Requirements Negotiation and Prioritisation

Abstract

The quality of a software application is highly dependent on its ability to meet the necessities of customers and users, so it is crucial to capture and specify all requirements that the system must possess. However, many projects have more candidate requirements than those that can be built without exceeding the available time and budget. This chapter presents and discusses the need to allow stakeholders to collectively negotiate the requirements in order to decide which ones shall be incorporated in the system. The chapter also discusses the importance of assigning priorities to requirements, in the context of a system project. The aspects related to requirements prioritisation and some of the most popular prioritisation techniques are also presented.

Appendices

Further Reading

There are many books about negotiation in general, a subject that is quite fascinating and that helps in addressing technical problems, but also daily ones, with strategies that have proved useful. A classic book is authored by Fisher et al. (1999). Another book Fisher and Shapiro (2005), by the same principal author, deserves to be read, since it adds the emotion to the negotiation process. A very complete book, written by Lewicki et al. (2010), is also a good reference.

The literature that addresses negotiation in the more specific context of the requirements is not abundant. The more relevant material is the chapter written by Grünbacher and Seyff (2005). An analysis of the practical utility of the different scales in the context of the requirements prioritisation can be found in Karlsson et al. (2006). Some tools also exist to support the negotiation of requirements, for example, Boehm et al. (2001).

Interesting and useful is also an incursion into game theory, whose application in negotiation contexts is often highly pertinent. A well-presented textbook on classical game theory is Peters (2008), that provides a rigorous coverage of the subject. The book by Bazerman and Moore (2012) is worth reading, especially Chap. 10.

In addition to AHP, there are other multi-criteria decision algorithms that can be adopted for supporting the prioritisation process. A good source for studying some of those algorithms is the book organised by Figueira et al. (2005).

Exercises

Exercise 6.1 (Naveda and Seidman 2006, pp. 71–72) While eliciting requirements, the analyst has registered, for each requirement, the name of the person that firstly proposed that requirement and the date in which that proposal was made. Which of the following missing pieces of information will have a bigger impact (negative, due to its absence) on the activities associated with change management?

(a) Traceability, (b) Requirements type, (c) Priority, (d) Source.

Exercise 6.2 (Naveda and Seidman 2006, pp. 37–38) Which of the following sentences better portrays the requirements management process?

• 1. Measuring the requirements quality permits saying that they remain unchanged over time.

• 2. A requirement, once rejected, should remain rejected.

• 3. Change is inescapable, so the requirements management process should take into consideration that fact.

• 4. One should use a computer-based tool to trace a set of (functional and non-functional) requirements.

Exercise 6.3 Consider that you have six requirements ($R_1$ to $R_6$). Rank them, according to the importance (weight 70 %) and user satisfaction (weight 30 %) criteria, with the AHP method.

Exercise 6.4 Which of the following criteria/dimensions should exist in a mechanism for the classification of requirements negotiation processes?

(a) Support to renegotiation, (b) Strategy to resolve conflicts, (c) Level of automatisation, (d) Support to documentation.

Exercise 6.5 Describe each possible posture that a stakeholder can have with respect to the negotiation process.

Exercise 6.6 coordinated by the instructor: Divide the participants in the session in pairs for playing arm wrestling (also known as bras de fer). If you are reading this book outside the context of an academic course, find a friend or family member to play with you. There are 30 s to play and the objective for each player is to maximise the number of points. At the end, each participant indicates the number of points obtained, the winner being obviously the one that was able to accumulate more points.

Exercise 6.7 (Raiffa 1982, pp. 262–267) coordinated by the instructor: Divide the students in groups of three persons, designated by A, B and C. If you are reading this book outside the context of an academic course, find two friends or family members to play. The objective of the game for each participant is, during at most 30 min, to negotiate within her group an agreement that maximises her individual gain, taking into account the point limits defined in the following table:

Agreement	Persons	Total gain
1	Just A or just B or just C	0
2	A and B	118
3	A and C	84
4	B and C	50
5	A and B and C	121

Before the negotiation starts, each participant should carefully analyse this table, establish a strategy, and write it for further analysis. In this phase, the participants do not communicate among them. During the negotiation, if two persons wish to have a private conversation, they can do it just once, during 2 min. At the end of the negotiation, the group must indicate which persons are in the agreement and how (i.e., with numbers) the gain is divided among them. Those that do not participate in the agreement have zero points. When all negotiations are concluded, the results for the three different roles (A, B and C) in each group are announced. Based on that information, each group analyses the performance of each element and discusses what happened in the negotiation.

Exercise 6.8 coordinated by the instructor: Imagine that a given city is threatened by a bombing attack. There are 12 persons interested in seeking protection in an air-raid shelter, that however can only accommodate six persons:  

men

• A 40 year old violinist, addicted in cocaine;

• A 25 year old lawyer;

• A priest with 75 year old;

• A 20 year old atheist, responsible for several murders;

• A 28 year old physicist, that only enters the shelter if carrying a gun;

• A 21 year old poet, that adores to declaim his poems;

• A 47 year old homosexual;

women

• The wife of the lawyer, that is coming out of the madhouse; the lawyer and the wife prefer to stay together, even if outside the shelter, than to be separated;

• A 34 year old prostitute;

• A university student that did a vow of chastity;

• A 12-year-old girl with a low IQ;

• A 32 year old woman with mental disabilities, that has epilepsy.

• 1. Make your list, deciding which six persons shall enter the shelter.

• 2. Form a group with three other persons and decide upon a unique list with the six persons to enter the shelter, base on the individual lists.

• 3. All the students should now produce a list of six persons, based on the lists of each group.