Data infrastructure optimization software
Data integration and quality software
Data availability and security software
Cloud solutions

The CFO’s Guide to Hadoop and Getting Started With Big Data Analytics

Until now, the CFO could comfortably take a ‘let’s wait and see’ approach to big data. Organizations using Hadoop and big data analysis ranged from ultra successful (Google) to companies looking to build success. Without a proven ROI, however, most CFOs couldn’t or wouldn’t opt in. Now that Hadoop and data analysis are paying off (Hadoop is a lot easier, more secure, and better supported), it’s time to get off the fence and into the game.

SSI CFO Blog - 4-18-16

The most common uses CFOs find for Hadoop and big data are fraud detection and prevention, improving revenue streams, establishing better employment practices, boosting customer lifetime values, and preparing the business for changes in industry trends predicted by the data.

Before you get started, it is worthwhile to visualize how Hadoop and big data will benefit your organization. To the CFO, big data is most useful in the areas of identifying revenue streams, evaluating employee performance, guiding new hire practices, foreseeing and planning for industry trends, and identifying fraudulent activities. Here are some real-world examples:

  • Customer A buys 10 large widgets and 5 small widgets every three months. Using this, you can improve your supply chain and inventory practices, as well as boost the customer’s lifetime value by offering a discount to order more frequently or to increase the amount Customer A orders on a regular basis.
  • Customer B used to pay their bill every 30 days. Now, payments come every 90 days. The savvy CFO can use this data to develop ways to help struggling customers, while also getting that much-needed revenue in a more timely manner.
  • Employee A hasn’t taken a vacation day or a sick day in over 18 months, even though her habits over the past three years were to take one week in the summer and another right after the holiday season. Employee A hasn’t upped her game, because her productivity levels are actually declining. Your data points to the reasons for the changes: Employee A is likely to be involved in embezzling money from the company, and she refuses to take off for fear that her illegal activities will be discovered in her absence.
  • According to your data on industry trends, another recession is heading this way. Your data lets you know which of your products, incentives, and special offers keep customers coming in even when money is tight. You are able to get sales and marketing on board, and your company hardly notices a downtick in revenue while your competitors struggle for customers and can’t afford to keep their full staff.

These are but a few ways that data can improve the performance of the CFO, as well as their value to the organization (think job security!)

Where to Get Big Data

Okay, great, you’re sold. The problem is that your data is strewn about the organization, locked up in data silos and separate systems, all in different formats. That’s where data integration comes in. Data integration requires collecting the data from all of the disparate sources, reformatting the data for processing, and building an architecture to house the data, while keeping security at the forefront. You will also need to develop applications that make use of the consolidated pool of data you create. This usually involves offloading mainframe data into Hadoop, which for many organizations includes important reference information needed to make sense of enterprise-wide data when collated and analyzed.  It could also involve integrating other back-end systems, as well (ERP, marketing automation software, data from IoT devices, etc.) In addition to getting enterprise-wide data, including mainframe, into Hadoop another challenge for your IT organization is to get a handle on the rapidly evolving ecosystem of Hadoop tools and technologies – which is why it is important to choose vendors as partners who architect their solutions to future-proof your implementation.

How to Get Started with Hadoop

SSI CFO Blog - Elephant

Hadoop can accept unstructured data, which means you can include data from sources like mainframe transactions and logs, email, documents, and back-end systems with odd formatting like ERP and CRM software.

Hadoop is a significant undertaking, but yields tremendous benefits to the CFO willing to take it on. Begin by querying the data for the most pressing issues facing your business. Make sure your goals are aligned with those of the other C-level executives. Be clear about the scope of the Hadoop initiative — begin by setting goals for the ROI and establishing a timeline for having certain phases of the operation complete. Just don’t hamstring your team to the point that they’re more worried about budget and time constraints than they are the value that Hadoop and big data can deliver.

Finally, build an Hadoop infrastructure that is capable of scaling upwards as your data undertakings grow. Eventually, you will want to migrate all of your data and data streams into Hadoop, so build an infrastructure that can handle your organization’s volumes of email, documents, video and audio files, back-end systems, mobile technologies, data derived from cloud applications, etc.

With careful planning, you can keep your CapEx and OpEx expenditures on big data at a minimum while maximizing your return on investment in terms of better operational and business intelligence, improved employment practices, better customer acquisition and retention rates, a new level of protection against fraud, and much more.

 

Related Posts