By Richard Parcia
I?ve had enough. Every person in IT circles keep invoking this word. Aside from the sexy terminologies like Internet of Things (IoT) or Digital Transformation, no word has gotten so much bastardization in the last five years or so as the word analytics. From vendors to CEOs to HR folks, everyone is invoking the word when the topic of the ?Future of IT? is brought up.
Frankly, I don?t mind it when it is mentioned by some business owners I know. In fact, I also share their enthusiasm. One prominent businessman friend of mine had it as a topic in one of our conversations and it was a pleasure to discuss with him the potential of the field.
Aside from imagining what it can bring, the gentleman knows his hard math. If there is such a thing as a credible conversation, it was accurately one of those.
However, when it is conjured by those who are obviously trying to sell a product or worn up in their sleeves by IT social circle butterflies, it can be amusing to the point of irritation. It used be pure amusement. It slowly got converted in a wry smile and one raised eyebrow. Nowadays, it?s irritating.
Some months ago, I was featured in a magazine published by one of our major telcos. I appreciated the opportunity, and was tickled by the thought of getting my mug published on a glossy material. The topic was analytics and I tried to explain to the interviewer what is it all about and the future of it.
The story that was published later on didn?t really capture the essence of what I was saying (I sounded like a duffus). In fairness to them, they did give me editorial latitude (truth is, I was so busy to read it before I said ?go?).
Looking back, I realized that the reason why the editors of that magazine didn?t capture what I said was due to the fact that they themselves hardly knew anything about the topic. They know it?s the ?in thing?. But I suspect their only knowledge of the topic was due to Google.
Allow me to straighten things. First of all, let me state it: Analytics is NOT new.
Analytics as a field has been around for decades. My first foray into it was in the late ?90s when I was a graduate student under the tutelage of Dr. Santiago M. Alviar of UP Los Banos.
It wasn?t a deliberate discovery. While I was writing my thesis for my MBA, I was practically penniless and couldn?t afford to pay a statistician to do my data analysis. So my strategy was to get one of the best in the field, thinking that he would it for me — or so I thought.
It didn?t work the way I planned it. Instead of doing the analysis, Dr. Alviar told me to study it on my own, and since I was a full-time assistant instructor at that time in a local college, he practically ordered me to volunteer myself to teach the subject matter because he believed that ?if you want to master something, teach it.?
Suffice to say, I devoured the subject matter, and has used all of those tools and valuable advice from my old advisor — 20 years ago — all throughout my professional and academic life.
I lost count the number of times I used it to make decisions, present information, simulate, and geekily play with data. And in my part-time teaching gigs, I lost count of the number of graduates (Bachelors, Masters, and PhDs) who received their degrees because I prepared, processed, and presented, their data for FREE. Yes, I did it for free. It was an unstated promise I made to pay it forward for what my mentor taught me.
By the way, in a short-sighted move, a lot of MBA schools some years ago removed the ?thesis? requirement from their programs. Using the reason that the professional field will have no use for a ?thesis? (I disagree), they removed it because it was widely known inside the academic circles that a lot of students fail to graduate because of that ?thesis?, particularly, that hardly understood, chapters 3 and 4. Those chapters of an applied research (the common MBA thesis then) in the old programs are practically analytics (more on this in a future column). Now, I won?t call myself an expert but like my businessman friend who was a former math professor, I know my numbers. But I digress.
So what is analytics? It?s simply applied statistics. The range of its application range from the simple descriptive type to the multi-variate ones. If you read your online materials, they will say it?s a combination of statistics, computer programming and operations research.
Frankly it?s not a good definition because, while the fundamentals of statistics and operations research are identical (it?s math, after all), computer programming only comes to the picture because of one primary reason: It was only feasible to process large data with computers.
It?s almost humanly impossible to process multi-variate statistics manually. That is the reason why if you visit the research libraries, where you find all those old theses of our old professors, you will hardly find anything beyond the T-Test as analysis tool.
Even then, doing that took hours and hours to be analyzed from sheets of Manila paper pasted on walls and books of general ledgers. It was that tedious.
However, those who had access to computers were doing a lot of number crunching. They were the ones who studied models and verified them. They painstakingly tallied the data in what you called datasets (its a spreadsheet type file) and ran them in those computers to produce the desired information. They had been doing this for decades.
I am familiar with the use of advanced and sophisticated efficiency methods like DEA or SFA that had been a staple of researches that were done in UP Los Banos and UST. I?m quite sure the other universities with a good research culture are doing the same.
Now the question: So if the universities had been using this, why only now? Because beyond experiments wherein most of the data is controlled, there is hardly no value for business at all.
Analytics, to be valuable for business, must have the element of timeliness. Without it, you cannot make decisions swiftly. In the university, time is used for discoveries and wonder. In business, time is money.
Before, even with the right tools and the brains, analytics was hardly an option. For example, if you are a fast-food company or a mall with multiple branches nationally, you will only get the accurate reports after being compiled in boxes and sent over via plane or ships.
If you were the yet-to-be-invented analytics head, you will only receive it in a week at best, and you will still need the time to ?tally? the data of all those branches. By the time you produced the information, the data is already stale. It won?t make sense.
There is more to this topic and one column will not cover it. At the very least, I hope I made it more understandable to folks. A lot of companies are buying ?Analytics Solutions? from vendors and consulting firms but hardly understand what it is or the use for it.
In fact, I know of a huge local retail company who has one of the best analytics software in their environment but use it only to produce descriptive statistics. They don?t know that their own ERP, if properly setup, can actually produce the same. Why buy a multi-million solution when you are getting an output that can be produced by using a spreadsheet?
So there. Next time somebody invokes this sexy topic of analytics, it won?t be a vague topic of discussion. Don?t be fooled into buying something just because they say it?s the way of the future. Worse, don?t hire somebody who keeps invoking it as if it?s a new thing. Most likely, that person does not understand it, what?s more, even practiced it.
There is more to analytics than the pie and bar graphs. It?s both an art and a science. And, well, mathematics. It?s beautiful. Understand it properly and it can be powerful.
The author is an associate professor of information systems at the Graduate School of the University of Santo Tomas in Manila. He has held various global leadership roles at Intel, Unitedhealth Group, and LafargeHolcim. Currently, he is a senior consultant for an international consultancy firm based out of Paris, France. You can reach him via LinkedIn
