How to improve product user stickiness?

Recently, more and more customers have been asking about user stickiness indicators, and DAU/MAU is the most frequently used indicator. The higher the ratio of daily active users to monthly active users, the higher the user stickiness to the APP .

DAU, namely: Daily Active User, refers to the number of daily active users;

MAU, or Monthly Active User, refers to the number of monthly active users.

From an extreme perspective, if the same users are active every day, for example, there are 10,000 DAUs every day, then these 10,000 users are active every day within 30 days, and the MAU is also 10,000, so the DAU/MAU is 100%, and the user stickiness reaches the upper limit. WeChat is an example of close to 100%.

Let's analyze another extreme example. If the daily active users are different, for example, 10,000 DAU per day, then the daily active users are different within 30 days, the MAU is 300,000, and the DAU/MAU is 1/30, and the users have no stickiness at all.

For common apps, the user stickiness ranges from 3% to 100%. Apps in different fields will also have different benchmark values. For example, mobile games will use 20% as the baseline, while tool apps will use 40% as the baseline.

In daily project and product operations , the daily DAU/MAU values ​​are often affected by the cycle (weekdays/weekends), version updates and activities, and user stickiness fluctuates greatly in the short term. Therefore, it is usually necessary to use the long-term average as a reference , such as one month or between two major versions.

Taking a certain customer APP as an example, the DAU/MAU value is about 50% on weekdays, while the value on weekends and holidays is around 20%. When talking about user stickiness, the daily average value of DAU/MAU is usually 39.37%, which is very close to the baseline value of 40%.

Further discussion of the algorithm:

The current mainstream algorithm uses yesterday's DAU and the MAU of the previous 30 days. For example, if today is August 31st, then the DAU is August 30th and the MAU is August 1st to 30th. The advantage of this algorithm is that DAU and MAU are calculated on the same day, which makes calculation easy. The disadvantage is that it only calculates the proportion of DAU on the last day of the complete 30-day period in MAU. Many rigorous customers ask why we cannot choose DAU on other dates?

For example, by taking the first day DAU1 of the complete cycle as the numerator and dividing it by MAU, we can get another interpretation of stickiness: the proportion of active users on any day within 30 days can be a definition of user stickiness, that is, DAU1/MAU, DAU2/MAU, DAU3/MAU...

Recalculating the DAU1/MAU stickiness of the above customers, the result is 39.41%, which is not much different from the traditional definition. After calculating DAU2/MAU, DAU2/MAU, etc. in the same way, the numerical results are not much different.

Further tracking, the value of each DAUn divided by MAU within a 30-day period also shows obvious fluctuations. The fluctuation trend here is consistent with the trend of 30 DAU (just divided by the same value)

Adding up the above DAU1/MAU to DAU30/MAU one by one, we get another commonly used indicator: the average number of active days per person in the month, which is also an indicator for evaluating user stickiness.

So how big is the correlation between average active days per person and DAU/MAU?

Since the average number of active days per person in a month is the sum of 30 DAUs divided by 1 MAU, and user stickiness is 1 DAU divided by MAU, the two will naturally differ in magnitude by about 30 times.

Comparing the average monthly active days and user stickiness DAU/MAU after dividing by 30, we found that their trends are basically the same. The difference lies in the slight lag in fluctuation amplitude and response time. From a numerical point of view, the average number of active days in a 30-day month is 11.93 days. After dividing by 30, the result is 0.3977, which is very close to the DAU/MAU of 39.37%. Therefore, using monthly average active days and DAU/MAU in work has the same effect.

Author: Analysys International, authorized to publish by Qinggua Media .

Source: Analysys International