If you come from a business and finance background like me you’ve probably realized when browsing through data science or business intelligence articles that it seems that finance is playing catch-up when it comes to mixing it up with using analytics or business intelligence for financial purposes. Areas such as marketing, procurement, sales, and operations are not limited to only descriptive analytics but depending on how far up the curve the company is, prescriptive analytics are not only sought but used.
One of the goals of this blog will be to cover financial concepts and provide some ideas of how data science, business analytics, or business intelligence could add value. It’s important to note that value doesn’t have to always equal direct monetary paybacks, and believe me I know that one of the hardest parts of selling these concepts is that there might not be an immediate benefit on the company’s coffers.
Value could also come in the way of:
- Productivity enhancement.
- Insights that lead to agile and objective executive decisions.
- Stronger relationship with customers.
- Visualization to understand a problem.
- Time-saving on daily activities.
- Increase competitive advantage.
- Lowering risks.
Sometimes we’ll step a little bit into other functional areas given that implementing one of the mentioned concepts may have an impact on the company’s financials; for example, the Sales department using BI to understand what happened to gain insights that allow them to replicate what they did right and avoid what they did wrong or working with prescriptive analytics to realize the actions that would lead to the most successful outcome…both of which if carried out correctly would result in higher revenue.
Defining Business Intelligence, Big Data, and Data Science
I feel compelled to take the opportunity to define the difference between business analytics, big data, business intelligence, and data science for the purpose of this blog. I know how confusing it must be to know the definition for each one; especially, when some of them are used interchangeably, they overlap or work towards a similar goal.
Business Intelligence vs Business Analytics
When I was studying my master’s degree one of the most common questions that I got was “what’s the difference between business intelligence and business analytics”? One of the definitions of analytics states that the difference between these two is that analytics deals with future events by trying to predict based on the use of algorithms, taking into account the company’s data and models such as classification (trying to predict what class to assign it to) or regression (to predict a value). But didn’t we also just saw an image where we saw analytics as studying past events, in order to find out “what happened” and “why it happened”? This is where it gets tricky…
First, right off the bat sticking to a high-level definition of analytics such as “The process of exploring data and reports in order to extract meaningful insights, which can be used to better understand and improve business performance” (Source) gives us more flexibility of use. While the “improve business performance” can be interpreted lightly and is mostly done by job postings as talking about a future-only focus; in reality, this definition alone gives a clear picture that it also encompasses gaining insights about the past (after all how will a company carry out a statistical analysis for future predictions without past data). So what does this mean?
It means that despite business intelligence using only past and present data it contains analytical capabilities because of its efforts to gain insights. Just like business analytics it collects, analyzes, and visualizes data. It also presents and organizes the information for visualization purposes (ex. Tableau, Power BI). But unlike business analytics it limits itself to the questions “What happened?” (descriptive analytics) and “Why did it happen?” (diagnostic analytics). Business analytics instead of focusing on these questions takes the data to change future business operations by making either predictions (predictive analytics) or fixing the root problems that it found (prescriptive analytics).
So in summary:
Data Science vs Business Analytics and Business Intelligence
The first clear difference between data science and business analytics and business intelligence is that while the latter seek to answer a question, data science pursues the questions that should be asked based on the data. A data scientist should be able to mix business acumen with the technical skills to work with data all the way from the source to visualization; however, some sources don’t put business acumen as a must given that unlike business analytics or business intelligence data science isn’t strictly about improving a company’s results. Data science is used in academia, scientific research, environment, city management, etc.
Another key difference is that data science has become very popular because of its use to extract value out of unstructured data, while business intelligence and business analytics work more with structured data…more on this when we see big data.
While data science sounds very separate from the other two because of its use outside of business and its work with unstructured data, there is also some overlap and quite a big reason why they tend to mix it with analytics. As I mentioned before a data scientist must have a lot of skills because he doesn’t only work with analyzing the data but also works with the data since its raw stage. Preparing the data for analysis is in fact around 80% of the work. Modeling and deployment is the other 20%.
Just like business analytics shares features with business intelligence, so does data science serve as an umbrella for analytics. Data science mixes up programming, statistics, data analysis, and visualization or reporting. The efforts made by the data scientist to build a data pipeline can be later used by both data analytics and business intelligence by accessing the stored information. It is because data science covers the other two subjects that I was tempted to name this blog Data Science and Finance as sometimes the posts will also be about using analytics or business intelligence for finance. I figured Data and Finance was more appropriate given the focus is to study ways or cases where value could be extracted from data for financial purposes.
Defined in a simple manner, big data is extremely large data sets both structured and unstructured that can’t be processed with traditional applications (Excel, Access) because it is too large to be stored in a single computer. As the name implies one of its defining traits is its volume; especially, when considering data sources such as social media. Another one is velocity, you’ve probably already noticed this but data is created, replicated, and shared at a very fast pace and big data solutions must work to smoothly accept this incoming data and avoid saturation. Big data also poses a challenge due to its variety, meaning it may come in the form of structured data (data models, text only in a defined manner, information in a relational database such as credit card numbers, or customer names, etc.) and/or unstructured data (images, sounds, video, social media posts). Most importantly there is also, veracity, can you actually trust the data collected? Finally, although not every source lists it as an important dimension of big data, there is value. Aside from being trustworthy the data should also provide value (I would place its importance as strong as veracity, because if it’s not going to offer any worth why use it?). There will definitely be more than one post about big data due to its increasing use by financial services.
I appreciate your visit and hopefully you stick around as I explain finance and business concepts and then spice it up with the areas we learned about a bit in this post. The intended audience are people coming from non-technical areas but if there’s a specific technical question please ask away. Thank you and I’ll see you around, cheers!