Buried point data completeness check and result analysis

1. Data verification function entrance

Entrance: Data Management - Data Quality - Data Verificationinsert image description here

2. Operation steps

2.1 Statistical time interval

Choose according to your own needs. Generally, the time after burying a serious bug fix is ​​selected to avoid too much dirty data from affecting the analysis results.

2.2 Data Sampling

Choose according to your own needs. The full amount of data is large, and the analysis will be more complicated, but the result is more accurate. Partial sampling can be selected according to your own needs. It is recommended to select it together with the statistical time interval. If the time interval is short, full or 1/4 sampling can be selected; if the time interval is long, a small part of data sampling can be selected for verification.
insert image description here

2.2 Upload collection plan

The project will have a requirements collection plan (formulated according to the "Data Collection Requirements Template"), but there are generally some other irrelevant information during the development and testing process. Therefore, please check carefully before uploading. You can refer to the "Data Collection Requirements Template" to ensure that unnecessary information is deleted. What is included in the template? See previously uploaded resources for details. What should be paid attention to when collecting required content? See Chapter 3 for details

2.3 Start verification

Click to start the verification, and the system will soon generate an excel result file. QA can analyze the result. If there is any data quality problem, contact the development to promote the problem solving and verification.

3. What should be included in the "Data Collection Requirements"?

The conceptual "Buried Points" column has been mentioned in previous articles, which can be used for reference. The following only explains the points to be paid attention to to avoid stepping on the pit.

3.1 Description

No need to change, just follow the template content

3.2 Public properties

Just replace the following parts, copy and paste the public property sheet in the project. First check that the content is complete.
Focus on checking whether the attribute value and data type meet expectations, otherwise the analysis result will report an error, which will be mentioned later.
insert image description here

3.3 Custom event table

You only need to replace the following parts, copy and paste the custom event table in the project. First check that the content is complete . In order to avoid stepping on the pit, it is best to be strictly consistent with the template format.

Note:
1. The format is strictly consistent with the template.
2. Only paste the custom events that need to be analyzed, not all of them.
3. Don’t worry about the order of events, but the event code must be filled in.
4. Format issues such as merging cells
insert image description here

3.4 Event table of all buried points

Most of the preset events have been filled in the template, so there is no need to change them. If this form does not include the ones to be verified, you can add them according to your needs, and the format should be consistent with that in the template.

3.5 User table

Only need to replace the following part, copy and paste the user table in the project. First check that the content is complete.
Focus on checking whether the attribute value and data type meet expectations, otherwise the analysis result will report an error, which will be mentioned later.
insert image description here

3.6 Preset properties

No need to change, just follow the template content

4. What have been verified?

4.1 Event Integrity Check

For the preset events and custom events in the collection plan, check
only what is missing and what is added.

差异说明
缺少:事件设计中存在,但实际环境中没有的事件
新增:事件设计中没有,但实际环境中存在的事件

4.1.1 Analysis method

引起差异原因说明
1、事件缺少原因可能因为:数据未采集;数据采集但该行为没有被触发过;数据未成功入库;事件被隐藏等
2、事件新增原因可能因为:新增事件;事件英文名和事件设计不符;错误数据未隐藏等

关注点:
1、重点关注自定义事件的新增和缺少,尤其关注缺少。
2、缺少时

Guess you like

Origin blog.csdn.net/weixin_44691253/article/details/131808896