After six months of effort, the product finally began to tend to be stable, the previous version has been exploring, needs often change, no time to bury the point system. With the steady and in-depth work product, increasingly recognize the importance of data, the data began to bury points related matters. This time witnessed the process of product (APP) data from scratch Buried, to share out for everyone to see a complete APP data Buried What is the process.
Next, these aspects will be set forth
First, what is the point of data buried
Second, the application specific data Buried
Step three, the data Buried
IV Notes
First, what is the point of data buried
Buried relevant personnel data such as product or can operate according to specific needs, customized to statistics more complex user data. For example, want to track user behavior, observing the relevant page click data, critical path conversion, campaign effectiveness analysis when an event, you need to advance Buried data, the corresponding data can be observed after the APP on the line, were analyzed.
Data points can be buried in their collection and statistical background, you can also use third-party data analysis platform, this time mainly on how to use third-party data analysis platform for data buried point.
Second, the application specific data Buried
Our company uses a Friends of the Union, so the next will be set forth as an example to Friends of the Union.
1. Statistics path
APP general all pages will be buried point, Friends of the Union will count users access path on the page after the buried point, the overall user behavior data can be obtained by accessing the path.
2. Custom Event
Custom event is a specific buried point, buried in APP point, what specific data are presented here. Custom events into the event counter and a computing event.
Count the number of events of major statistical events. The corresponding values calculated event type main statistical event.
For example, a purchase event, the statistics how many people buy success belongs to count events, statistics buy people buy success belongs to compute the distribution of the amount of events.
3. Event conversion rate
Commonly known as the funnel model, through the funnel model, you can see conversion events set for each step, each step of the assessment results.
Step three, the data Buried
1. Buried clear destination point, be Buried on demand.
Buried before to think about what the needs are and what the purpose is to achieve this goal, which requires statistical data to statistical data, which pages need to bury point? Buried in what position on the page, through what form Buried, statistics button clicks, or enter the number of pages?
比如这次上线了商城的功能，数据埋点其中一个目的想要统计购买过程的转化率，那么需要的是购买过程各个步骤的数据，整个购买流程涉及到的页面包括商品列表页、商品详情页、确认购买页、支付页、支付成功页。也就是需要对上面提到的页面进行埋点，统计进入到各个页面的数据。
2.与开发沟通讨论
梳理好要埋点的数据后，要多跟开发沟通，讨论埋点合理性与可行性，把埋点的目的跟开发描述清楚，一方面开发可以帮忙进行梳理，查缺补漏甚至提出更好的埋点思路；另一方面开发了解清楚后埋起点来更加胸有成竹，效率更快，防止出错。
3.开始进行埋点
使用第三方数据分析平台，在APP里埋点后，还需要在第三方平台上传相应的事件ID与事件名称，一定要代码中的ID与名称一致。ID与名称一般是产品这边整理命名，iOS 与Android统一。
4.漏斗模型
数据埋点完成后，如果要统计分析事件转化率，则需要提前添加漏斗模型，添加漏斗模型后第二天才会开始统计数据。
四、注意事项
1.如果想要整个APP全方位无死角进行埋点，工作量是非常庞大的，面对庞大的数据反而会造成干扰混乱，无从下手，所以在埋点前，一定要明确埋点的目的，不要为了埋点而埋点。要统计的数据庞大时，建议分阶段分版本进行埋点，先对主要事件关键路径进行埋点，一步一步完善。
2.埋好点后及时进行跟进，落实埋点的完整性与准确性。
3.测试的数据跟用户的真实数据没有可比性，如果有测试环境与正式环境，让开发注意不要把测试时数据也统计进去了。
4.不同第三方平台对于时间ID与名称可能会有不同限制，在命名时需要注意。友盟事件ID长度在128字符内，名称在32字符内。
5.每个点都有一个专属ID，ID之间的区分尽量明显点，这次就掉入了一个坑，有一个漏斗模型一直统计不到数据，分析了各种原因，找了好久，最后才发现原来第一个步骤选ID时由于有两个ID之间只相差一个字母，没有察觉选错了，后面的步骤也就没法统计到数据。
6. In fact, many of their own background data can be found, the statistics will be more flexible in the background, more targeted, and more detailed data collection, and may be easier to achieve. So keep up the development of multi-point communication before were buried, what the statistics look at third-party platform, which in their own statistical background, finding the optimal solution.
Written on the back
Buried End point is only the first step in the completion of data analysis, how to analyze the data collected, and found problems with the law, to guide product optimization is the key!
This paper consists of: @ second snow publish original in PMCAFF product manager community, reproduced Please keep this information.
Original Address: http: //www.pmcaff.com/article/index/929345781595264