Re-study of user-paid penetration rate

"" Reader

Paid penetration Rate

What we are talking about today is actually about the content of the new pay user research model, when we talk about models, sometimes we are too myth, the model finally is a set of methodologies, I think this is their own thinking thinking of the final landing of a carrier, only through the process is a model.

Today's content is a further study of the rate of user-paid penetration, to be sure, our previous research on the pay-for-user pyramid included all paid users, and our previous approach was to measure users ' contribution or value, dividing users into whale users, dolphin users, and small fish users. But today we will analyze this problem from the perspective of the user's life cycle, and then draw a further study of the pay penetration rate.

composition of paid users

Paid users are a very complex group, the first level is common to us, and we often use the level of data analysis, but from the beginning of the second layer of our many subsequent analysis is very useful. I have a guess: if the user pyramid is stable, then the increase in the rate of pay penetration is necessarily meaningful.

In our study of different pay groups, we found that the characteristics of group users have been formed in the first stage, in other words, we speculate that a user's ability to pay for a game is basically a delineation. Of course, many people will have doubts about this, because many games are paid through the game "trap", stickiness and extended consumption to further expand demand, stimulate consumption. There is indeed a point in this. However, if you analyze the data carefully, you will find that many players ' consumption is basically within their own range and under pressure.

We do not rule out extreme users, such as deep obsession with the game so that users are fully engaged in the game, but the proportion of such users is very small. From this point of view, everyone's ability to pay is basically fixed (want to extend and stimulate consumption, you have to update, operate), then we continue to pull high permeability in fact, not much effect. If your game is worth the money they spend, people who have paid very little will spend very little or even lose in the end. Too many games, too many choices, too many temptations. In this sense, the permeability is limited.

So in this case, we can do one thing, that is, in the early stages of paying users to predict and judge the pay user's ability, rather than through late data to verify which is the real whales, which are dolphins, which are small fish. This also reflects the value of data analysis, using the past to find the future, rather than the future validation of the past.

The structure of the paid penetration rate

The so-called structured, is layering to establish a pay penetration rate, because we've built a pyramid model for paid-user research, and the way we used a pay-for-penetration metric in the past needs to be further refined, which is not to say that the original approach is wrong, because in some high-level analyses and speeches, We need one of these indicators to be OK.

However, as an analyst, in the specific face of the business, we can not be such extensive use of a pay penetration rate to analyze the problem, because it will cover up a lot of problems. So the structure I propose is tiered pay permeability:

W-pur: Paid penetration rate for whale users

D-pur: The paid penetration rate of dolphin users

F-pur: Paid penetration rate for small fish users

There may be a problem here, it is estimated that we all have this question, how do we calculate this pur? The calculation method is as follows

Number of whale users/active users

Here it needs to be explained that whale users here are based on historical whale user characteristics calculated by this month's whale users, which are themselves a predictive data, but must be paid users, active users that are Mau.

serialization of paid penetration rates

Retention Rate I think we are all familiar with, for example, the next day, 3rd, 7th, 30th, this is a way of time serialization, which I think for the rate of pay permeability we can also do time serialization.

That is, the first day, the next, 7th, 30th paid permeability research, but it is clear that the users here are the added users. Its definition is as follows:

N-Day paid penetration rate

New users within a limited time, N-day-paid user/limited time new users

Assuming October 8 500 new users, the first day of 50 people pay, then the first day of pay rate for 50/500=10%;

Assuming that October 8 has 500 new users, October 9 (that is, the next day) there are 25 people pay, the next day pay rate for 25/500=5%;

The rate of payment in this way is a separate and three-dimensional breakdown of the paid penetration rate we've previously counted. The pay-for-penetration breakdown is a clear way to pay for new users and active users, as some new users start paying for the first day, while some new users choose to pay for a certain period of time. But the standard for active users is not up to par. This also helps us to study the natural pay cycle of active users in more detail.

The above is for a specific daily rate of pay penetration analysis, of course, like the retention rate study, we can limit the time to weeks, that is, the week of new users in the next week of the pay permeability study, which is feasible. It depends on what you need.

Here is a pay-for-penetration study based on the retention rate model, the method is essentially the same as before, a little change, as to whether the method meets your product needs and analysis needs, according to their actual situation, the content described here is for reference only, as for exploration and discussion.