregistered personnel, is a potential user of the product, regardless of the use of free trial, or purchase + gift and other promotional methods, as long as the user registration, you can think of a conversion, conversion rate can be used as an effective indicator of evaluation activity benefits.

• conversion ratio (Conversion rate): The proportion of registered new users after participating in the activity, the higher the conversion rate, the higher the product's attractiveness to the user, the better the benefit of the activity.

Analyzing the source and conversion paths of new users helps to determine which sources bring more effective conversions:

• Source of new users : free trial, purchase + gift, etc.
• conversion path (or conversion funnel) for new users: The conversion path for new users registered with free Trial (Trial, Activated Trial, Activated Trial in 7 days)

The stickiness of the new user to the product can be assessed by the retention rate of the new user:

• retention rate for new users : Percentage of users who are still active in the same period during the continuous billing cycle

3, indicators for observing the behavior of old users

The old user refers to the user who has registered before the event, this batch of users is a loyal fan of the product. By analyzing the behavior of old users, it can also be used as a way to evaluate the effectiveness of activities. How do you measure the benefits of hosting this event for older users? Loyalty can be measured as an indicator of the old user, loyalty refers to the user frequently use the product, the product provides a dependency on the service.

Old users are already paid users of the products, the amount of their production to maintain a level, if the activity can promote the use of products by older users more frequently, then the amount of old users will increase, dosage growth can be used as an indicator to evaluate the behavior of old users:

• Usage growth of old users: After participating in the activity, the usage of the old user in front of a pay cycle, compared to the increase is reduced?
• Percentage of users with increased usage: Percentage of users who have increased usage after participating in the activity

Consider some special cases, how many old users have been recalled, and how many old users have been lost:

• Zombie Account activation (New billed Customers): How many old users have left and re-paid for the product or service;
• Paid user churn (churned Customers): Users no longer pay to purchase products or services during a billing cycle.

4, indicators for observing the dosage

The dosage of the product is also an effective indicator of the effectiveness of the evaluation activity, in order to assess whether the increase in consumption is related to the activity, can be achieved by comparison of total consumption and activity-related dosage, which requires the addition of a dimension, which is the service type, the user uses the product this behavior, is actually a consumer product services.

Because the activities have specific keynote speeches, when the service is related to the topic of the lecture, it can be considered that the growth of the service dosage is related to the activity. If the use of the subject-related increase, the total amount of growth is relatively flat, then the activity has promoted the increase in dosage, if the use of the topic related to the growth of moderate, the total amount of growth is also flat, then the activity of the increase in dosage is not useful.

Compared with comparability, in the processing of the data, the impact of the new user on the data should be excluded, with the event time as the dividing point, only compared to the old users in the first n (1-3) Pay Cycle and the post N (1-3) The amount of the payment period. In order to calculate the dosage more accurately, it is necessary to avoid the influence of individual factors, which can be used in the mean value of 2-3 payment cycles.

Usage growth analysis for old users:

• Percentage of old users with increased usage
• What is the increase in consumption of old users and the percentage of growth after participating in the activity?
• Average usage per user

Usage growth analysis related to activity topics:

• Percentage of old users with increased usage (related to activity)
• What is the increase in consumption of old users and the percentage of growth (related to activity) after participating in the event?
• Average usage per user (activity-related)

Improvise Adjust indicator

It should be noted that data analysis should be based on the business, in accordance with the main problems, set up key data indicators to answer the challenges encountered in the decision-making, for example, when the product does not have visibility, site visits can be a key indicator; when there is a large number of user base, the number of registered (or trial number) as a key indicator When the number of registrations increases to a certain extent, conversion rates (that is, free users are converted to paid users) are key indicators.

The above indicators, is a data analysis of the re-disk, for reference only, attached to the website analysis of the data indicators commonly used, you can identify which is the vanity indicators?

1, website Traffic metrics

• Pageviews (page VIEW,PV), visitors are logged every time a page is opened.
• Number of independent visitors (unique VISITORS,UV): The number of unique visitors to the site within one day.
• Number of repeat visitors (Repeat visitors,rv): Repeat visits to visitors to the site within one day
• Page views per User: The average number of pages visited by each visitor is high, which means that the user is highly viscous, that is, the visitor shows interest in the site and is willing to stay for a long time.

2, User behavior indicators

• Bounce Rate: Indicates that the user is only browsing one page and leaves the site, the bounce rate shows how much visitors are interested in the site: the lower the bounce rate, the more interested visitors are to the site.
• Average length of visit: Refers to the duration of each visit, the larger the indicator, the longer the visitor stays on the webpage, the more interested in the website;
• Avg. Pages: Refers to the number of pages visited per visit, the larger the indicator, the more interested visitors to the site;

3, conversion indicators

• Conversions: Refers to the number of times a visitor arrives at the conversion target page, which means that the visitor does what the website manager wants the visitor to do, and is related to the promotion purpose that the website operator expects to achieve;
• Conversion rate: Refers to the efficiency of access conversion, conversion rate = number of conversions/visits, the higher the conversion rate, the better the promotion effect of the site.
• Conversion path: A path refers to a series of intermediate pages that a visitor passes before arriving at the target page you set, which means that the potential customer completes the action you expect, which is closely related to your promotion purpose and the definition of the promotion effect. By tracking the conversion path, you can see how visitors are accessing each step of the conversion.

4, retention indicators

• Number of retained persons: number of users who have continued to use the site for the next period after their first visit to the site
• Retention (Retention rate): The number of people who continue to use the site in the next cycle, the percentage of registered users in the current period, the higher the retention rate, the higher the user's stickiness to the site.

5, source Analysis

• Source: How visitors access websites, direct access, or search engines
• Search term Analysis: What search terms are used to find and access websites on various search engines

6, Visitor attribute analysis

• Age of the visitor
• Visitor's Area
• Old visitors/New visitors:

7, Other dimensions ...

Reference Documentation:

Benefit Data Analysis