Big data can deliver competitive advantages to financial planning and analysis (FP&A) teams—and the organizations they work for—but implementing and harnessing big data remains an aspirational goal for many organizations. In this interview, Keith Bailey, Director of Finance, ACL, discusses how finance teams can leverage the power of big-data analytics and how to get started.
“Big data” and “analytics” have become real buzz words in the past couple of years—is all that attention warranted? If so, what does it mean for finance teams?
I generally find that FP&A teams have heard about the potential of big data analytics and many are self-consciously pressuring themselves to take advantage without stepping back and asking themselves what problems they’re trying to solve. Many organizations are somewhat late to the party: for years, marketers, auditors, sales organizations, and many other business professionals have been effectively exploiting patterns in big data to wring greater customer acquisition from expenditures, or detect patterns in financial control breakdown for remediation. Finance leaders are often the issue—they just aren’t interested in, or pressing for, the intelligence gleaned from big data.
Finance teams sit at the confluence of business strategy and transactional data and are actually uniquely positioned to transform their roles, detecting patterns, and confidently leading the effort to both address risk and press ahead with strategy.
If finance teams aren’t already using analytics to leverage big data, where should they start?
It’s a bit of an indulgence because it is sitting there in plain sight but by far the easiest and most valuable way to get started with a big-data risk analysis/opportunity illumination technology like ACL is payments data. It doesn’t matter how tight your financial controls are, or how scrupulously-shaped your ERP processes are; I’m here to tell you that if your organization sees a volume of transactions measured in the millions or billions, there is a 100% certainty that you have fraud, waste, and policy abuse material enough to be concerning.
Analysis of corporate credit card usage is typically fruitful. For commercial businesses with a sales force, regression analysis on the sales pipeline is a fabulous opportunity to add value to the revenue forecasting effort and really bring data analysis to bear on strategy, as is analyzing patterns in customer data.
Once you’ve proven the analytics ROI for a small tactical problem, what’s next for the finance team?
If you follow my advice and make your first foray into data analysis impactful by focusing on low-hanging patterns in payments data, you’ll find you readily get a mandate to look further. Critically, in order to move on to more strategic analysis, you will need to pursue continuous monitoring enabled by technology that lets you crunch massive volumes of data routinely and without your intervention. Exceptions, errors, breaches of policy or outright incidents of fraud need to be remediated automatically. Ultimately, predicting the future is where it’s at when it comes to FP&A’s role in big-data analysis.
Partner with leadership to understand what is important to them and simply start collecting data that nobody else is looking at. If managing cost is an important issue—and it usually is—sources of big data might include things as diverse as elevator maintenance data to predict repairs or fleet management data, again to predict repairs or drive efficiency in usage.
What advice can you offer in terms of equipping a finance team with the skills to develop and maintain analytics capability?
This is something I’m really passionate about. Police forces will tell you that the best investigators are those who are naturally curious, people who just have an inclination to seek patterns. And so it goes for data scientists in the FP&A team. Find those people in your organization who show an aptitude and comfort level with data and technology. See how far you can push the envelope. If you are demonstrating strategic value, you’ll be able to bring in people with specialized skills from outside the team.
Meaningful big-data analytics work really does require time so you will have to start by being deliberate about carving out time for someone to focus on it. If a data science team already exists within the broader organization, tell them about the data you’re monitoring and ask for their time. Those teams will relish the access to rich transactional data and the opportunity to add value at an organizational-strategy level.
Finally, you talked about identifying anomalies, how do you make sure that dealing with those anomalies doesn’t overwhelm other departments?
You are the experts at determining what is actually impactful to the business. Ask yourself “how big of a problem is it really?” Be confident. As I mentioned, take advantage of technology that streamlines both the detection of patterns/exceptions and the actual remediation.
When you have observed what looks to be an exciting pattern, don’t just run to leadership with your observations. Usually, the greatest benefit is in going to the teams, the departments, the people who are living in the front lines of creating the data you’re looking at. Bring them opportunities and highlight for them the risks in what you’re observing. Then, together, take your solutions to leadership for maximum impact. Imagine joining customer service leadership to present a new strategy that could materially improve customer retention.
If you’ve noticed a pattern that indicates strategic-calibre risk, make no mistake: taking the subjectivity out of your argument by presenting analysis of real, transactional data flowing through the business will earn you a seat at the decision-making table.