IT leaders trying to apply analytics to run their businesses better should listen to Immelt's advice, based on his experience leading GE's Industrial Internet vision.
General Electric CEO and chairman Jeff Immelt has made a big bet on data analytics driving his company's growth, and on Wednesday he shared some hard lessons he has learned over the past four or five years about using and building analytics software. Immelt didn't lay those lessons out in a Letterman-style top 10 list; instead, they just sort of trickled out throughout a two-hour customer presentation, where Immelt and GE customers discussed the company's latest thinking about the Industrial Internet, which is GE lingo for the Internet of things.
Immelt's observations strike me as relevant for IT leaders thinking about how they'll apply analytics to run their businesses better. GE is trying to sell software that helps people more efficiently use the jet engines, wind turbines, CAT scans, locomotives and other industrial gear the company makes. You're implementing analytics software for a variety of different reasons. So here's my take on what Immelt said and how it applies to you:
Analytics Changes Expectations
A decade ago, GE was in the mode of "the product breaks, we fix it," Immelt said, acknowledging that such a mentality can lead to "misaligned interests." Today, GE has more than $100 billion in revenue tied to service contracts, whereby it gets paid based on a product -- a power plant turbine, a jet engine, a locomotive -- being in service. It needs analytics software to help customers avoid downtime and thus make those contracts profitable. "We don't see GE being a software company per se. It's about how we enhance our customer relationships," Immelt said.
Make no mistake, GE is selling software. This week it announced it has added 14 new industry modules to its Predictivity line of software, adding to the 10 it had to manage and analyze industrial data. It sells Predix, a data management platform (much of it using NoSQL approaches) designed to store and use machine-generated data. It just announced new partnerships with AT&T, Cisco and Intel to advance the networks, sensors and on-machine computing needed for the growth of industrial data. But most of that effort is focused on upselling GE's huge installed base. To do that, GE must convince customers that its software can help them wring more production out of a piece of GE equipment.
Be Practical About The Value
Last year Immelt's message was all about the macro impact of the Industrial Internet -- predictions of up to $15 trillion in economic gains in the next 20 years from companies making smart use of sensors, networked machines and data analytics. This year he has a micro marketing message: "no unplanned downtime" due to GE equipment failures. Industrial Internet is the big vision, but "no unplanned downtime is what counts for customers," he said.
Immelt was quick to point out "we're not there yet" on delivering zero downtime. It's a goal for the Industrial Internet, just like helping to drive that forecasted $15 trillion in economic growth. But if you're asking a CEO to envision ultimate success, is it easier to picture $15 trillion in economic growth or an airplane that's never out of service for a maintenance failure?
Be Humble Enough To Learn
GE execs got to the top of the heap by knowing how to sell products -- often highly engineered, highly sophisticated equipment. Software is different, Immelt warned, and GE had to retrain its execs on how to do business unit reviews related to the software and analytics component.
"I'm a 30-year GE guy. I used to run this [healthcare] business. I can sit down and have a CT review with the best of them," Immelt said, pointing to an imaging device on stage. "On analytics and software, we had to retrain the leadership team -- basically the big iron team-- and we had Bill [Ruh, VP of GE's software and analytics center] and his team come before the leadership team and say, 'Here are the 10 questions you ask when you're doing a review.' So we all have to relearn stuff. Don't be too proud."
GE hasn't stopped bragging about its industry experience, since that knowledge is why it thinks it can create useful software for industry players. But Immelt provides a warning that execs who know existing products and processes might not see the opportunity that new data and analysis options create.
Analytics Can Lead To Awkward Questions
With equipment, GE could be the expert about what it's capable of, or how to approach repairs. It has experts who can tell a power plant engineer everything there is to know about running or fixing a turbine. Predictive analytics results are much more open to interpretation, and in fact are only valuable if they're shared with people capable of acting on them -- often by challenging the status quo.
"Products tend to be closed systems; analytics are open systems," Immelt said. "So you have to have this transparency with the customer, and the data has to be in place, and that's uncomfortable for us. We grew up in the industrial world where it's basically 'we have 200 patents on every CT and you can't touch it.' You have to get your team ready for that."
New Skills Are Needed
GE has long prided itself on being able to manufacture outstanding business leaders with the same proficiency that it cranks out power plant turbines. But Immelt said GE needs to bring in a new kind of talent in software and analytics. "We never made real progress here until we brought people in from outside GE," Immelt said, referring to its analytics effort. "One of my lessons learned is the importance of really good people who have different insight into how this world really works."
6 Tools to Protect Big DataMost IT teams have their conventional databases covered in terms of security and business continuity. But as we enter the era of big data, Hadoop, and NoSQL, protection schemes need to evolve. In fact, big data could drive the next big security strategy shift.
Big Data Brings Big Security ProblemsWhy should big data be more difficult to secure? In a word, variety. But the business wonít wait to use it to predict customer behavior, find correlations across disparate data sources, predict fraud or financial risk, and more.