Blogs

Look at Big Data in Different Ways to Find Business Value

There are many ways to view big data, but which way reveals the business value possibilities?

Owner, Forest Rim Technology

There are lots of ways to look at big data. One way is to look at the sheer volume of data. Some types of data are more voluminous than others. The trading of stock on any of the stock exchanges is one example. When looking at every trade of every share, a day’s worth of trading would result in a huge number of occurrences of data. Or look at meteorological data. Consider the huge amount of data collected every time a temperature in any location changes by one degree. There are undoubtedly many other types of changes that would result in multiple instances and tremendous volumes of data.

Another way to look at big data is by age. Some data is very fresh, and other data is old. For example, if you look at the sales made by Walmart since 1965, there would be quite a lot of data. And the vast majority of it would be legacy data. Another similar consideration might be records of genealogy. Tracing a family tree back a thousand years could lead to unearthing some very old data.

Yet another way to look at big data is by the method of its collection. Consider the tracking of railroad cars by an analog computer. The computer sits in a farmer’s field, looks at the train cars as they pass by, and records which cars have been in transit. Other data is gathered by satellite. The satellite looks at patterns of clouds and ocean currents and other measurements of the earth—such as oxygen levels in the ocean.

And there are many other ways data can be collected. For example, data is often gathered as a result of transactions. A purchase made at a Walgreens pharmacy or at Safeway, for example, is placed in a database somewhere. And nurses and doctors register the birth of a child every time a baby is born in a hospital.

Take another look

Although there are many other ways to look at big data, take a different look from the perspective of business value. Where can business value be found in the pools of data that exist in a corporation?

Arguably, there is big data value everywhere, if one looks hard enough and is smart enough. And certainly some amount of business value can be derived from almost any collection of big data. But some big data is rich in value and other big data is bereft of value. So where is there a lot of business value in big data, and where is there relatively little business value in it?

One place to start is by looking at clickstream data. There’s some business value there, but the problem with clickstream records is that there are a lot of them and finding those few records containing business value is not easy to do. What about telephone call detail records? There are literally billions of call detail records and a few of them have business value. But not many of them do, so there is some business value there. However, there aren’t many records that have value in proportion to all the telephone calls that are made.

Now consider the analog records made by a manufacturer as the product comes rolling off the production line. There are a lot of records that are created by the analog computer. And, sure enough, a few of them have some business value. But like telephone call detail records, not many analog records have value, and they are hard to find.

Looking elsewhere for business value in big data, how about customer call center records? These records are created every day as customers chat with representatives of the company. Almost every call center record has business value—a lot of it. Then there are warranty records for the manufacturer. Suffice it to say, every single warranty record has a lot of business value. Feedback on the quality of manufacturing is perhaps the most valuable business information a manufacturer can get.

And then there are medical records. When it comes to research and understanding a disease afflicting a population of patients, there is outstanding value to be had in every medical record. Those medical records can be the key to life and death for patients ailing with the disease and its symptoms.

Look no further

When analyzing what is going on with big data, a large pattern emerges. Those big data records that are repetitive—in which one record looks very much like the next record—don’t hold all that much business value, or the business value is at least hard to get at. In contrast, those big data records that contain nonrepetitive data—where there are conversations, dialogues, and narration—are teeming, oozing, even bursting with business value. The conclusion to be drawn here is that anyone interested in deriving business value from big data should look no further than its nonrepetitive data.

Please share any thoughts or questions in the comments.

 

[followbutton username='IBMdatamag' count='false' lang='en' theme='light']