Do you really need automated testing?

A few days ago in the technical exchange group , Some students threw a mind map about the implementation of automated testing , Ask if you have any good suggestions .

The mind map is as follows ：

After reading it, I asked some questions ：

• What is the current pain point of the team ;

• What are the possible solutions ;

• Automation is not the best solution ;

• Implement automation , Whether research and evaluation have been conducted within the team ;

Many students do automated tests , The common question is what I should do , With what frame , Separation of script and data , Fall into a familiar dependency path , But lack of thinking .

This article , I would like to talk about before the automated test is implemented , Some thoughts and personal opinions on whether automation is needed and when to do it .

Do you want to do automated testing ？

In fact, the concept of automation has been around for a long time , At first, it was mainly used in the field of industrial production , It refers to the control of mechanical tools through instructions or software to complete some highly repetitive work . Later, with the development of computer technology and the booming trend of the Internet industry , It has been applied in the field of software development and testing .

Whether we are familiar with various automated tests such as UI automation 、 Interface automation or even unit automation testing CICD, Its essence is to help us complete our daily work with the help of tools , Manual and time-consuming work .

The advantage of automated testing lies in the fact that software tools are entrusted with tasks that are highly repetitive and time-consuming , Free up human resources to do other more creative work , Increase the ratio of input to output , Create more value with less cost .

Do you want to do automated testing in your work ？ The answer is yes .

But the new question is coming , automated testing = Applicable to any team at any stage ？ Does not necessarily .

When to do automated testing ？

We all know , software test （ Quality assurance ） In fact, the pursuit is 2 Goals ： Quality and efficiency .

In essence , Automated testing is an auxiliary means to improve efficiency , Not the purpose of solving the problem , And it is not the only way to improve efficiency .

When to conduct automated testing at work , How to develop , What type of automated testing do you do , Whether there are enough resources to invest , They all need to be carefully investigated and evaluated , Not to automate for the sake of automation , This is tantamount to discarding the basics .

Our daily testing work is guided by the methodology of software engineering , Follow certain process specifications . Methodology of software engineering , In fact, it is essentially three words ：

cost + earnings + risk . That is, try to use lower cost investment to obtain higher income , And the risks undertaken are controllable

.

You can't have all three , A certain amount of balance and sacrifice is required , To achieve the ultimate goal of quality and efficiency .

So before evaluating whether to conduct automated testing , It is necessary to consider the above factors as much as possible .

for instance ：

A startup company , At present, it is in the period of product start-up and rapid iteration , The pursuit is to launch quickly MVP The product is on the market , Business is unstable , Shortage of human resources , The technical infrastructure is poor , At that time, it is not suitable to carry out automated testing .

As mentioned earlier , 

Automated testing is suitable for jobs with high repeatability , And it is not easy to solve the problem by plug and play . Relatively stable iteration of business requirements is required 、 Relatively mature and stable R & D team and certain technical infrastructure construction , And a more standardized process can be better implemented , Achieve the purpose of improving efficiency

.

So if we want to implement automated testing and achieve the purpose of improving efficiency , What are the factors to consider ？

What to think about before landing automatic test

In terms of my own practical experience , Before the landing automation test , Some of the following factors need to be considered ：

• What are the pain points we are facing now ？

• What are the reasons behind the pain points ？

• What can be done to solve the problem ？

• Automation is not the best solution ？

• Whether the current situation is suitable for automated testing ？

• What factors should be investigated and evaluated before conducting automated testing ？

• Select the pilot scope ;

• Automated tool research ;

• Team member technology stack matching ;

• How much human time and resources should be invested ;

• What is the expected input-output ratio ;

Just like a classmate in the group said ：“ I had a meeting with the business department yesterday , Some department leaders pointed out that there is still a big gap between the full link voltage testing platform and the commercial platform , It needs good construction . I said, , You basically don't need full link voltage measurement , I put a lot of energy into it , But nobody uses it ”.

The essence of an enterprise is to pursue greater value , In fact, I don't care about the technical means . Automation is one of the auxiliary means to improve efficiency , Not the purpose of doing things .

Don't automate for the sake of automation , When I see a nail, I want to find a hammer to hit it , Compared to what to do , What needs to be considered more is the reason for doing this and the value it brings .

This article , What I'm talking about is a way of thinking , Rather than a specific practice path .

Next article , I will talk about how to implement automated testing in specific work , Coming soon .