Big Data & Analytics in the Supply Chain
Big Data, Big Data, Big Data, is arguably the most overused, least understood, term floating around the business world today. Add in the sweetness of Analytics and you have two great tastes that taste great together, but are for the most part, to most people, empty calories. These are two topics that everyone talks about but very few do anything about them or with them. Why? Because we all know we should be doing something with this to help our businesses, but very few know what to do.
Big data has always been around, we’ve just never had any really good way of capturing it or storing it before. A good example of that is the electric meter on your home. It was for the longest time a very dumb appliance that was really good at its one task. The wheel spun around and the dials moved and once a month the electric service provider, formerly known as the utility, would read the meter and send you a bill. Twelve meter readings a year. Twelve points of consumption data annually per customer record. A good sized utility with 1,000,000 customers would have 12,000,000 data points annually to analyze. That might sound like a lot, but by today’s standards it’s nothing. Oh, and by the way, all of this data was stored on very expensive cylinders on a mainframe.
Your home now, most likely, has a very smart meter attached to it. That meter captures consumption by the minute. Given that there are 1,440 minutes in a day that translates to 525,600 data points per house per year. Across that same 1,000,000 customer base you now have 525,600,000,000 data points annually to analyze. If you thought 12,000,000 was a lot, well it’s seriously time to rethink that. And all of this data is stored on cheap hard drives or in the cloud. But even though it’s big, there’s really no more information in those 525,600 data points per house than there was in the original 12. Data, big or small, is not information.
Okay, so we’ve established that there’s a lot of data out there coming from places we’d never expect it to come from. So what? Pretty much, so nothing. Having lots and lots of data, dare we say, Big Data, really doesn’t get you anything. Data is not information and data is typically not actionable. To change data into actionable information you need Analytics. Analytics is a scary term for some people as they imagine data scientists in lab coats and wild hair finding out everyone’s deepest darkest secrets. I’m not saying that’s not happening, I’m saying it’s not as common as imagined. Analytics is simply applying analysis to data in a way that answers questions and helps predict the answer to future questions.
People shy away from Analytics because it seems hard and involves statistics and numbers and crunching data. That’s why Halo made Analytics easier. Because without timely information you can’t make actionable decisions and if you can’t make good decisions your business suffers accordingly. Key to being easier are wizards that step you through the Analytics process and help you select the proper modeling techniques for your data and to achieve the desired results. In addition, data cleansing and refining techniques bring in only the part of your Big Data that you need, creating a much more efficient process.
But what does all of this have to do with the Supply Chain? One of our partners, Zelcom, is a RFID provider. RFIDs in the manufacturing process produce all sorts of real-time data and lots of it. All of this data is great, but again, it’s just data. By using analytics you can create actionable information. That information allows you to make better business decisions. If you were able to track specific pieces of rework to certain operators or machines, you’d know that training or repairs were needed and you could take the proper action. If you knew what color and sizes of products were completed and being shipped, you’d know how many orders were being filled in full and on-time. If you needed to create a completely traceable production lot, you’d have all of that information. By having the Analytics necessary you’re able in real-time make proactive decisions that affect your upstream and downstream supply chain.
With all of this RFID generated data, you no longer have to guess about any stage of production, material delivery or finished product shipments, it’s all available to you if you have the capacity to turn this data into information and turn the information into good business decisions. These business
decisions and corresponding supporting information can be collaboratively shared with internal and external customers to insure the most effective supply chain possible.
There can no longer be any complaints of not having the data. It’s everywhere. There still may be complaints about having the right data. But most of the time, when you hear someone say they don’t have the data, what they probably really mean is that they don’t have the Analytics to do anything of value with the data they have.
Great article John. I've seen too many people buy tools that don't get used. Funny the image of analytics being in the hands of "wild haired data scientists" but they probably aren't the consumers of the data. With new information being provided who is going to use it? And on the other side of the fence what questions can you answer with this new data and analytics? As a Performance Improvement consultant I like to have people get regular data on process performance. I've seen senior management teams with complex operations miss quarterly expectations because they just weren't on top of their operations and the data was all there to predict it (if not the information). How can a customer and solution provider make sure that the Analytic toolset and information actually get used? (in my case I'd like to make sure those key process health metrics are being watched by the accountable consulted and informed individuals).