What if the LOB had digital democratization

4 recommendations to set your IoT analytics strategy on the best course

abstract chart graph analytics

Everyone is enthusiastic about the flow of insight that will be unleashed by big data and the internet of things. CEOs look forward to new strategic and competitive insights. Sales and marketing managers can’t wait for the precision they’ll see in measuring where, when, and why new product or service rollouts are succeeding. Investors are seeing new opportunities, new ideas, and innovations by the dozen.

Even better, this influx of opportunity is coming at a time when the enterprise is ready. It’s becoming quickly digitalized. Clouds are rolling in from all directions, bringing IT independence to LOB (line-of-business) managers, finance managers, and all others. It’s a type of digital democratization, and it extends all the way to the analysts—now called citizen analysts—who are starting to mine some of that IoT data.

Gartner was on target when, in 2015, it predicted that by 2017 most organizations’ business users and analysts “will have access to self-service tools to prepare data for analysis.” The data is coming, and it’s ready for mining.

But as a business unit manager, are you ready? No one should blame you if you’re less than enthusiastic. The job of actually pulling meaning from the coming torrents will be far from simple, and success far from guaranteed. As Gartner’s VP Rita Sallam said, “Data preparation is one of most difficult and time-consuming challenges facing business users of BI and data discovery tools, as well as advanced analytics platforms.”

Gartner also warned of some disruptions that this analytics democratization might cause. These include breaches in privacy and security, and, in general, a kind of analytic sprawl that would lead business managers and citizen analysts in too many directions at once. 

To guard against these dangers, and to set your IoT analytics strategy on the best course, I offer four key recommendations:

1. Gain executive buy-in

This is true of any boundary-stretching project, and even more important when the project’s goals may be loosely defined at the start. You know that the right implementation could pay off handsomely at the bottom line, but you can’t promise specifics. Your ability to communicate the value of big data/IoT analytics is vital to getting the support—and funding—you’ll need. Building an effective project prototype may help gain executive support, especially if your process and results can be conveyed visually to the C-level and other LOB managers.

2. Prepare citizen analysts for their new roles

The term citizen analyst can cover plenty of ground between business expert and data scientist, and can include business-savvy power users, developers, and other professionals. Ideally you’ll find a mix of job titles on your team, each informing the other to the betterment of the entire team.

Seek out and nurture key skills in team members, such as creativity, enthusiasm, communication skills, and, importantly, awareness of how their work will fit in with the organization’s strategic business goals. Do what you can to set and maintain the conditions for success. Encourage open debate in the name of creativity and challenge them to use the data to prove a point, even if it means challenging you personally.

3. Call on IT for governance

Realize that the role of IT is changing, from that of report generator to one of advisor, or enabler, of the newly democratized analytics effort. As IoT analytics become widespread in your organization, it will be vital to maintain a strong but agile governance capability, and this is best done from a centralized source.

Call on IT to help in a number of ways, from data cleaning and validation to helping create consistent reporting styles for use by analytics groups across the organization’s various units. This will be valuable when, for instance, multiple business units want to collaborate on IoT analysis for presentations to the CEO.

4. Consider the architecture before the analysis

It may be tempting to sign up the newest cloud-based application for a specific, immediate analytical need, but beware: this can cause way more trouble than it’s worth, as multiple small clouds take over your unit’s formerly tidy architectural landscape. A so-called point application may look appealing on the surface but realize that it’s the underlying architectural platform that will do the heavy lifting. Choosing the right platform will pay benefits for years ahead, and will prepare you for future evolutions, with even greater quantities of data than you’re seeing now.

Copyright © 2018 IDG Communications, Inc.