Chapter 1 - A new approach for complex problems
--------------------------DE-TOUR--------------------------
The Scrum Guide directly introduces Scrum as a framework for developing and sustaining complex products. However, it is important to know the old ways of developing products in order to better understand the contrast that Scrum brings. In this chapter, the following are included for clarity and context. They are not part of The Scrum Guide:
Waterfall, Traditional project plan, Late feedback, Cone of uncertainty, Scientific approaches
--------------------------DE-TOUR--------------------------
How a Product is built using the traditional way – Waterfall
Organizations create strategies for business purposes. Some of these strategies aim at building product capabilities. For example, a software product building strategy may include a Customer Relationship Management System, Billing and Payments, Mobile Channel, New Product Introduction, etc.
Many organizations use a process methodology called waterfall in their projects to build products. Two types of people are needed to define the business and development aspects of project plans.
The business people: They define what the product should do. To define the product needs, they make long-term predictions about the future. Some examples of predictions include what the value of the product will be, how it will be received in the market, and more importantly what that market will be when the product is actually released. Predictions are based on many assumptions and projections built on what is known today, which may be years before the product is actually released.
The project managers or planners: They plan how to build the product. Based on the product needs deduced from the business predictions, they come up with a sequence of activities like Analyzing, Designing, and so on. They also predict (estimate) the future cost and time of these activities to find the total project cost and time. Such estimations are based on many assumptions and future projections of the product needs, people competency and behavior, project technology, etc.
Why the name waterfall?
The outcome from the waterfall-based project planning is a detailed plan.
Later, during the execution of this plan, future changes may invalidate the current plan. Such changes may require re-planning project activities, sometimes from the basic product definition, and re-doing all the activities. Such revisiting, re-planning, re-doing, is called Iterating. In waterfall, iteration is seen as the result of bad planning. So, when people in waterfall encounter a change, they reactively go through an elaborate change control process to resist the change in order to keep the original plan intact.
A waterfall-based project plan expects that the sequence of steps in the plan will go forward without requiring changes. They resist changes when they do occur. So, the plan is designed such that the activities only go forward without iterating back again, and hence the name waterfall. It is just a waterfall, the water falls only forward, not backward.
Fig. 1- Waterfall way of building products
-------------------Question- 1--------------------------
In the waterfall methodology, the duration of the activities, like development, testing, etc. are
a) Planned to be fixed irrespective of any calculation (time-boxed).
b) Predicted using some calculations based on “today’s weather.”
c) Never decided upfront.
-------Answer-------
The waterfall activities are estimated and calculated well in advance based on today’s knowledge and assumptions. Their duration will vary depending on the calculations. They are not time-boxed; they do not have fixed durations. Correct answer is ‘b.’
-------------------Question- 1--------------------------
Waterfall – The tale of late feedback
After executing the complete plan, usually after a long period of time, the product is delivered to the business/users in a big one-time release. These people see the outcome and provide feedback about the product. The business may choose to test the market and release the product. So, the market also may provide feedback.
If the feedback is positive and indicates the larger acceptance of the product, everyone is happy.
However, this is not always the case. The assumptions made by the business about user behavior may be invalid. The interpretation of the Project Managers/Planners about what the business people wanted may be incorrect. Also, the external factors, market receptiveness, and assumptions might have changed.
Sometimes, the feedback may be about a new insight that requires major modification to the product or the product itself is identified as obsolete. Such feedback is called Late Feedback since the waterfall project may already be closed or it may involve too large an effort to incorporate the feedback so late.
Fig. 2- Late Feedback in Waterfall
-------------------Question- 2--------------------------
In waterfall, the project team often gets early feedback from customers or markets about the product they are building, so they can act on this feedback on time.
a) True
b) False
-------Answer-------
Waterfall-based projects rarely offer any opportunity to get early feedback about the product. Correct answer is ‘b.’
-------------------Question- 2--------------------------
Complex problems
Product building is a complex problem. Most often it is not just an isolated product development but may involve integration of the product into the larger organizational system.
For example, in software development as technology evolves and markets change, organizations need to continuously adopt the newer software developments and enhancements into their existing complex web of technology infrastructure.
This complexity is further multiplied by the presence of many other factors such as different people, processes, technology components, and so on.
Complexity becomes more dynamic – Complex Adaptive Problems
Product building is not only complex. Multiple factors involved in building complex products also vary over time. For example, in a software project a software component that worked in a small-scale system, which has limited users today, may not work when the system becomes huge with large numbers of users tomorrow. Similarly, some of the developers that worked today may not be available for the project tomorrow, and the productivity may vary.
Such time-dependent complex problems are also called complex adaptive problems.
In these problems, the amount of unknown is really huge in the initial stages and will likely trend down over the course of product development. This trend is called the cone of uncertainty. Future change is certain in complex adaptive problems.
-------------------Question- 3--------------------------
What factors will increase complexity? Select all that apply.
a) Larger number of people on the project
b) Longer duration of the project
c) Batching of a large amount of features into one big release
-------Answer-------
All these factors – more people, longer duration, and only one big release – increase the complexity. Correct answers are ‘a,’ ‘b,’ and ‘c.’
Note about multiple correct choices: In the actual PSM 1 assessment, if you need to choose more than one correct choice, each choice is provided as a check box (not radio button). Additionally, the question will ask you to select two / three correct choices. In this book, we have used ‘Select all that apply’ to indicate there are multiple correct choices. This is intentional so that you can learn more by thinking through each choice.
-------------------Question- 3--------------------------
Is waterfall-based Project Planning a scientific approach?
At the initial stage of a project, planners use techniques such as Critical Path Method (CPM) to scientifically calculate the duration of the project. These calculations need some inputs such as product definition, people productivity, etc. The planners make assumptions to arrive at these inputs.
Since the project is at the initial stage, the cone of uncertainty is high and hence the assumptions have high probability of becoming incorrect.
So, if these assumptions are incorrect, output from the scientific calculation will also be incorrect. Therefore, project planning in the waterfall method is based on scientific calculation, but it does not mean that this planning is foolproof and risk-proof.
Are scientific approaches good for complex adaptive problems?
Scientific calculations and predictions are helpful when the problems are deterministic, stochastic, etc. In these types of problems, the future behavior or result can be modelled or predicted.
Complex adaptive problems are not deterministic. Deterministic requires the elements of the problem to be either constant or follow a definite mathematical model. The elements of complex adaptive problems are not predictable and do not follow a definite mathematical model.
Complex adaptive problems are also more complicated than being stochastic. While the elements of stochastic problems are random, the range or the boundaries of variation can be predicted. This boundary can be derived from data of the past. An example is the toss of...
