At this point, everyone understands the importance of and need for analytics.
But understanding the need and actually knowing how to go about leveraging analytics are two very different things.
Adding to the confusion is the ambiguity around the term itself. As often happens with “hot” concepts that intersect the worlds of business and IT, marketers latched onto the idea of analytics as the next big thing and did exactly what you’d expect: marketed it like crazy.
Analytics Buzzword Bingo
As we all know, when marketers (these are my people, so I know of what I speak) get going, the buzzwords start rolling in. And do we have ourselves a jumble of analytics buzzwords these days.
Business Intelligence. Business Analytics. Advanced Analytics. Data Mining. Machine Learning. Artificial Intelligence. Deep Learning. I could keep going if it wasn’t for those pesky editorial word counts.
I’ve been involved with analytics my entire professional life and I’ll be the first to admit, I often don’t know what someone is referring to any more when they toss around these buzzwords — and most of the time, neither do they.
This is a problem. And it's one we need to solve before we can truly make progress with our desire to implement meaningful analytics initiatives. With that in mind, let’s take a step back and look at what some of these buzzwords really mean.
Parsing the Analytics Nuances
Despite the proliferation of buzzwords, we’re (in general) referring to the same basic thing: the idea of combining data with math in order to derive insights.
There are, of course, many different ways to go about doing that — hence all the buzzwords — but at the highest level, that’s what we’re all after. But that doesn’t mean all analytics concepts are the same. Understanding the nuances and elegance of each is critical to knowing which to apply and when.
Business intelligence, the old stalwart, is about using data to understand what’s happened in the past. It’s about running reports on systems of record — ideally using well curated and trusted data — to deliver metrics or KPIs that provide insight into what’s happened with your business over a defined period of time.
For example, you’d use business intelligence to assess what changes have occurred in the way your top customers have transacted with you over the past 12 months. BI is learning done in hindsight, but valuable nonetheless.
Business analytics relies on finite sets of data and is grounded in the use of logic to undercover potential business improvements. For example, you’d use business analytics to learn not just how your top customer has transacted with you, but why they’ve behaved that way, thus paving the road to an actionable decision based on that logical outcome.
This is where we start to go deeper. Advanced analytics is an umbrella term that encompasses both predictive and prescriptive analytics.
In short, advanced analytics is about applying sophisticated mathematics to the right subset of data in order to make predictions about what is likely to happen next. Advanced analytics usually requires the expertise of a data scientist skilled at building complex data models.
You’d leverage predictive analytics if you wanted to better understand the likelihood of a patient being readmitted to the hospital following a surgery. If someone is marketing a business intelligence or business analytics solution that purports to help you predict customer behavior, either the vendor doesn’t understand the difference between BI and advanced analytics — or they’re hoping you don’t understand the difference.
Data mining is an equally advanced mathematical process that focuses on uncovering complex relationships hidden within the data. Data mining is often leveraged to answer questions that are vague in nature and leads to outcomes that you otherwise could not have anticipated or articulated on your own.
Data mining differs from advanced analytics in that you don’t necessarily know what you’re initially looking for. It’s about exploration in the hope of uncovering something unanticipated.
A term being used more and more each day, machine learning is a discipline rooted in the computer sciences aimed at making sense of seemingly endless amounts of data (far more than humans or traditional systems can process) and finding patterns and commonalities among our systems and data.
Which Analytics Solution Is Right for You?
Obviously there’s a lot more to each of these terms, but this should give you a better sense of what people are talking about —or what they should be talking about — when they refer to these concepts. And if they’re descriptions don’t match up to your newfound understanding, chances are you’re on the receiving end of a marketing-inspired buzzword pitch.
So which of these mechanisms is right for your business?
Once again, if someone tells you there's a “right” answer or that using one is “always” better than using another, chances are, that someone is a marketer. Or someone who doesn’t understand the distinctions.
The answer, as is always the case, is it depends. It depends first and foremost on the nature of your business and the nature of the outcome you’re looking to achieve. No matter how fast the number of buzzwords grows, the starting point for any analytics initiative remains the same: a clearly defined business question.
If you just want to know how much your top customers transacted online versus in person in the last 12 months, a BI solution will probably suit your needs. If you want to understand why those customers did what they did, you’re probably in the market for a business analytics solution. If you want to understand how those customers are likely to behave in the next 12 months, you’re in need of advanced analytics.
And if you’re not sure exactly what you’re looking for, but feel certain your company’s huge volumes of data holds insights, data so large it can’t possibly be processed by traditional means, you may be ready for a data mining or machine learning approach.
The important thing to remember is that it’s not about the technology or the capability — it’s about you and your business needs. No matter how many buzzwords the analytics industry gives us, that will never change.
Join me next month as I share some best practices to prepare your organizations for all the different analytical emanations.