ADP's Chief Technologist, Jim Haas, explains how cluster technology is one of the technologies turning business analytics into products, and profit.

What if there was a technology, one that existed right now, that could slowly turn the darkest corners of a corporate server system into valuable, salable products? Technologists have a lot of clout in today's business world but, even for them, that sort of alchemy would be impressive indeed. Yet, that's what Jim Haas, Chief Technologist at ADP, does for a living, going beyond simple business analytics to provide exciting new opportunities. The tool he uses to achieve this is within the reach of just about any organization.

Big Data Made Practical

As with so many revolutionary modern technologies, this one all begins with data. Whether the database in question is filled with data from app-use monitoring or customer behavior tracking, it simply needs an enormous volume of structure or "unstructured" data — that is, files can be in either ordered or haphazard order, highly manicured or much like the form in which they were originally collected.

"We get the data and look at it and see what kinds of products we can build out of it," Haas says. "We use all sort of techniques, but primarily we do that work with Hadoop clusters."

Thankfully, a Hadoop cluster is a fairly simple idea, just a very large collection of relatively low-cost computers that work together to analyze enormous, often unstructured data sets and produce insights. The true potential of cluster computing isn't in any one particular hardware setup, but rather in the applications older hardware could see.

The idea is certainly ambitious, meant to streamline everything from payroll to HR to time management, but it's also realistic thanks to the source of the insight itself. "The data is there already," Haas explains. "Cluster analytics simply turns it into revenue."

Clustered Insights

With powerful computer clusters for business analytics like the one helmed by Haas, previously worthless databases can actually start to drive profits. That's because hidden in these databases are insights valuable enough to spark major interest from large, prominent organizations, as The Washington Post notes, even large organizations can see a significant ROI on this sort of analytics, since they also have the largest volumes of lead-like data waiting to be mathematically transmuted into business gold.

Haas refers to this as "building a new revenue stream off one that didn't exist by taking data and transforming it into something that's useful in the outside world." One of the best current examples of this sort of application is payroll.

When making hiring decisions, and designing offers for desired candidates, it can be extremely difficult to keep up with compensation standards for top-tier positions. With the right data set and analytical technology behind them, however, organizations could produce an up-to-date report on what competitors are doing, and how it's been working out for them.

"You get a benchmark of what other people are doing and what's the norm," Haas says. "That's hard to know in a fast-changing environment." This exact sort of analysis could be applied just as easily to determine which physical location is best for a new facility, or even to point out that the overall cost-benefit ratio of that facility is inferior to simply outsourcing the entirety of the work.

Turn Cost Centers Into Profit Centers

Since, unlike many types of business analytics technology, cluster computing can happily crunch through unstructured data, it requires far less work in advance — it requires less planning ahead of time, and less time spent massaging the data into the perfect form. This means that just about any company older than a few years can expect to have at least some collection of information that could be telling them more than it currently does, as noted by Computerworld UK.

And it's this ability to create all-new sources of revenue, as opposed to augmenting existing ones, that gives this sort of analysis such incredible potential. Further, modern computer clusters are quick enough that they can start to produce near real-time insights, meaning that even previously available information could be newly valuable when it can be generated with extreme timeliness.

Current cost centers, portions of a business that consume funds but do not directly produce it, could ironically have the greatest volume of potentially salable data, allowing experts like Haas to "take cost centers and turn them into profit centers." And of course, cluster computing is now also itself a service, allowing businesses to easily commission experts like Haas to use cluster computing to turn leftover data into new all-new sources of revenue. There is now no viable excuse not to turn a business's least valuable data into one of its most valuable resources.

Stay up-to-date on the latest human capital management insights for finance leaders: subscribe to our monthly e-newsletter.

Tags: compliance predictive analytics