Build a super detailed AARRR model data indicator system in 5 minutes

The AARRR model is our classic model for user analysis and is a typical funnel structure. From the perspective of the life cycle, it describes the five steps that users need to go through to enter the platform and ultimately obtain commercial value. Value comes not only directly from user purchasing behavior (earning revenue), but also from the revenue generated by users as recommenders (self-propagation) and content producers (retention rate).

01

What is the AARRR model

AARRR represents five words, which correspond to the five stages in the product life cycle:

Acquisition: How do users discover (and come to) your product?

Activation: What is the user’s first-time experience like?

Retention: Do users return to the product (repeat use)?

Revenue: How does the product make money (through users)?

Refer: Is the user willing to tell other users?

The AARRR model points out two core points:

Focus on users and take the complete user life cycle as the clue;

Control the overall cost/revenue relationship of the product. The success of the product means that the user lifetime value (LTV) is much greater than the sum of the user acquisition cost (CAC) and the user operating cost (COC).

02

AARRR data metrics

1. Acquisition

The acquisition stage is the product promotion stage and the first step in product operation.

Operators acquire target users through various promotion channels and in various ways, and evaluate the effectiveness of various marketing channels, so as to more reasonably determine investment strategies and minimize user acquisition costs (CAC).

The main indicators to pay attention to at this time are:

(1) Daily number of new users (DNU)

Definition: The number of users who register and log into the game every day.

Registration here is a broad concept. For some apps, it refers to the user who launches the app for the first time, so the definition of DNU can also be the user who logs in or launches the app for the first time. It should be noted that in mobile statistics, users sometimes also specifically refer to devices.

Solving the problem:

· User share contributed by the channel.

· Macro trends, determine the delivery strategy.

Whether there are a large number of spam users.

·Registration conversion rate analysis.

2. Increase activation

New users are converted into active users after sedimentation. At this time, we need to pay attention to the number of active users and the data on user usage frequency and stay time.

(1) Daily Active Users (DAU)

Definition: The number of users who logged into the game every day.

For some apps, launching the app means you are an active user, while for others, you need to register an account and form an online account to be considered an active user. The calculation of active users is deduplicated.

Solving the problem:

·Core user scale.

Product life cycle analysis.

Active users of the product are lost and active users are decomposed.

User activity rate: active users are counted as the number of users.

(2) Weekly Active Users (WAU)

Definition: The number of users who have logged into the app in the last 7 days (including the current day), usually calculated based on a natural week.

Solving the problem:

Periodic user scale.

· Cyclical change trends, mainly comparisons between promotion period and non-promotion period.

(3) Monthly Active Users (MAU)

Definition: The number of users who have logged into the app within the last 30 days (including the current day) of the month, usually calculated based on the calendar month.

The fluctuation range of MAU is relatively small. In terms of the stability of product user scale, MAU is a weather vane. However, during the promotion period, the impact of version updates and adjustments to operational activities on MAU is more obvious.

In addition, the trend of MAU changes differently at different stages of the product life cycle.

Solving the problem:

Stability of user scale.

· Promotion effect evaluation.

Changes in overall user size.

(4) Average daily usage time (DAOT)

Definition: Daily total online time/Daily active users.

Regarding usage time, it can be divided into indicators such as single usage time, daily usage time and weekly usage time. By performing interval distribution and average calculation on these indicators, we can understand the participation stickiness.

Solving the problem:

Analyze product quality issues.

Observe the average usage time in different time dimensions to understand the habits of different user groups.

One of the channel quality measurement criteria.

Retention is the basis for churn analysis.

(5) DAU/MAU

DAU/MAU can show the average number of days a user visits the App each month. For example, if an App has 500,000 DAU and 1 million MAU, its DAU/MAU ratio is 0.5, which means that the average monthly visit time for users is 30*0.5=15 days. This is also a relatively important indicator for evaluating user stickiness.

DAU/MAU ranges from 3.33% to 100%, but it is obvious that these two situations are basically impossible to occur in reality. Apps in different fields will have different benchmark values ​​for reference. For example, mobile games will use 20% as the baseline. The values ​​of Honor of Kings in June and September 2017 were basically around 31%. The baseline for tool apps is 40%.

The higher the DAU/MAU value, the stronger the stickiness of the App, which means that more users are willing to use the App. On the contrary, if the DAU/MAU value is very low, it cannot be directly said that the App is a failure. We also need to conduct a multi-dimensional analysis based on multiple conditions, such as product attributes (for example, apps for regular financial management/job hunting/house buying/renting may have a relatively low DAU by nature), time considerations (weekdays/holidays, etc.), version updates, operational activities, and ARPU values ​​in the user dimension, before we can draw conclusions. Therefore, it is important to correctly understand the meaning of DAU/MAU.

3. Retention

After solving the problem of activity, another problem was discovered: "Users come and go quickly." Sometimes we also say that the game lacks user stickiness or retention.

We need indicators that can be used to measure user stickiness and quality. This is a means of judging whether an APP can retain users and increase the number of active users in the early stages. Retention rate is one of the means.

Retention rate: The number of new users in a certain period of time is recorded as A. After a period of time, the proportion of users who are still using the app to the new users A is the retention rate.

(1) Day 1 Retention Ratio

Definition: Daily new users is the ratio of the number of users who logged in on the +1 day to the number of new users.

(2) Day 3 Retention Ratio

Definition: Daily new users are the ratio of users who logged in on the +3rd day to the new users.

(3) Day 7 Retention Ratio

Definition: The ratio of daily new users who logged in on day +7 to the number of new users.

Retention rate has gradually evolved into an important criterion for judging product quality.

While paying attention to retention rate, we should also pay attention to the analysis of churn rate. The retention rate is more concerned with a comprehensive analysis of whether the channels for acquiring users are reasonable from the perspective of user acquisition and whether the user scale of the product can grow. The churn rate is concerned with why some users leave the app. This may be a problem that exists in the user acquisition stage, but when the app has a stable user base, the loss of a paying user may cause a significant decline in the app's revenue.

The calculation of retention rate can be determined according to the statistical time period. For example, when calculating weekly retention, the weekly retention of new users is calculated by the total new users in one week and the retention in the subsequent weeks.

The +3 days or +7 days mentioned above are intended to emphasize the concepts of the 3rd day and the 7th day. Please note that when calculating the retention rate, the day of addition is not included in the number of days. In other words, the retained users we mention refer to the retention of new users on the 1st, 3rd, and 7th days after they are added.

Solving the problem:

APP quality assessment.

User quality assessment.

User scale measurement.

Churn: The number of users who leave the app at different times during the statistical time interval.

(4) Day 1 Churn Ratio

Definition: The proportion of users who logged into the app on the statistical day but did not log into the app in the following 7 days to the active users on the statistical day.

(5) Week Churn Ratio

Definition: The proportion of users who logged into the app last week but not this week to the weekly active users of last week.

(6) Month Churn Ratio

Definition: The proportion of users who logged into the app last month but not this month to the monthly active users of last month.

The churn rate is an indicator that needs to be paid special attention to when the APP enters a stable period. If paying attention to retention means paying attention to the situation of APP users entering the APP in the early stage, then paying attention to the churn rate means caring about the user stability and profitability conversion of the product in the mid- and late stages of the product. During the stable period, the revenue and activity are very stable. If there is a large churn rate, this indicator needs to serve as a warning, and gradually find out which users have left the APP and where the problem lies. In particular, the analysis of paid user churn requires special attention.

Solving the problem:

Active user life cycle analysis.

· Changes in channels.

Operational means to increase revenue and evaluate the impact of version updates on user churn.

During what period is the churn rate higher?

·Industry comparison and mid-term product evaluation.

4. Revenue

There are many sources of revenue, including: app payment, in-app feature payment, advertising revenue, traffic monetization, etc. The main evaluation indicators include ARPU (average order value).

Main concerns:

(1) Payment rate (PR or PUR)

Definition: The ratio of paying users to active users.

In layman's terms, the payment rate is also called the payment penetration rate. In the mobile APP market, most people only care about the daily payment rate, that is, the Daily Payment Ratio.

The high or low payment rate does not mean that the product’s paying users will increase or decrease. The payment rate also varies among products of different APP types.

Solving the problem:

·Product’s profit conversion capability standard.

User payment key points and conversion cycle.

·Evaluation of paid conversion effect.

(2) Active paying users (APA)

Definition: The number of users who successfully paid during the statistical time interval. It is usually calculated on a monthly basis and is also called MPU (Monthly Paying Users) in the international market.

In data analysis, we pay more attention to daily paying users and weekly paying users. The main reason is that the user life cycle is short and short-term payment becomes the focus.

The calculation formula for the number of active paying users is as follows:

APA＝MAU×MPR

Solving the problem:

The number of paying users of the product.

The composition of APA, the proportion of whale users, dolphin users, and small fish users, and their profitability.

The value of the paying group is the overall stability analysis.

(3) Average revenue per user (ARPU)

Definition: The average revenue generated by active users during the statistical period. Usually measured in months.

The formula for calculating average revenue per user is as follows:

ARPU = Revenue/User

Monthly ARPU＝Revenue/MAU

It is total revenue divided by total active users, usually calculated on a monthly basis.

The strictly defined ARPU is different from the ARPU known in China. The domestic ARPU = total revenue/number of paying users. Therefore, paid ARPU is often emphasized, and there is a special term here called ARPPU.

ARPU is used to estimate revenue at different scales in the early stages of product positioning and is also an important reference for LTV.

Solving the problem:

·Judgment of user quality in different channels.

Product revenue contribution analysis.

The relationship between average revenue per active user and advertising cost.

(4) Average revenue per paying user (ARPPU)

Definition: The average revenue generated by paying users during the statistical period. Usually measured in months.

The formula for calculating average revenue per paying user is as follows:

ARPPU = Revenue/Payment User

Monthly ARPPU＝Revenue/APA

It is total revenue divided by the total number of paying users, usually calculated on a monthly basis.

ARPPU is easily affected by whale users and small fish users, so you need to be cautious when analyzing it.

The combination of ARPPU, APA and MPR can analyze the retention of paying users, conduct in-depth analysis of the loss of specific paying groups, and ensure the quality and scale of payment.

Solving the problem:

·The paying ability and gradient changes of paying users.

The overall payment trends of paying users and the differences among different paying tiers.

·Explore the value of whale users.

（5）Lifetime Value (LTV)

Definition: The total revenue generated by a user during his or her lifetime. It can be seen as a long-term accumulated ARPU.

The average LTV for each user is calculated as follows:

LTV = ARPU × LT (average life cycle calculated in months or days)

Among them, LT is Life Time, that is, the time from the first time a user launches the APP to the last time. Generally, the average value is calculated. LT is measured in months, which is the average number of months a user stays in the APP. For example, if the ARPU of an APP is RMB 2 and LT is RMB 5, then LTV is 2×5= RMB 10.

The above calculation method is feasible in theory. In practice, we adopt the following LTV calculation method.

Track the number of new users on a certain day or week, calculate the cumulative revenue contribution of this group of users in the following 7 days, 14 days, and 30 days, and then divide it by the number of new users in this group, that is, cumulative revenue/new users = cumulative ARPU (LTV). This method can calculate the rough lifetime value of this batch of new users at different life cycle stages.

Solving the problem:

User revenue contribution cycle.

Profit contribution of user groups and channels, measurement of LTV and CPA.

LTV does not distinguish between paying and non-paying users, but looks at the overall value.

5. Refer

Self-propagation is also called word-of-mouth or viral communication. One of the important indicators is the K factor.

The calculation formula of K factor is not complicated, the process is as follows:

K = (the number of invitations each user sends to his friends) × (the conversion rate of people receiving invitations into new users).

Assuming that on average each user will invite 20 friends and the average conversion rate is 10%, then K = 20 × 10% = 2.

When K>1, the user base will grow like a snowball.

When K is less than 1, the user base will stop growing through self-propagation when it reaches a certain size.

Most apps cannot rely solely on self-propagation, but must be combined with other marketing methods. However, it is still necessary to add functions that are conducive to self-propagation during the product design stage. After all, this free promotion method can partially reduce CAC.

The above is the data indicator system of the AARRR model. Only by establishing a complete data indicator system can we conduct a more comprehensive analysis of user behavior in the future.

Author: From the South

Source: From Nanji