Do you ever feel adrift in a sea of buzzwords? Between the hundred Clouds on the horizon, the code ‘jedi’ you’re trying to hire and the blockchain disruptors leaking out of the incubator that you haven’t quite figured out yet, does it ever feel like a bit much to take in?
I thought so. And if your business is serious about data, there’s probably three that crop up time and again with only blurred definitions and vague marketing spiel to distinguish between them: business intelligence, business analytics and data science.
They may sound like much of a muchness, but there are real differences, and it’s important for businesses to pick the approach that best suits their needs. So: what are the business tools behind the buzzwords?
Business Intelligence (BI)
Business objectives: BI software is tasked with developing a set of key metrics to describe the past (not prescribe the future). It’s looking in the rear view mirror to answer questions such as, ‘what happened?’ and, ‘by how much?’
Though limited in analytical functionality, BI systems – when used successfully – optimise commercial or technical performance and streamline daily business operations.
Target user: Non-technical Analyst, Business Management
Technical sophistication: Low-Medium BI software usually has limited analytical complexity yet often has powerful reporting and rich visualisation functionality. Reporting can range from highly granular metrics to higher level business reports.
Popularity: According to Google Trends, searches for Business Intelligence have been declining over time.
Business/Data Analytics (BA)
Business objectives: BA seeks to answer the, ‘why?’ and, ‘what will happen in the future?’ Rather than only looking backwards, analytics help businesses answer more strategic questions in addition to day-to-day operational ones such as, ‘who are my most valuable customers?’ or, ‘what is the 6-month business forecast?’
Business analytics also allows users to query data and answer ad-hoc questions, which is not usually possible in traditional BI software.
Target user: Business user, Data Analyst, Data Scientist
Technical sophistication: Medium-High. Business Analytics will usually aggregate, enrich and analyse data – both unstructured and structured. Beyond simply collecting the data, advanced analytics can identify correlations between different (often siloed) sources of data and use machine learning to spot patterns, build models and make predictions.
Business Analytics introduces a higher level of statistical capabilities, data mining and predictive analytics to the more reporting-oriented Business Intelligence.
Similar to BI, Business Analytics is in the form of a software platform designed to operate across teams and users in an organisation. A key objective of an analytics solution should be to democratise access to data and insights in an organisation – to allow non-technical users to manipulate data in sophisticated ways.
Popularity: According to Google Trends, searches for Business Analytics have been increasing over time.
Data Science
Business objectives: Data scientists are often tasked with specific briefs such as to find the most valuable customers or predict churn – let’s call these the ‘known unknowns’. But they may also engage in data mining – to discover the ‘unknown unknowns’ that the business may otherwise not have found.
Target user: Data Scientist / Data engineer
Technical sophistication: High. While business intelligence and business analytics provide a software platform to view data and automate or reproduce analyses, data science requires more technically skilled professionals to use a variety of tools (many of which are open source) to answer complex business questions.
Data scientists are able to use the latest big data techniques in machine learning, natural language processing, image processing or neural networks to extract ‘hidden’ insight from the vast pool of data that businesses may have amassed as well as acquire new valuable sources of data. A good data scientist is a rare breed and as a result often difficult (and expensive) to hire; they should combine domain knowledge in the industry with both statistical and programming skills.
So as interest (as measured by Google Trends) in business intelligence has waned, appetite for business analytics and data science capabilities has increased. This may be partly down to greater industry knowledge about the more mature concept (meaning less need to Google it), but surely also due to businesses discovering they can do more with data than they ever could before.
The resounding winners in today’s data landscape are business/data analytics and data science. Whilst these areas overlap significantly, they approach big data in different ways. Advanced Data Analytics is often a more cost-effective way to automate and democratise data discovery and analysis using tried-and-tested techniques, whereas data science is more about finding new ways of analysing data by creating and deploying algorithms.
The virtue of data analytics is that it can be easily integrated into a company’s existing operations and architecture. What we will see in the future is further union between data science and analytics – more analytics software will adopt and absorb the latest data science innovations so that these tools, which were once only at the disposal of PhD-whizzes, can be in the hands of non-technical business users.
Ultimately, the right choice for a given business is a tool that is adaptable and able to move dynamically with changing business demands and new technologies. It’s important that anyone can access useful analysis to help them in their roles and this isn’t just kept to the select few as data becomes central to more and more areas of every business.
Jay Patani, Tech Evangelist at ITRS
Image Credit: Sergey Nivens / Shutterstock