Analytics is one of those fantastical words that can be used to mean just about anything we want it to mean. According to Merriam-Webster, analytics is “the method of logical analysis.” Dictionary.com defines analytics as “the science of logical analysis”. Unfortunately, both definitions use the root word of “analysis” in the definition which seems a bit like cheating.
Analysis is derived from the Medieval Latin (analȳticus) and Greek (analȳtikós) from the 1580’s, and means to break up or to loosen. In today’s world, analytics provides a structured approach to problem-solving.
For example, Barbara Minto (see The Minto Pyramid Principle) advocates inductive logic for most problem-solving exercises, stressing the importance of considering mutually exclusive and collectively exhaustive sets of ideas, which can be seen in other works on problem-solving (see, for example, Strategic Thinking in Complex Problem Solving.)
What Analytics is not..
If we define Analytics as a process or discipline, then by definition it cannot be a thing – that is,
- It is not an end in and of itself (that is, it is not a destination)
- It is not a technology (although technology is used to support the process)
- It is not simple mathematics
- It is not statistics, although statistics are used heavily in testing our theories about the data
- It should not simply describe but rather get at the root of the issue in order to influence change
Often vendors will describe their products as “analytics solutions,” which can be helpful if we are really using it as part of the solution. But often such offerings don’t help with the complete solution e.g. the full process of analytics: Problem definition, question design, business analysis, impactability, storytelling, and operationalization.
What is Analytics?
For the intent of this article, we define analytics as the scientific process or discipline of fact-based problem-solving.
We use logic, reasoning, critical thinking and quantitative methods to examine phenomena and determine its essential features. Analytics is rooted in the scientific method including hypothesis testing, inductive logic, and deductive logic. One of the confusing parts of analytics is that it is often paired with other words such as:
- Big Data Analytics
- Descriptive Analytics
- Prescriptive Analytics
- Business Analytics
- Operational Analytics
- Advanced Analytics
- Real-time Analytics
I view these pairings as qualifiers on the type of analytics that we are doing or special cases. For example, Big Data Analytics uses the process/ discipline approach to understand the massively large databases to create undeniable truths about the nature of the phenomenon being described or considered in the underlying data.
Advanced Analytics, on the other hand, is a specialization of analytics that uses highly developed, computationally sophisticated techniques with the intent of supporting the fact-based decision process.
How does Analytics compare to other concepts?
It is useful to highlight the differences between analytics and its nearest neighbors as this can prove useful in helping clarify what we mean by analytics.
Analytics versus Business Intelligence & Reporting
In a paper I wrote back in 2010, I defined Business Intelligence (BI) as “a management strategy used to create a more structured and effective approach to decision making…BI includes those common elements of reporting, querying, OLAP, dashboards, scorecards and even analytics. The umbrella term “BI” also can refer to the processes of acquiring, cleansing, integrating and storing data.”
Based on that definition, it would seem that Business Intelligence is more of a catch-all term than analytics – where analytics is merely a subset of the general class of problems defined by BI.
Since 2010, I can’t decide whether I agree with those that consider BI as the umbrella term or if analytics is a separate thing. I do consider BI as more about telling the story of what happened in the past rather than predicting the future (the realm of analytics.) The application of BI should involve the “analysis” of the data rather than simply awareness of facts but doesn’t necessarily involve advanced quantitative and mathematical techniques for doing so.
Note in the BI Portal depicted here (courtesy of Information Builders), we relay facts about the past such as sales, growth, percentages, volume but there is no synthesis of the facts to predict or prescribe change.
In sum, Business Intelligence (and its little subling, Reporting) are the techniques we use to display information about a phenomenon whereas Analytics goes beyond description to truly understand the phenomenon to predict, optimize and prescribe appropriate and inappropriate actions.
For another take on the difference between health analytics and business intelligence, take a look at this article from Jason Burke.
Analytics versus Health Informatics
Health Informatics is a healthcare-specific term. Health Informatics is defined by the U.S. National Library of Medicine as “…the interdisciplinary study of the design, development, adoption, and application of IT-based innovations in healthcare services delivery, management, and planning.”
In practical terms, informatics relates to the technologies we use to process data for storage and retrieval and sits at the intersection of people, information, and technology. The diagram above depicts the relationship between Health Information Technology and Health Information Management.
In sum, healthcare informatics relates to the ecosystem of data and systems that support operational and clinical workflows rather than the analysis of data found therein.
Analytics versus Big Data
Big Data is a way to describe the onslaught of information that organizations must deal with in their pursuit of turning data into insights. It was coined to encapsulate the complex data environment found in the modern enterprise. We contrast Big Data to traditional small day by its volume (how much data we have), velocity (how fast the data is coming at us), variety (numbers, text, images, video) and veracity (trustworthiness.)
If Big Data is a concept used to describe the current context of information, analytics is used to help us deal with Big Data in either proactive (prescriptive) ways versus the reactive ways (which is the realm of business intelligence.)
Analytics versus Data Science
It would seem that defining Big Data was a cake-walk as compared with Data Science as I found such little consistency in the term and lots of debate about what it means and whether it is different at all from analytics. Even those that would attempt a definition, I find that they do so by discussing the people (data scientists), the skills they need to have, the roles they play, the tools and technologies they use, where they work, and their educational backgrounds. For a great article that illustrates these differences, take a look at this infographic from American University.
Instead of describing Data Science by the people or the types of problems they address, I define it as the following:
Data Science is the discipline of using quantitative methods from statistics and mathematics along with technology (computers and software) to develop algorithms designed to discover patterns, predict outcomes, and find optimal solutions to complex problems.
I differentiate Data Science from analytics in that it can help support or even automate the analysis of data but that analytics is a human-centered process that takes advantage of a variety of tools including those found in Data Science to understand the true nature of the phenomenon. Data science tends to focus on macro, generalized problems where analytics tends to address specific challenges within an industry.