GE bets big on its own cloud operating system, Predix

Every year around this time, executives from some of the biggest U.S. industrial companies fly south to Florida for their annual confab.  This year, one would think they all mistook Longboat Key for Silicon Valley.The big buzzword at this week's gathering was software: Just about every presenter talked about ways to use technology to make their airplane engines and assembly lines smarter and more efficient. Some plans were more fully baked than others, with General Electric CEO Jeff Immelt memorably commenting that some of the digital strategies trotted out by his fellow presenters were "B.S." (He didn't name names.)

Pot shots aside, there is a notable difference in the ways industrial companies are approaching the Internet of Very Big Things, even among two of the biggest presenters: GE and United Technologies.

GE is taking the most radical approach. It doesn't just want to help make its own wind turbines and jet engines work better -- it wants to sell tools to rivals to help them do the same.

The industrial giant has spent more than $1 billion developing its cloud-based Predix software platform, which is essentially an operating system akin to Microsoft Windows or Google's Android. It works by collecting streams of data via sensors on industrial machinery and using that information to analyze the performance of that specific product and make predictions about future market trends.  GE is betting Predix will help it transform itself into a top-10 software company in the same league as Oracle and SAP. 

No other industrial company is trying to do anything like that -- at least not yet -- and by the time anyone catches up, GE may be light-years ahead. Most of its peers are still in the stage of developing internal digital strategies for their own products. United Technologies falls into that camp, with CEO Greg Hayes saying: ``We're not going to be a software company. We're not going to sell software. That's not what UTC does. We sell high-technology solutions to our customers.''

It's a subtle distinction terminology-wise, but a big discrepancy in terms of business models. Hayes' comments signal that he sees technology as a complement to what United Technologies already does, rather than a business in and of itself like GE does.

It's not totally clear that GE's plan will be a slam dunk. The big question will be whether it can get manufacturing rivals to trust it enough to hand over control over their data-management operations. For now at least, that doesn't seem to be a problem. GE's digital business had $5 billion in revenue last year and is on track for at least $6 billion in 2016. By 2020, GE Digital could make up as much as 25 percent of the company's earnings, according to Nick Heymann of William Blair. And boy is it profitable. On an Ebit basis (earnings before interest and taxes), margins for the software piece of GE's efforts will be around 35 percent, while those on the data analytics side could run in the range of 50 percent, Heymann said.  By contrast, GE's aviation unit -- its most profitable currently -- had an operating margin of 24 percent in the most recent quarter. 

Looking at those numbers, it begs the question of whether United Technologies should be thinking more seriously about developing a software platform of its own. That's revenue growth and profitability gains that the company is potentially missing out on at a time when both are hard to come by these days for industrial companies. The reality may be that CEO Hayes has too many other problems to worry about -- from managing the launch of a costly new jet engine to recovering market share in the elevator market to just getting sales to grow -- that he can't prioritize building a software business from the ground up right now. But an opportunity lost is someone else's gain.