Skip to content
Big data

How Big Data is bringing the back office to the forefront

Brian Peccarelli  Chief Operating Officer, Customer Markets

Brian Peccarelli  Chief Operating Officer, Customer Markets

The data geeks may eventually take over the world. But before that happens, they are going to first need a seat at the table with senior management. Here’s why that’s a good thing for big business.

By now we’re all familiar with the promise of big data. In theory, the ability to collect, process, and analyze vast amounts of discreet yet interrelated data points on everything from customer purchase patterns to supply chain logistics can give businesses the insights they need to increase efficiency, ensure compliance, and reduce fraud.

For many of us in the C-suites of large corporations, however, we’re also intimately familiar with the very real hurdles that exist between the current state of many business operations and the promise of big data nirvana. Getting different data collection and analysis technologies from different vendors to speak the same language; getting data scientists and frontline managers to speak the same language; and finding the right path from raw data to actionable insights are not small achievements. They require a great deal of forethought and a highly choreographed strategy to execute.

Fortunately, several pioneering businesses have already gone through the growing pains involved with this process to start bringing viable big data initiatives into the mainstream. While all have taken slightly different approaches, the common bond linking those that work the best is a clear line of communication between business-facing executives who are ultimately responsible for bottom line growth and data-facing scientists whose job it is to design analytics that help further that growth.

That’s not always the most intuitive process.

Data science is after all a science. Running a business is not always so scientific.

This dichotomy presents an opportunity for senior leaders to nurture a data translator who serves as an intermediary between the hard core quants and the senior management team. This person can either be a dedicated chief data officer role who has dual responsibility for data analytics and business growth or a senior technology leader who sees the full business picture. Whatever the title, the key is to develop a role that is equally facile at speaking quant and understanding the business needs of customers.

One example from my world of tax software and services where that balance between real-world business need and data science magic come together nicely is the tax auditing process. Auditing, by its very nature is an extremely labor- and data-intensive process. Add the complexity created by the litany of global tax reforms stemming from the BEPS initiative and the fear factor surrounding compliance with Know Your Customer (KYC) requirements designed to prevent money laundering, terrorism funding, and other financial crime, and the prospect of accurately auditing a multinational corporation’s financials has become almost insurmountable.

In the past, this process would involve an auditing firm – typically a large accounting firm – working through their clients to gain access to detailed financial and operational information. They would then select sample populations, test for exceptions, and extrapolate results based on a representative subset of data.

While that process was effective in the majority of scenarios, it still left the door open for so-called five percent events, or anomalies that were not accounted for in the modeled results, but existed deep in the underlying data.

With new big data analytics capabilities, auditors are able to directly access detailed client data down to the individual transaction level and apply analytics against millions of different touch points to identify risk, benchmark business metrics against other similar businesses, and even kick out insights on how to reduce operational inefficiencies.

By focusing our big data development on initiatives like this, which present both a clear-cut business mandate and a huge data science challenge, we’ve found it is easier to get business and data teams aligned on a common set of goals. We’ve also learned that by focusing the entire effort on producing measurable financial results, we’ve been able to keep the process from veering too far off into the land of the purely speculative.

Increasingly, what we’re finding as we dive deeper into the world of new tech-assisted business processes and artificially intelligence analytics, is that the process of managing those efforts to create practical, real-world solutions is becoming just as important as the innovation that goes into them.