BUS610: Business Intelligence and Analytics

2a. Outline how business intelligence teams "define the problem" by learning more about the context of "the problem" and its relationship to other aspects of the operation 

• What is meant by the term analytics mindset?
• What are the three major decision-making styles?
• What is the benefit of making decisions driven by data?
  

To make informed decisions based on data, managers must have an analytics mindset to understand how the data is derived, interpreted, and communicated. Therefore, they need to develop an analytical skill set that helps them know what makes sense and understand where analytics adds value. This enables them to be confident enough to ask pertinent questions of the analyst. Read this article to learn the benefits of managers developing an analytical mindset.
 
Decision-making boils down to choosing between alternatives against a defined set of selection criteria. There are usually costs versus benefits, advantages versus disadvantages, and alignment with preferences. The more factors to consider, the more challenging the decision will be. Adding time limits and personal emotions further complicates the process. Utilizing data to help optimize the alternatives informs the decision-maker with much more support to make their decision.
 
There are different decision-making styles. Each decision-making style is characterized by either a task or social focus and a high or low tolerance for ambiguity. Styles with a high tolerance for ambiguity can work with unknown variables as they come to a conclusion. Those with a low tolerance for ambiguity want as much clarity as possible in all the circumstances and information that lead to their decisions.
 
Decision-making styles can generally be defined into three categories:

1. Psychological, which is based on the decision-maker's needs, desires, preferences, and/or values.
2. Cognitive, which involves an integrated feedback system between the decision-maker and the environment's reaction to those decisions. This style includes iterative cycles and regular reassessments to measure the impacts of the decision.
3. Normative, which derives decisions based on the communication and sharing of logic within an organization's normative construct. The decision has to fit in and support the organization's mission and goals.
   

A major part of decision-making involves analyzing a defined set of alternatives against selection criteria. These criteria usually include costs and benefits, advantages and disadvantages, and alignment with preferences. The ability to make effective decisions that are rational, informed, and collaborative can greatly reduce opportunity costs while building a strong organizational focus.
 
Data-driven decisions are supported by facts, not guesses or hunches. Making decisions supported by data versus guesses or hunches helps the organization make better decisions, enables replication, and minimizes liabilities from making unsupported decisions.
 
The classic model of decision-making has been in existence for many decades and forms the basis from which other, more modern decision-making models build. A solid understanding of and the ability to apply the classical model is an essential core skill of any decision-maker.
 
The following figure provides an overview of the managerial decision-making process.

To review, see Overview of Managerial Decision-Making.
 

2b. Compare the different methods BI uses to support management teams, such as data mining, reporting, and visualization, trend and statistical analysis, and predictive analytics 

• What are the possible outcomes of the decision-making process?
• What types of analysis tools can be used to support decision-making?
• What methods are used to store, manage, and mine data to support decisions?
  

Data-driven decision-making uses a variety of machine learning approaches for data analysis by characterizing a decision problem and ascertaining the connections between the problem variables (input, internal, and output variables) without having explicit knowledge of the physical behavior of the decision model.
 
The processes included in BI architecture are data collection, data integration, data storage, and data processing.
 
For a business problem to be well-defined, it must be measurable and the operation repeatable over a specific period. To be measurable, the results must be able to be measured or counted to determine if the prediction was accurate. A repeatable operation requires that the chosen attribute to measure must occur regularly and have a repeatable pattern. A specific period requires the variable to have a specific beginning and end, such as a week, month, or quarter.
 
Intelligence analysis and decision-making can have elements of both art and science. This is because standard approaches and tested techniques can codify and help make the processes orderly and productive. However, these methods are only as good as the individuals who implement them. Humans can be unpredictable, and even the best forecaster can be hindered by the outcome of a project, even when the analysis has been perfect.
 
Technological advances have changed the practice of BI exponentially. The ability of sophisticated software to collect and process data from myriad sources allows so much more information to be available to analysts and managers, and it can overload them. Dashboards are important for presenting data. Something as simple as an electronic catalog from which you can search for library materials could be considered a basic dashboard. You put in your search terms, which act as data filters, and the system shows you the best matches.
 
To review, see Decision-Making Tools.
 

2c. Describe various management tasks, from policymaking to performance evaluation to improving procurement strategies to identify relevant trends to understand how they benefit from using BI 

• What are the differences between the four different analytical models to frame tactical and strategic questions?
• What are the various analytics domains that could be deployed in an organization?
  

A well-defined business problem should start with a question that needs to be answered. It needs to be measurable and the operation repeatable over a specific period. An organization needs to have a starting point of what it would like to know, such as a relationship between x and y. Defining the requirements and the objective of a decision is a critical first step in the whole process. If the requirements are not defined well, the subsequent steps in the process will only take the decision-maker further from their objectives. If we ask the wrong question, we will get the wrong answer. In defining the objectives and requirements of a given decision, the decision-maker will also need to look at the information in new and novel ways. The managerial catchphrase you sometimes hear is thinking "out of the box" and not letting past assumptions limit your ability to see the bigger context of a given decision-making problem
 
Analytics can assess and visualize decisions, describe the implications of historical data, predict and model future expectations, and optimize internal processes. Navigating and deriving value from big data is critical to successful organizational management.
 
There are four primary tactical and strategic analytical models managers may use to frame analytics:

1. Descriptive analytics: what happened
2. Decision / diagnostic analytics: why it happened
3. Predictive analytics: what will happen
4. Prescriptive analytics: how we can make it happen
   

Forecasting is how an analyst or team uses analysis to develop estimates on what is likely to happen in the future. This works much better in the short term than in the longer term, as conditions are unlikely to change as much in the next 6-12 months as they are in the next 5-7 years. However, as the COVID-19 pandemic has shown us, sometimes there are shocks in the environment that make even the best short-term forecasting look unreliable in hindsight. Keep in mind that such extreme external shocks are rare. The last global pandemic, for instance, was a century ago. Thus, there is definite value in data-driven forecasting.
 
Analytics can impact nearly every domain in an organization, including finance, marketing, talent, customers, risk management, transportation, and sales.
 
To review, see Why You Think You're Right Even if You're Wrong.
 

Unit 2 Vocabulary 

This vocabulary list includes the terms that you will need to know to successfully complete the final exam.

• alternatives
• analytical model
• analytics
• analytics mindset
• business problem
• classic model
• cognitive analytics
• dashboard
• data mining
• decision-making style
• descriptive analytics
• diagnostic analytics
• directive
• forecasting
• measurable
• normative
• psychological
• prediction/predictive
• reporting
• visualization