How should the KPIs of technical staff be set?

Let’s read a set of comics first. The author Xi Qiao (ID: coderstory) feels sorry for the HR of this company.

Thanks to Zhu Feng for providing the source of inspiration. Thanks to Mr. Huo for writing the code for all scenes. If you are not good at writing, please contact him.

Complained about HR

Let’s continue to discuss how the KPIs for technical staff should be set.

Alibaba Technology has an article "How to Quantitatively Assess Technical Personnel's KPIs". The author is Zhang Jianfei. He believes that technical personnel's KPIs can be broken down into "three contributions": business contribution, technical contribution and team contribution:

1. Business contribution: including demand control, business projects and business innovation.

2. Technical contribution: including design refactoring, technical influence, Code Review, innovation efficiency and code quality.

3. Team contribution: including recruitment, training of new employees and team atmosphere.

Focuses on the second "technical contribution". The author uses some cases at work to describe:

1. Application quality:

Responsible or ***The application quality score that the same person is responsible for (can be examined from the dimensions of code duplication rate, cyclomatic complexity, layering rationality, etc.), and what work the technical staff has done to improve the application quality score.

2. Design reconstruction:

In the customer communication project, the technical staff conducted domain modeling and design of the CRM sales domain, and the abstraction was reasonable;

Found that the package classification in Infrastructure was unreasonable, redesigned and refactored

Found that the error codes in the current system are confusing, sorted out new error code specifications, and completed the code reconstruction.

3. Technical influence:

The team shared strategic models and was well received by classmates.

Accepted the invitation and shared the SOFA architecture at industry conferences.

4. Code Review:

When reviewing a certain code, I found that there may be thread insecurity risks.

When reviewing a certain code, I found that the design was unreasonable. The use of the responsibility chain here can solve the problem very elegantly and has certain scalability.

5. Innovation and efficiency improvement:

I found that starting Pandora Boot for local testing was a waste of time, so I wrote a Test Container to greatly improve the efficiency of self-testing.

I found that some boilerplate code did not need to be written, so I added annotation support for optimistic locking and paging to simplify the code.

On a certain project or technical point, a patent was produced: business configuration based on domain model.

6. Code quality:

Number of bugs after testing, number of online failures (the system can extract it, you don’t need to fill it in yourself)

Improved XX Unit testing of modules and found problems in automated regression many times.

Next, I would like to share with you a set of drawings. The author is technical KPI made by Big Data Engineer.

Big Data believes that although technical staff prefer freedom environment, but for the entrepreneurial team, if everyone does not follow the standards, the company's technical team will be in disarray.

What KPIs should technical personnel use?

Based on the comments in the previous Alibaba technical article, the author Ming believes:

KPI is active, and KPI should have focus at different stages.

1. In the entrepreneurial stage, KPIs should be tilted towards business as much as possible. When survival is a problem, talking about technical ideals is simply utopia.

2. When the income gradually stabilizes, you should try to lean towards technology. Only at this time can you talk about robustness. When the whole thing is on the right track and the system is robust enough, it will be more appropriate to continue to maintain a 6/4 (business/technology) ratio. Of course, this should be flexible according to the actual situation of each enterprise. Regulation.

This also puts forward a multi-dimensional assessment standard for TL's own business, technology, and market. Pure technology or pure business is not advisable. After all, everyone works in commercial institutions, and business Profit is the first priority, and everything else must revolve around this. Technical optimization also serves commercial profits.

Therefore, it is a kind of utopian thinking to put aside technology and talk about business or put aside business and talk about technology.