提取Firebase / BigQuery DAU，WAU和MAU [英] Extracting Firebase / BigQuery DAUs, WAUs and MAUs

问题描述

我不想过于复杂化这个问题，所以我会尽量避免混淆。

我需要的结果是双重的。我希望
在以下内容中为移动应用程序确定DAU，WAU和MAU：a）Google Analytics以及b）Firebase Analytics。我想可以把更多的注意力放在b）上，因为他们正在转向架构和分析之间更紧密的整合，以便将来的应用程序开发。

在Google Analytics中计算DAU，WAU和MAU ：

目前GA / Firebase报告1,7，（14），30天活动用户（ https://support.google.com/analytics/answer/6171863?hl=zh_CN ）：
- 1天活动用户数：启动会话的唯一身份用户数
您的网站或应用在1月30日（日期范围的最后一天）。
- 7天活动用户数：1月24日至1月30日期间（您的日期范围的最后7日
日）在您的网站或应用上启动会话
的唯一身份用户数。
- 14天活动用户数：从1月17日至1月30日（日期范围的最后14天）从您的网站或应用发起会话的唯一
用户的数量。
- 30天活动用户数：1月1日至1月30日期间在您的网站或应用上启动会话的唯一用户数（日期范围内的整个30美元b b天）。

我的问题是：

• GA或Firebase报告1,730天活跃用户我的理解与DAU / WAU / MAU不同？或者这看起来完全一样？本文解释了另一种方法（使我相信WAU和7天活动用户不一定相同）使用Custom Dimensions来计算WAU： http://www.notingon.com/dau-mau-measurements-in-google-analytics/
  目前，我的做法是选择例如9月1日至30日，并将30天活跃用户等同于该月的MAU，7天活跃用户等同于WAU，以及1天活跃用户等同于DAU，我想知道这是否是正确的方法，或者我应该使用自定义维度进行自定义计算以获得DAU / WAU / MAU？
  

>

计算Firebase中的DAU，WAU和MAU

Firebase控制台中的1天，7天和30天活动用户，位于GA： https://support.google.com/firebase/answer/6317517#active-users 。看起来，如果有人想计算其他任何东西，你将不得不通过Blaze程序设置Google BigQuery？ （ https://firebase.google.com/pricing/ ）。

我偶然发现了计算1/7/30日活动次数的两个例子，但我仍然认为这与DAU，WAU和MAU不同：
Firebase - > BigQuery如何获得当月，周，日的活跃用户
活动用户指标的差异在Firebase Analytics仪表板和BigQuery导出之间

我的问题是：

• 我应该区分计算（1）1/7/30 Day Actives与（2）DAU，WAU和MAU吗？如果是，与上面列出的两个示例相比，查询看起来如何？我需要在流程中应用唯一ID /自定义维度吗？


• 在GA中，可以设置用户ID无论是在网络还是移动应用视图中，并将它们结合在一起，但Firebase Analytics中的这种方式会如何解决？是否可以将此唯一ID设置为用作自定义维度以构建自定义查询？或者是已经传递的唯一设备ID？


• 从BigQuery提取数据时会有数据延迟吗？

非常感谢！ D $ / $>

解决方案

GA或Firebase报告1,730天活动用户， b $ b的理解与DAU / WAU / MAU不同？
我应该区分计算（1）1/7/30日活动
与（2）DAU，WAU和MAU吗？

虽然概念相似，但这些指标在GA和Firebase Analytics中具有不同的语义。在GA中，活动用户是指在特定日期与您的应用启动会话的计算机，其计算取决于应用的显式检测（即开发人员必须手动记录点击）。在Firebase Analytics中，活动用户是在给定日期记录user_engagement事件的人员。当应用程序花费时间在设备的前台时，用户参与事件会自动记录。因此，Firebase Analytics中的活跃用户是在前台与应用进行互动的人。 Google Analytics中的活跃用户是开发者发送匹配的活动用户。

在GA中，您可以在网页和手机上设置用户ID应用程序视图并将
绑定在一起，但在Firebase Analytics中如何处理此问题？

您可以调用Firebase的setUserID方法为该用户赋予一个ID，并且您可以通过该ID删除活动用户。或者，app_instance_id从Firebase和（可选）传递到BigQuery，广告标识也是如此。请参阅Firebase Analytics BigQuery架构此处。

是否可以将此唯一ID设置为
自定义维度来构建自定义查询？或者是一个唯一的设备ID
已被传递？

是的。自定义用户标识在架构中记录为字段user_dim.user_id。

从BigQuery中提取此数据时会有数据延迟？

数据每天从Firebase导出到BigQuery。有些数据来自设备迟到（例如，如果设备在记录事件时最初处于离线状态），并且该数据会在随后的几天内发送。

I don’t want to over complicate this question, so I will try to ask it as clear as possible to avoid confusion.

The outcome I require is two-fold. I want to determine the DAUs, WAUs, and MAUs for a Mobile App within: a) Google Analytics, as well as in b) Firebase Analytics. I guess one can focus more attention on b) because their is a shift towards tighter integration between Architecture and Analytics for future app development.

a) Calculating DAUs, WAUs, and MAUs in Google Analytics:

Currently GA/Firebase reports on 1,7,(14),30 Day Active Users (https://support.google.com/analytics/answer/6171863?hl=en): - 1-Day Active Users: the number of unique users who initiated sessions on your site or app on January 30 (the last day of your date range). - 7-Day Active Users: the number of unique users who initiated sessions on your site or app from January 24 through January 30 (the last 7 days of your date range). - 14-Day Active Users: the number of unique users who initiated sessions on your site or app from January 17 through January 30 (the last 14 days of your date range). - 30-Day Active Users: the number of unique users who initiated sessions on your site or app from January 1 through January 30 (the entire 30 days of your date range).

My questions are:

• GA or Firebase reports on 1,7,30 Day Active Users which by my understanding is not the same as DAUs/WAUs/MAUs? or is this seen as exactly the same ? An alternative approach (which led me to believe that i.e. WAUs and 7 Day Active Users are not necessarily the same) was explained in this article which used Custom Dimensions to compute WAUs: http://www.notingon.com/dau-mau-measurements-in-google-analytics/ Currently, my approach would be to select for instance 1 - 30 September, and take the "30 Day Active Users" as equal to the MAUs for that month, "7 Day Active Users" as equal to the WAUs, and "1 Day Active Users" as equal to DAUs. I want to know if this would be the correct approach or should I apply Custom Dimensions to do a custom calculation to get to DAUs/WAUs/MAUs ?

b) Calculating DAUs, WAUs, and MAUs in Firebase:

We see the same 1-day, 7-day, and 30-day active users in the Firebase console, that is in GA: https://support.google.com/firebase/answer/6317517#active-users. It seems that if one wants to calculate anything else, you would have to setup Google BigQuery through the Blaze program ? (https://firebase.google.com/pricing/).

I stumbled unto 2 examples that computes the 1/7/30 Day Actives, but I would still see this as different from DAUs, WAUs and MAUs: Firebase -> BigQuery how to get active users for that month, week, day Discrepancies on "active users metric" between Firebase Analytics dashboard and BigQuery export

My questions are:

• Should I have a distinction between calculating (1) 1/7/30 Day Actives vs (2) DAUs, WAUs, and MAUs ? If yes, how would a query look like compared to the 2 examples listed above, and would I have to apply Unique IDs / Custom Dimensions in the process ?
• In GA one can set User IDs on both a web and mobile app view and tie them together, but how would one approach this in Firebase Analytics ? Would it be possible to also set up this unique ID to be used as a custom dimension to build a custom query ? or is a unique device ID already being passed ?
• Would there be a data delay when pulling this data from BigQuery ?

Thanks in Advance ! D

解决方案

GA or Firebase reports on 1,7,30 Day Active Users which by my understanding is not the same as DAUs/WAUs/MAUs? Should I have a distinction between calculating (1) 1/7/30 Day Actives vs (2) DAUs, WAUs, and MAUs ?

Although the concepts are similar, these metrics have different semantics in GA and Firebase Analytics. In GA, an Active User is one who initiates a session with your app on a given day, and its calculation is dependent on explicit instrumentation of the app (i.e. the developer must log hits manually). In Firebase Analytics, an Active User is one who logs user_engagement events on a given day. User Engagement events are logged automatically when the app spends time in the foreground of a device. And so, an Active User in Firebase Analytics is one who engages with the app in the foreground. An Active User in Google Analytics is one for which the developer sends hits.

In GA one can set User IDs on both a web and mobile app view and tie them together, but how would one approach this in Firebase Analytics ?

You can call Firebase's setUserID method to ascribe an ID to that user and you can dedupe your active users by that ID. Alternatively, app_instance_id is passed in to BigQuery from Firebase and (optionally) so are advertising identifiers. Refer to the Firebase Analytics BigQuery schema here.

Would it be possible to also set up this unique ID to be used as a custom dimension to build a custom query ? or is a unique device ID already being passed ?

Yes. The custom user ID is documented in the schema as the field user_dim.user_id.

Would there be a data delay when pulling this data from BigQuery ?

Data is exported from Firebase to BigQuery on a daily basis. Some data arrives late from devices (for example, if a device was offline initially when events were logged) and that data is then sent over on subsequent days.

这篇关于提取Firebase / BigQuery DAU，WAU和MAU的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！