Business Intelligence and Analytics are actually two separate disciplines and there are vast differences in skill sets between the two. I like to refer to this as "All-Time Horizons." That said, both BI and Analytics use data and are about data. We just work with the data using different strategies, tools, and objectives — often with the same datasets. The difference is really what we are looking for in the data and what we want to do with the data.

In the diagram above, the highlighted teal zone represents what we are calling Business Intelligence or BI for short.

Now that we understand the differences between Business Intelligence and Analytics, let's look at the concepts we will cover that apply across all time horizons. The industry is flooded with vendors and tools that perform the same operations and functions on the data. Just like there are many automobile manufacturers, there are as many — or possibly more — software vendors in the Business Intelligence and Analytics space.

Learn one tool especially well for each thing you are attempting to accomplish. Focus on the concepts and not the tools. Once you have a firm understanding of the concepts, you can pick up any toolset in short order.

"An approximate answer to the right question is worth a great deal more than a precise answer to the wrong question."

— John Tukey

With no industry standard naming and definitions, understanding the Data Science space can be the most daunting part of entering the field. If you Google "data science three pillars" or "data science Venn diagram" you will be presented with a variety of similar visualizations. The two below give a realistic representation of the three skill areas employers are looking for in analysts. The experience levels they look for do not change the tenets themselves — just how much training you need in each area.

The IMS Diagram (2014)

This first diagram is from an Institute of Mathematical Statistics posting from 2014. I invite you to read the article — it does help to frame what BI, statistics, and data science mean in the real world.

Fig. 1 — Data Science Simplified Diagram. Based on: Yu, B. (2014). IMS Presidential Address: Let Us Own Data Science. Institute of Mathematical Statistics.

The Geringer Diagram

This second diagram by Geringer facetiously includes Unicorn in the intersection of all three areas, and also shows where technologies and tools fit within the Venn. Note the Machine Learning, Traditional Research, and Traditional Software entries — we will explore the technologies deeper in the next section, Class Scope and Alignment.

Fig. 2 — Data Science as Convergence. Based on: Armengaud, E. et al. (2017). Industry 4.0 as Digitalization over the Entire Product Lifecycle; adapted from Geringer.

With a better understanding of the three pillars of data science, let's take a deeper dive into the areas and associated roles under the data science umbrella. This blog entry from 365 DataScience.com does a good job explaining the areas of data science in a matrix of technical areas across an all-time horizons x-axis. Although the blog entry is focused on differences between common terms you will often hear mentioned, the info diagram allows for a high-level scope of the technical classes in the program.

Fig. 3 — Data Science Scope Matrix. Based on: Valchanov, I. (2021). Data Science vs Machine Learning vs Data Analytics vs Business Analytics. 365 Data Science.

Even though the importance in learning should be on concepts and not tools, we still need to use tools in our learning! I have picked the core tools we will be using in class based on current industry trends. Obviously, Canadian business demands are a priority in what I select for teaching, and I continue to revisit the tools each year.

Technology changes frequently and there could be minor changes as each year progresses. For example, through 2020 I had been teaching a Statistical Learning course using the R programming language. Due to the increased demand for analysts to be exposed to the modules and libraries available in Python, I am now using Python for both an Intro to Data Science and an Applied Data Science course.

Although most new learners are excited about the tools we use in Business Intelligence and Analytics, it is important to understand: it's the skills that make the analyst valuable. This is often overlooked in career preparation, project planning, and even in job postings.

I don't know how many times I have been told, "we need another Data Stage ETL person," or "we need another Siebel report author." Most often in these cases, what was needed was somebody who could transpose data into information. The chosen tool didn't matter, and mastering a tool does not ensure a person can understand the needs of the business and provide the information needed to guide decision making.

I came across a great post by Chris Dutton on LinkedIn that I agree with completely. I know my students have heard this from me often in class, but it helps that the message is getting out there in industry circles as well.

I am often asked: "How strong do my mathematics skills need to be to write code required for machine learning?"

A simple answer is an understanding of Statistics, Linear Algebra, and Calculus. The more you understand and can apply about each of those disciplines, the more desirable your skills will be in the market. Because of available tools and varying roles, there is a spectrum of skills required depending on the role.

Three Types of Interaction with Machine Learning

The analogy to music from the book provides a helpful explanation for the spectrum of mathematical depth different roles require:

Astute Listener — The democratization of machine learning by open-source software, online tutorials, and cloud-based tools allows users to not worry about the specifics of pipelines. Users focus on extracting insights from off-the-shelf tools. This is like listening to music: choosing between different types, benefiting from it. More experienced users are like music critics, asking important questions about ethics, fairness, and privacy.

Experienced Artist — Skilled practitioners can plug and play different tools and libraries into an analysis pipeline. The stereotypical practitioner would be a data scientist or engineer who understands machine learning interfaces and their use cases. This is like a virtuoso: highly skilled practitioners bring existing instruments to life. Using mathematics as a primer, practitioners extend and generalize existing algorithms.

Fledgling Composer — Developers of machine learning need to develop new methods and extend existing algorithms. They are often researchers who need to understand the mathematical basis of machine learning and uncover relationships between different tasks. This is like composing music: within the rules of musical theory, creating new and amazing pieces.

Most BI & Analytics professionals operate at the Astute Listener or Experienced Artist level. The Fledgling Composer level is the domain of ML researchers and engineers.

Many students ask what they should brush up on before starting a Business Intelligence and Analytics program. In a nutshell: make sure you are comfortable with advanced worksheet operations in Excel — pivot tables, lookups (i.e. VLOOKUPs), built-in functions, and data tables. Reviewing the statistical package in Excel as well as reviewing inferential statistics is also a good idea for when you begin working with statistical learning models.

I would also recommend exploring an open source database like SQLite (see link below for free resources).

Core Resources

LinkedIn Learning Resources

Many public libraries and schools in Canada have an agreement with LinkedIn Learning. With technology changing as fast as it does, traditional textbooks cannot keep up — videos help bridge the gap. You can often access LinkedIn Learning free with your student ID or public library card.

I am often asked why Business Intelligence listings on job sites require a minimum of 2–5 years of experience.

The reason why you see an experience requirement for Business Intelligence Analyst roles is that the role itself is not an entry-level position. As you may have noted in the Learning Scope and Alignment section, there is an expectation of the analyst to possess a broad skill set.

This desired skill set spans:

This breadth of requirement is why the program is structured to build skills across all three of these domains simultaneously. Entry-level positions that lead toward a BI Analyst role include data analyst, report author, and ETL developer roles.