https://www.cnblogs.com/yingyingja/p/10754794.html
Test Management: Using quantitative management thought
1. Problem
We first want to make a point, in the enterprise work, not to "benefit" as the goal of the work is bullying. Whether the daily organization of work, or a long-term technology planning, whether it is the production of products, or training of personnel in the enterprise, its ultimate goal should be "effective." The benefit, it must be a measure to quantify the way, otherwise we can not objectively and accurately measure its value.
How to quantify the work, is the actual problem in today's enterprises, many enterprises and enterprise managers include junior staff, do not have to quantify thoughts and ideas. This leads us often can see some ironic phenomenon at work.
- In Enterprise Performance Management, we often find themselves constantly "qualitative evaluation." For example, a performance indicator is the completion of the work. Then the work is done well or not? If the "well done" should be how to define?
- In the field of technology applications, many projects just for the purpose of the latest fashion, and technology upgrades, but it has no effective means to measure the value technology brings.
- In the personal scope of work, how to show their work performance and results to the leadership? We often tend to be personal report "remember running account", the lack of bright spots, but also unconvincing.
If a job you never find a way to quantify the value it generates, then we can say that the work had little value. Some people may question this view, they often say, for example drivers, cleaners or cooks, how to quantify their work, I asked them, what their main purpose is to work you will say:?! Ah that provide services on the right. Since it is a provision of services is their main objectives, it can be used to measure satisfaction index, which is not quantified yet?
For software testing is there are difficult to quantify the problem, test the most intuitive quantitative data is the amount of test execution and the number of defects. Both data can reflect some value, but can not objectively reflect all of the testing effort and value. But testing should be quantified, if not directly quantified, you should take the indirect quantification method converts indicators to achieve a quantitative measure.
2. The guiding ideology
We SMART principle as the guiding principle of our work to quantify.
SMART principles:
Specific- clear
Transaction goal must be clear, not too broad.
Such as: strengthening internal training to improve the team level? How to strengthen? How to improve?
Measurable- Measurable
There should be a clear set of data as a basis for measuring whether or not to achieve their goals. If there is no way to measure the goals, we can not judge whether this goal achieved.
Such as: through internal training so that staff skills level is improved?
Attainable- achievable
The goal is to be able to accept the debtor's current organizational skill level attainable.
Relevant- related
Correlation means to achieve objectives associated with this goal case with other objectives. If you achieve this goal, but totally unrelated to other targets, or very low correlation, even if that goal is reached, the significance is not great.
Time-Bounded- time-bound
Time-bound target property refers goal is to have a time limit.
It can be seen to follow the SMART principle, can make our business organization, planning more specific, and measurable. Transaction account in order to avoid blur, the results of qualitative determination flow.
3. Case
Case 1: quantitative targets Management
Enterprise management approach in a kind of popular affairs, OKR law.
OKR Work Act means Objective, Key Result, is widely used Google, LinkedIn and other top companies, both for large companies, but also for small teams.
To measure the results (KR-Key Result) staged by clear objectives (O-Object), using the key metrics. The key lies in quantifiable measure KR, otherwise management by objectives is just empty talk.
Let's compare the before and after quantization OKR:
Qualitative objectives:
O：高效的完成测试工作
KR：测试工作按计划完成，工作成果得到领导认可
Quantitative objectives:
O：高效的完成测试工作
KR1：A模块测试用例200条全部执行完成
KR2：缺陷泄漏量为0 KR3：工作进度偏离低于10% Case 2: Quantifying the benefits of technology
Our case to apply automated testing techniques.
Everyone talks about how technology can improve test automation test team productivity and solve difficult problems, the liberation of testers. However, how to determine the test automation technology can really achieve these goals?
We can use a formula to measure the quantifiable benefits of automated testing:
- 自动化收益=全手工测试成本–自动化首次成本–自动化维护成本
- 全手工测试成本=分析成本+设计成本+测试次数X执行成本 - 自动化首次成本=分析成本+设计成本+自动化搭建成本+编写成本+执行成本 - 自动化维护成本=维护次数X单次维护成本 Interpretation:
- The more the number of tests required, the higher the earnings automation
- Repeat testing of lower demand, the lower the automation benefits, and can be negative
- The more the number of automated maintenance, automation income lower
So automated testing can save costs, need full investigation, with the data speak.
Case 3: Resume quantify reflect
A lot of people's lack of biographical highlights, it is difficult to obtain the favor of the employer. Especially for senior engineers, automation test engineers, management personnel. How to make your resume stand out from the many job seekers? Quantify! It is a very good way.
We can use methods STAR (Situation, Task, Action, Result) to re-organize your resume. As follows:
Result in which the data portion of the project is the biggest highlight of this experience.
In summary, quantitative thinking should run through our work among the results of quantifiable, measurable process in order to better plan their work and reflect individual performance and value.
1. Problem
We first want to make a point, in the enterprise work, not to "benefit" as the goal of the work is bullying. Whether the daily organization of work, or a long-term technology planning, whether it is the production of products, or training of personnel in the enterprise, its ultimate goal should be "effective." The benefit, it must be a measure to quantify the way, otherwise we can not objectively and accurately measure its value.
How to quantify the work, is the actual problem in today's enterprises, many enterprises and enterprise managers include junior staff, do not have to quantify thoughts and ideas. This leads us often can see some ironic phenomenon at work.
- In Enterprise Performance Management, we often find themselves constantly "qualitative evaluation." For example, a performance indicator is the completion of the work. Then the work is done well or not? If the "well done" should be how to define?
- In the field of technology applications, many projects just for the purpose of the latest fashion, and technology upgrades, but it has no effective means to measure the value technology brings.
- In the personal scope of work, how to show their work performance and results to the leadership? We often tend to be personal report "remember running account", the lack of bright spots, but also unconvincing.
If a job you never find a way to quantify the value it generates, then we can say that the work had little value. Some people may question this view, they often say, for example drivers, cleaners or cooks, how to quantify their work, I asked them, what their main purpose is to work you will say:?! Ah that provide services on the right. Since it is a provision of services is their main objectives, it can be used to measure satisfaction index, which is not quantified yet?
For software testing is there are difficult to quantify the problem, test the most intuitive quantitative data is the amount of test execution and the number of defects. Both data can reflect some value, but can not objectively reflect all of the testing effort and value. But testing should be quantified, if not directly quantified, you should take the indirect quantification method converts indicators to achieve a quantitative measure.
2. The guiding ideology
我们将SMART原则做为指导我们量化工作的思想。
SMART原则：
Specific-明确的
事务目标必须明确，不能过于笼统。
如：加强内部培训，提高团队水平？如何加强？如何提高？
Measurable-可衡量
应该有一组明确的数据，作为衡量是否达成目标的依据。 如果制定的目标没有办法衡量，就无法判断这个目标是否实现。
如：通过内部培训，使人员技能水平得到提升？
Attainable-可达成
目标是要能够被执行人所接受的，当前组织技术水平可达到的。
Relevant-相关的
目标的相关性是指实现此目标与其他目标的关联情况。 如果实现了这个目标，但对其他的目标完全不相关，或者相关度很低，那这个目标即使被达到了，意义也不是很大。
Time-Bounded-有时限的
目标特性的时限性就是指目标是有时间限制的。
可以看到，遵循SMART原则，可以让我们的事务组织，工作规划更为明确，可度量。从而避免事务交代模糊，结果流于定性判断。
3. 案例
案例1：量化目标管理
企业中流行着一种事务管理办法，OKR法。
OKR工作法意为 Objective, Key Result，被广泛应用谷歌、领英等顶级公司，既适用于大型公司，也适用于小团队。
通过阶段性明确工作目标（O-Object），并使用关键性度量来衡量结果（KR-Key Result）。其中的关键就在于KR的可量化衡量，否则目标管理就是一句空话。
我们来对比一下量化之前和之后的OKR：
定性目标：
O：高效的完成测试工作
KR：测试工作按计划完成，工作成果得到领导认可
定量目标：
O：高效的完成测试工作
KR1：A模块测试用例200条全部执行完成
KR2：缺陷泄漏量为0 KR3：工作进度偏离低于10% 案例2：量化的技术效益
我们以自动化测试技术的应用为例。
大家都在谈论测试自动化技术如何能提升测试团队效率，解决难点问题，解放测试人员。然而，如何去判断测试自动化技术能否真的实现这些目标？
我们可以用量化的公式来衡量自动化测试的效益：
- 自动化收益=全手工测试成本–自动化首次成本–自动化维护成本
- 全手工测试成本=分析成本+设计成本+测试次数X执行成本 - 自动化首次成本=分析成本+设计成本+自动化搭建成本+编写成本+执行成本 - 自动化维护成本=维护次数X单次维护成本 解读：
- 所需测试次数越多，自动化收益越高
- 重复测试的需求越低，自动化收益越低，并且可以为负值
- 自动化维护次数越多，自动化收益越低
所以自动化测试能否节约成本，需要充分调研，用数据说话。
案例3：个人简历量化体现
很多人的个人简历缺乏亮点，难以获得用人单位的青睐。特别是对于资深的工程师，自动化测试工程师，管理人员而言。如何让简历能够从众多求职者中脱颖而出？ 量化！就是非常好的方法。
我们可以采用STAR（Situation, Task, Action, Result）的方法来重新组织个人简历。如下所示：
其中在Result部分的数据，就是这段项目经验的最大亮点。
综上，量化的思想应该贯穿于我们的工作当中，成果可量化，过程可量化，才能更好的规划自己的工作，并且体现个人业绩和价值。