10 Software Process Management Best Practices

10 Software Process Management Best Practices

Regardless of software process management type (waterfall, scrum, iterative etc.), there are some main rules which should be considered about software process management. In this post we will mention most important of these practices theorically:

 
1. Define roles and tasks of team members: For productivity, roles of team members should be defined clearly. These roles may be project manager, team lead, develeoper, tester etc. Furthermore, authorizations and responsibilities of these roles should be defined very clearly. Task-assignment based development should also be applied for avoiding redundant effords and chaos.
2. Define meeting types: Meetings are very important if we are talking about software process management and should be defined in detail (meeting participants, context, average duration etc.). Team members should also obey meeting rules. This will bring more productive meeting and development process, and will avoid losing time unnecessarily.
3. Define documentation strategy: Every single software process has a documentation strategy, even agile or extreme methods (backlogs, lessons learned info, bug items etc.). These documentation types should be defined clearly (document physical properties, standard sections, document update period, version numbering etc.) for consistency, understandability and more effective software production.
4. Define software quality metrics and measure them:
There are so many metric types (line of code, cyclomatic complexity etc.) in software world. According to the properties of your software type, some of these should be chosen to measure quality and growth of code. These info may be discussed periodically and  production quality will be increased after assessments. 

5. Perform issue/requirement tracking: Requirement management or issue tracking is one of the key points in software development. They determine the scope of software, also supports a traceability base for functional tests. Those issues/requirements are preferred to be saved and managed with useful tools which have more features than text editing. 

6. Perform version controlling: Version controlling is also crutial in production. This should include code and other documents. It will support co-working on code, returning to older versions of code and will ease versioning and dependency management. Versioning strategy (version numbering,versioning periods etc.) should also be determined clearly to obtain consistency. 

7. Perform testing: Testing is one of the main phases of software development. Unit testing must be performed for every type of software, except a few exceptions like some user interface code. Other testing types (system, user, integration etc.) should be defined clearly and applied consistently as defined. This will increase quality of production and decrease bugs. 

8. Perform dependency management: As software projects grow, more external libraries (jar, dll, ...) or projects (external projects or internal company projects) are included. If those items are added to projects imprecisely, later updates or version changes will bring chaos and so much time consumption. Dependency management strategies and tools should be used for efficient productivity. 

9. Perform code reviews frequently:  Code review brings high quality code. First of all it enforces developer to produce better code, because it will be controlled by others. Besides, a junior developer will learn better coding quickly by corrections of senior developers. So, code review is a partial type of pair programming and it enhances productivity. 

10. Save "lessons learned" info for your projects: Even if there are experienced staff in projects, there may be unforeseen events which may obstruct or retard development process. This can be a complex item configuration, error, production experience etc. Those happenings are highly preferred to be written in "lessons learned" documents and shared in public locations. This will avoid recurrence of time loss and provide more productive software process management.