Liam and Emma get home from work on a Friday afternoon. Tired from the week that has passed, neither of them want to make supper, so they plan to go out. Liam is in the mood for Mexican, but Emma wants seafood. They get into a heated discussion, bringing up things like who always gets their way and whose turn it is to choose, but after 30 minutes the correct decision is finally made: they get seafood and this song starts to play:

This is a day-to-day example of a decision-making situation. I had obviously jumped to a conclusion, but in software development decision analysis should be much more processed and structured. Recommending alternative business systems is one of the very first steps in business analysis and software development. Essential to the analysis phase is the solution approach. The solution approach explains the way a solution will meet the requirements necessary to solve the business problem. There are various techniques used to determine the solution approach; benchmarking, brainstorming, SWOT analysis to name a few. All of the techniques have their advantages and challenges and are equally important, however I want to focus on one of them – Decision Analysis.

Decision analysis involves using tools to evaluate all relevant information in order to generate an unbiased guide towards making the correct decision. These tools encourage a systematic approach to visually represent a quantitative evaluation of the important choices confronted by the business. Importantly, the result of decision analysis will contribute to defining the solution scope as usually the best solution will be agreed upon once the analysis is complete. This post will discuss four decision analysis tools that are widely used in the decision-making process.

NOTE: It is important to note that I will use the terms “proposed change”, “proposed solution” and “option” interchangeably and these should be treated as synonymous.

Force Field Analysis

Force field analysis involves examining forces that affect the proposed solution and creates a classic ‘pros and cons’ list. A score is given to each force based on its influence on the solution and you total up the pro and con side to see which one bears the greater total.

It is important to note that the influence score of forces will vary on the decision or decision makers. For example, when deciding on a new car to purchase, some may find that the cost holds a greater influence score than impact on environment, while others will feel the exact opposite.

Force field analysis is particularly helpful to communicate the reasoning behind a proposed change and potentially cement your choice. It works best when you have a number of propositions that have similar forces affecting it.

Impact Analysis

Impact analysis is a brainstorming technique used to identify the full impact of the proposed changes. There are 5 steps involved in impact analysis:

1. Prepare for the Impact Analysis by clearly defining proposed solution
2. Brainstorm major areas affected and who or what may be affected
3. Identify all areas and elements with the major areas
4. Evaluate impacts deciding on whether they are positive or negative
5. Manage consequences and decide whether you want to go ahead with the decision

A more in-depth look at each of these steps can be found at this useful source

Impact analysis is a great tool for gaining foresight on potential challenges and develop contingency plans for them. Since this task could be quite cumbersome, it is important to not focus on the micro-detail and identify just how far into detail you should go to get the most effective results. Finally, it is vital to remember that few changes happen in seclusion, so be mindful of the context the proposed change will function in and the reaction it may get from its users.

Decision Matrix:

The decision matrix is a highly rational and sequential tool for gathering a quantitative representation of each proposed change. The process involves the following steps:

• Identifying all options or proposed changes as row labels on a table and all of the factors that you need to consider as columns.
• Rate the importance of each factor using some kind of scale (e.g. 1 to 5) and document these next to or under each factor.
• Rank each option in relation to each factor using the same scale as used to weight each factor.
• Multiply the rank of each option’s factor by its importance rating and document the resulting figure.
• Total the now weighted ranks for each option at the end of each option.

The option with the highest rank should now represent the best option as well as give you a ranked list of the alternative options. A decision matrix is a good way of documenting how each proposed change compares to its alternatives. It objectifies the decision making process by quantifying the decisions and hence is good for when high stake or emotional decisions need to be made.

Decision Tree:

The decision tree is a very technical and visual tool for forming a balanced picture of risks and rewards for each decision. It involves mapping out each decision and its potential outcome, as well as associating a monetary value to each outcome.

The process for decision tree analysis is as follows:

_NOTE: As a result of its technicality I will simplify the process of conducting decision tree analysis, although I would advise reading more about it. _

1. Illustrate where a decision needs to be made with a square node and draw out one path for each option and associate an estimated cost for each path.
2. Illustrate where there is more than one outcome with a circular node and draw out one path for each outcome.
3. Associate a score or estimated monetary value for each outcome as well as the estimated probability that each one will occur.Multiply the value of the outcome by its probability and total the results for each node. So if a node has 3 possible outcomes, calculate each value * probability and total the 3 results. This total willbecome the value of the node.
4. You should now subtract the cost of the option from its now calculated value to give you a ‘nett value’ of the option.
5. Once you have done this for each decision path, you can weigh up which is the better choice based on its value.

As you can imagine, decision tree analysis is very involved and cumbersome, thus it is usually used when a decision is extremely difficult to make. Its technicality should not worry you though because the information it returns is extremely valuable. It allows analysis of potential consequences and is also a framework for quantifying the value and probability of proposed changes.

How do I decide how to make decisions?

While this is not a complete list, these have been seen as the most effective decision analysis techniques. Deciding which tool to use depends on factors like budget, time constraints, complexity of decisions etc. Usually it is best to use a combination of these tools to get the best and most complete decision.

We make decisions every day and subconsciously use these tools to make them (You even decided to read this blog post, for some reason… Probably because of the pretty pictures). In the realm of software development, decision-making should be made conscious. Even if the process seems arbitrary because the decision appears obvious, it is very important to document this process to safeguard yourself from the infamous scope creep.

Decision Tree:

The decision tree is a very technical and visual tool for forming a balanced picture of risks and rewards for each decision. It involves mapping out each decision and its potential outcome, as well as associating a monetary value to each outcome.

The process for decision tree analysis is as follows:

_NOTE: As a result of its technicality I will simplify the process of conducting decision tree analysis, although I would advise reading more about it. _

1. Illustrate where a decision needs to be made with a square node and draw out one path for each option and associate an estimated cost for each path.
2. Illustrate where there is more than one outcome with a circular node and draw out one path for each outcome.
3. Associate a score or estimated monetary value for each outcome as well as the estimated probability that each one will occur.Multiply the value of the outcome by its probability and total the results for each node. So if a node has 3 possible outcomes, calculate each value * probability and total the 3 results. This total willbecome the value of the node.
4. You should now subtract the cost of the option from its now calculated value to give you a ‘nett value’ of the option.
5. Once you have done this for each decision path, you can weigh up which is the better choice based on its value.

As you can imagine, decision tree analysis is very involved and cumbersome, thus it is usually used when a decision is extremely difficult to make. Its technicality should not worry you though because the information it returns is extremely valuable. It allows analysis of potential consequences and is also a framework for quantifying the value and probability of proposed changes.

How do I decide how to make decisions?

While this is not a complete list, these have been seen as the most effective decision analysis techniques. Deciding which tool to use depends on factors like budget, time constraints, complexity of decisions etc. Usually it is best to use a combination of these tools to get the best and most complete decision.

We make decisions every day and subconsciously use these tools to make them (You even decided to read this blog post, for some reason… Probably because of the pretty pictures). In the realm of software development, decision-making should be made conscious. Even if the process seems arbitrary because the decision appears obvious, it is very important to document this process to safeguard yourself from the infamous scope creep.