Introduction to Business Analytics, Part 2 [BAS 120]

Introduction to Business Analytics, Part 2 [BAS 120]


[music] This video will cover introductory information about business analytics, including the role of the business analyst, what makes a business analyst successful, an overview of business analytics tools, project outcomes, and the analytical process. There are 3 main reasons Business Analytics is an enticing career choice. The first is that there is a high demand for business analysts, and the relatively low supply of skilled workers means that salaries are higher in this field. Another reason is for the challenge of solving interesting problems. Analysts are typically people who are interested in solving complex puzzles. Finally, those who are curious about how things work can use their skills in analyzing data to uncover previously unknown truths. A business analyst can take many roles depending on the data and type of project. The most common roles are that of an interpreter, in which the analyst uses Descriptive Analytics to tell the story of what happened; an oracle, in which the analysts uses Predictive Analytics to predict future events; and a consul, in which the analyst uses Prescriptive Analytics to provide advice on the best course of action. An analyst becomes successful due to a combination of “hard” and “soft” skills. The hard skills are more tangible, and refer to what the analyst can do and with what tools, while the soft skills are less flashy on a resume, but equally or even more important than the hard skills. Analytical tools can be separated into two categories: software that requires coding, and software in a graphical user interface (GUI) that is based on point-and-click interaction. The main benefit in writing code is that it allows for more flexibility – there are more features and allows the user more possibilities. The drawback to coding is the extended learning curve. Within the last decade, there have been many new GUI programs that make analytics easier to implement without the need for writing code. Programs such as Tableau, Alteryx, and RapidMiner have started gaining market share, going along with older tools such as SAS Enterprise Guide; but none of these tools have yet to replace the overwhelming popularity of code-based software such as SAS, R, Python, or SQL. There are many different goals that an analytic project can strive for. Typically, these goals fit into one of two categories. The first is geared towards providing information about the business, such as reports and dashboards, for business stakeholders. Reports or presentations provide one-time insights to explain events that have occurred and predict future events, and dashboards are used by stakeholders for ongoing monitoring of key aspects of the business. The second category deals with the production of analytical products. In these types of projects, the business’ data becomes the input for a complex process that automatically produces an action. This can take the form of features to offer a better experience for consumers. For instance, Amazon has automated algorithms that determine products that you might like to buy. Analytical products can also be built to make internal business processes more efficient. An example of this is how credit card companies test every transaction for the probability of fraud. An analytical project should start with the goals well-defined. Very rarely are projects started to simply explore the data or find hidden truths. Collecting an inventory of all the relevant data sources is also an important step in beginning an analytical project. Finally, every analysis should begin with an effort to better understand the data. This should be done before any analytical techniques are used – simply observe the data files, and write down a list of observations, questions, and any other ideas you have. This process, called “creating disfluency,” will enhance the data dictionary and help the analyst internalize elements of the data. Cognitive disfluency is the principle that enables students who take lecture notes by writing to retain more of the material than students who type notes, even though those who type can take notes more efficiently. Sometimes, the more work we have to do to process the information, the better we can understand that information. Using this principle at the beginning of a project, before using any advanced analytical techniques, enhances the analyst’s capability of understanding the data. After the initial stages of an analytical project, there are a few more steps in the analytical process. The majority of time will be spent exploring and preparing data. In the exploration stage, the analyst learns more about the variables, including distributions and frequency of values, and also identifies variables with missing, or Null, values. In the preparation stage, the data becomes more useful, as variables are transformed and anomalous values are rectified – this is referred to as “data cleaning.” Another common task during this stage is to join data into as few sources as possible. Perhaps one of the most valuable techniques to use during the preparation stage is called “feature engineering,” in which the analyst transforms certain variables into new variables containing slightly different data. This allows “hidden” aspects of variables to be analyzed. For example, a dataset with a date variable can have a new variable added to determine whether that date is on a weekday or the weekend, and this new variable could add important information that was not readily available in the original data source. This is one of the areas of the analytical process where creativity is needed. The last two steps in the analytical process are to build models and then put those models into production. A model is a type of mathematical equation that describes relationships among variables in a data set, often for the purpose of predicting an outcome. By putting a model in production, an automated decision can be made when new data is observed. In some cases, models are not the goal of a project; rather, the goal is to analyze data and communicate the findings in an analytical report, presentation or dashboard. Visualization of the data is also a key component throughout the entire analytical process, as the analyst attempts to learn more about the data. This concludes our introductory video about business analytics. Today we covered the role of the business analyst, what makes a business analyst successful, an overview of tools, project outcomes, and the analytical process. [music]

You May Also Like

About the Author: Oren Garnes

Leave a Reply

Your email address will not be published. Required fields are marked *