What are the most popular software development models?

As anyone who has tackled a software development project can tell you, building digital solutions involves going through a complex process. This process has different stages that range from the idea and design of the software to its launch and maintenance. Development teams refer to this process as the software development life cycle (SDLC).

Maybe you know it, maybe you don't, but there are many ways a team can tackle the SDLC. In fact, we talk about different software development models depending on how the team organizes the steps and how they approach the workflow. There are more than 50 SDLC models in use today, each with its own way of organizing the development process.

However, teams mostly turn to just a few of these models as they have proven time and time again that they can deliver excellent results. What models are these? The following 8 are the most popular SDLC models.

1. Waterfall

Waterfall is one of the most well-known SDLC models as it has been in use for decades. In it, all development stages are sequential, that is, the team needs to complete one stage before moving on to the next. This way, tasks “cascade” to the next one, which is where the model’s name comes from.

Waterfall is a very structured model. This means that each step has its own set of deliverables and documents that the team needs to achieve to move to the next step. This approach to development provides high predictability in terms of schedule and resources, but makes it impossible to accommodate new requirements that may have arisen during development. Additionally, testing is often the last stage, so any issues engineers find are very expensive to fix as the product is almost finished by the time it reaches this stage.

When to use:

• Small and simple projects
• Projects with non-modifiable requirements
• Projects that need predictable budgets and schedules
• Highly regulated projects (such as those related to healthcare)

2. Model V

Many people see the V model as an extension of the Waterfall model, as it uses the waterfall stages in its first part only to move the process upward after the coding stage. This initial downward movement and subsequent upward movement form a V shape, giving this model its name. It is also known as the Verification and Validation model because each stage in the verification phase has a corresponding stage in the validation phase.

The difference with the Waterfall model, then, is that the V model has tests performed at each stage and not at the end of the process. This increases the overall quality of the product, but it also means it takes more time and money to release the software. Furthermore, requirements gathering happens at the beginning and cannot be changed during development, which implies that the V model is not flexible.

When to use:

• Complex projects that require high levels of predictability and the best possible quality (such as those related to navigation or medical systems)

3. Iterative and Incremental Model

This model combines iterative design and incremental construction model for development. The goal of the Iterative and Incremental model is to use repeated development cycles (an iterative process) to build a product that advances one small step with each iteration (an incremental process). By doing this, engineers can learn what works and what doesn't in development and apply that knowledge to refine subsequent iterations.

Thus, each iteration brings a new module to the software that builds on the previous iteration. This ensures consistency and allows some leeway in collecting requirements, as some of them can be modified as development progresses. However, it is important to note that these changes cannot be radical. The initial requirements form an unshakable foundation that can be adjusted somewhat, but still provide a rigid framework for iterative development.

When to use:

• Large projects made up of multiple moving parts
• Projects based on microservices or web services

4. Spiral Model

This model is a risk-oriented approach to software development. This means that the Spiral model is mainly driven by the potential risks of a given project. To accomplish its tasks, this model encompasses elements of different models, including incremental development, waterfall, and even evolutionary prototyping.

There are 4 main activities in the Spiral model: risk planning, risk analysis, prototyping and delivery evaluation. Thus, the project begins with a risk assessment that defines how the team will work. After the first iteration of the project is completed, the team analyzes it to improve it in the next cycle. Therefore, this model can help mitigate the risks inherent to development. However, significant customer engagement is also required in the early stages. Furthermore, the Spiral model can extend its cycles such that work can extend beyond the agreed schedule.

When to use:

• Projects without clear requirements
• Innovative projects with broad requirements
• Big projects
• Research projects involving new services or products

5. Rational Unified Process (RUP)

Another iterative model, the Rational Unified Process (RUP), is an adaptable process framework that teams can customize for their specific projects. The model itself has 4 phases: initiation, elaboration, construction and transition. Each of these phases goes through an iterative process that happens simultaneously, although with different intensity. Thus, the initial phase will have tasks focused on gathering requirements, while the construction phase will expand this focus to cover other activities such as design, implementation and testing.

Of all the models that use linear development approaches, RUP is the most flexible, as it allows changes in focus, even if the construction phase is advanced. However, RUP is not as flexible as the models listed below, mainly because it is still not as fast and versatile as Scrum or Kanban.

When to use:

• High risk projects
• Projects that require rapid and higher quality development

Agile Models

You don't need to be very involved in software development to know a little about Agile. This is because Agile is the main mindset that software development companies use to work on their solutions. Agile practices provide a lot of flexibility, which allows developers to make adjustments and corrections to the project as they go along.

Agile is less concerned with rigorous documentation or requirements gathering than the models we've seen so far. This mindset involves the collaboration of self-organizing teams that embrace adaptive planning, evolutionary development, iterative work, and continuous improvement to deliver highly functional solutions.

The Agile umbrella is broad and all-encompassing. Behind it, you can find different models that follow the Agile principles of collaboration, continuous improvement, value-driven development and technical excellence with simple, high-quality results. Given that all Agile models share a common philosophy about development, their use cases are all very similar.

When to use Agile templates:

• Innovative projects where early feedback is essential
• Projects where requirements cannot be adequately detailed
• Large projects that can be developed iteratively
• Most modern software development projects

Many models use Agile as a guiding mindset, including the three most popular: Scrum, Kanban, and Extreme Programming (XP).

• Scrum The most popular Agile model, Scrum, uses time-bound iterations (called sprints) to work on specific parts of the final product. The idea is to divide the product into small goals that can be completed in a limited time frame (usually between 2 and 4 weeks). Scrum is all about collaboration and communication, which is why the team meets every day in meetings called daily scrums. There they monitor what the team is doing and where it is going while discovering problems and possible solutions. When a sprint ends, the entire team comes together again to show the work that was done in that time (which usually consists of a working version of the final product). The same meeting also gives the team the opportunity to analyze the process and check which improvement opportunities they can take advantage of in the next sprint.
• Kanban Kanban leaves sprints aside in favor of a visual plan represented by a Kanban Board. In it, the team explains the tasks they need to perform, the functions and the progress of each one. Therefore, this model is more about transparency about the project's progress, allowing the team to detect problems and prioritize tasks more effectively. Another important aspect of Kanban is that work is not pushed to the team according to a schedule. Instead, work is allocated according to the team's ability to tackle it. This way, teams can decide what, when and how to work on the different tasks that need to be accomplished. Kanban is highly flexible as it allows requirements and changes to be introduced at any time. This certainly increases the team's agility to work on any project, but it also drastically reduces predictability.
• Extreme Programming (XP) Extreme Programming (XP) is another Agile model that also advocates frequent releases in short development cycles. The main difference with other models is that XP does not require developers to create any features unless they are really necessary. If this seems a little extreme, that's because it is. In fact, the name XP comes from the obsession with taking traditional software engineering practices to the extreme. XP's flexibility is somewhat in the middle of that of Scrum and Kanban: not as structured as the former, but not as free as the latter. Changes can be introduced after each iteration. To ensure final product quality, XP teams use pair programming, test-driven development, and continuous integration practices.

Various models for different needs

In today's software development landscape, it's easy to think that Agile is the way forward. Most software development companies use it and consider it the best mindset for the job. However, blindly adopting Agile models can have disastrous effects on your projects. Each project you need to carry out has unique requirements and characteristics, so it is important to analyze all available models to check which is the most suitable for developing that idea.

This doesn't mean you need to change development models every time. But you need to keep in mind that each model described here has its own set of advantages and disadvantages. Learning about them is the first step to enjoying their benefits.