Sie sind auf Seite 1von 16

For the exclusive use of U. Putro, 2017.

IN1395

Digitization of an Industrial
Giant:

GE Takes on Industrial Analytics

09/2017-6313
This case was written by J. Stewart Black, Professor of Management Practice in Global Leadership and Strategy, with
assistance from Anne-Marie Carrick, both at INSEAD. It is intended to be used as a basis for class discussion rather
than to illustrate either effective or ineffective handling of an administrative situation.
Additional material about INSEAD case studies (e.g., videos, spreadsheets, links) can be accessed at
cases.insead.edu.
Copyright © 2017 INSEAD
COPIES MAY NOT BE MADE WITHOUT PERMISSION. NO PART OF THIS PUBLICATION MAY BE COPIED, STORED, TRANSMITTED, REPRODUCED OR DISTRIBUTED IN
ANY FORM OR MEDIUM WHATSOEVER WITHOUT THE PERMISSION OF THE COPYRIGHT OWNER.

This document is authorized for use only by Utomo Sarjono Putro in 2017.
For the exclusive use of U. Putro, 2017.

On a warm summer’s day in 2009, Jeff Immelt, CEO of General Electric (GE), was listening
to the company’s scientists rave about sensors being installed on the latest GE jet engines that
could generate mountains of data. A flight from New York to Chicago, for example, could
produce over a terabyte. 1 While impressed, he wondered how that data could be used. Would
it simply help engineers build better engines in the future? Given that GE developed a new
generation of jet engine only every ten years or so, that seemed a long time to wait. Couldn’t
GE use the data in ways that generated value today? In order to capture that value, Immelt
determined GE “had to be more capable in software.” 2

Ironically, however, the insight about the importance of software came at a time when his
priority was to take GE back to its roots as an innovative industrial hardware company. From
his appointment as CEO in 2001 until the unexpected announcement of his retirement in June
2017, Immelt orchestrated 380 acquisitions, spending over $175 billion on industrial
companies in sectors such as oil and gas and wind power generation; and divested 370
businesses worth a total of $400 billion, including long-time GE businesses such as
appliances and plastics. 3

GE innovative industrial roots began with its founder, Thomas Edison. Though he was best
remembered for inventing the first commercial light bulb, he was granted more than 2,000
patents across a variety of domains – most of them industrial. In fact, it was Edison that
positioned GE as an industrial (rather than a consumer) company with investments in areas
such as the generation and transmission of electricity.

Despite over half a trillion dollars in acquisitions and divestitures, GE’s revenues and profits
under Immelt seesawed but generally trended down, especially after the financial crisis of
2008 (see Exhibit 1). The financial crisis hit GE harder than most other industrial companies
largely because of GE’s direct exposure to it via its massive GE Capital unit, and this was
reflected in its more severe stock price decline (see Exhibit 2). However, its stock’s
underperformance was not confined to this this critical period but persisted across Immelt’s
entire tenure (see Exhibit 3). Indeed there was speculation that this was the reason for his
unexpected retirement in 2017. 4

1 Immelt, Jeffery and Rik Kirkland, “GE’s Jeffery Immelt on Digitizing in the Industrial Space,” McKinsey
Quarterly, October 2015, www.mckinsey.com/business-functions/organization/our-insights/ges-jeff-
immelt-on-digitizing-in-the-industrial-space , accessed 25 June 2017.
2 Lohr, Steve. “G.E., the 124-Year-Old Software Start-Up.” The New York Times, The New York Times, 27
Aug. 2016, www.nytimes.com/2016/08/28/technology/ge-the-124-year-old-software-start-up.html?_r=0 ,
accessed 16 Mar. 2017.
3 Gara, Antonio, “For GE’s Jeff Immelt, Hundreds Of Deals And $575 Billion Didn’t Yield A Higher Stock
Price,” Forbes, 15 June 2017.
4 McGregor, Jean and Thomas heath, “GE’s CEO, Jeff Immelt, to Step Down After 16 Years,” Washington
Post, June 12, 2017.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 1
For the exclusive use of U. Putro, 2017.

Products
At the beginning of 2017, most of the products sold in GE’s main business units represented
major CapEx purchases for its customers. Years of selling such products had generated an
installed base of over $2 trillion for GE (i.e., the replacement cost of the products in use). 5 To
win these sales, its salesforce emphasized their product’s features, superior quality and
reliability, or lower cost compared to competitors. But winning a deal could be a lengthy
process. The higher the price, the longer the sales cycle – which could take six to 18 months
to complete. Although margins varied across GE’s major business groups (Power & Water,
Oil & Gas, Energy Management, Aviation, Healthcare, Transportation and Home and
Business Solutions), they rarely exceeded 35% on product sales.

Product sales tended to be transactional in nature – GE made the product, the customer bought
it and the deal was done. Thereafter, the risks and benefits of ownership resided with the
customer; for GE the after-sales risk occurred only if the product didn’t perform as promised.
Since many products had long life cycles, any potential damage to its reputation and sales
might not arise for decades. GE sales teams primarily interacted with customer procurement
professionals, and with operational executives to a lesser extent. Pricing was fairly
standardized and volume discounts applied within prescribed limits. Often the sales effort was
reactive – in response to a detailed RFP (request for proposals) that customers put out to
multiple providers for tender.

In seeking to reinvent GE, Immelt had more than tripled investment in R&D, but creating
differentiated products was becoming increasingly difficult. Whereas GE had previously
enjoyed a lead in the sciences associated with its products, its knowledge advantage had
decreased over time—not because GE was slowing down but because rivals were catching up
and in some cases overtaking. Albeit sophisticated, GE’s products were increasingly being
commoditized as differences with competing products were shrinking. As is the case
whenever commoditization occurs, GE’s operating margins had been compressed from
approximately 25% when Immelt took over in 2001 to 14% in 2016.

Customer Service Agreements (CSAs)


Although product sales were important for GE, CSAs (customer service agreements) had
overtaken products in terms of total backlog and profits. Jet engines were a case in point. GE
could sell a jet engine only once, but it needed to be maintained and repaired multiple times
over its useful life. Because GE had designed and made the engine, it argued that it could
maintain and repair the engine better and cheaper than customers themselves or other
maintenance service providers. This argument was made not just for jet engines but virtually
all the products it sold. The fact that by 2016 CSAs accounted for 46.3% of sales (excluding
GE Capital) and 74.6% of GE’s $320.8 billion backlog testified to the compelling nature of
the pitch.

5 CNBC. “CNBC Exclusive: CNBC Transcript: Microsoft CEO Satya Nadella and GE CEO Jeff Immelt
Speak with CNBC’s Jon Fortt on ‘Squawk on the Street’ Today.” CNBC, CNBC, 11 July 2016,
www.cnbc.com/2016/07/11/cnbc-exclusive-cnbc-transcript-microsoft-ceo-satya-nadella-and-ge-ceo-jeff-
immelt-speak-with-cnbcs-jon-fortt-on-squawk-on-the-street-today.html , accessed 17 Mar. 2017.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 2
For the exclusive use of U. Putro, 2017.

In most cases, GE proposed the CSA when the product was sold. Most CSA contracts were
standard – a fixed price for maintenance and repair work, with guaranteed levels of uptime,
maintenance turnaround time, and so on. Ideally the duration was tied to the product’s
expected lifespan. Compared to product sales, CSAs were financially attractive to GE because
their long-term contractual nature provided a reliable and steady stream of future revenues,
unlike the cyclicality of product sales.

For many customers, lower maintenance and repair costs were not the only (or even the
major) attraction; the real value was in second order cost reductions. For example, an offshore
oil rig was down on average five to six days in a year and each down-day could cost $5-$7
million. If maintenance and repair could reduce the down time by just one day for a company
with 100 rigs, the customer could save $500-$700 million dollars a year!

Software
When Immelt had his insight about becoming more of a software company, GE was no
stranger to software. In 2010, GE employed over 5,000 software engineers worldwide, 6 the
vast majority working on software that made GE’s products more effective. For example, it
was software that made GE’s MRI machines turn magnetic refractions into the colour-coded,
high-resolution images that doctors used to diagnose injury and disease. However, when it
came to software, each business unit was fairly autonomous and had its own platforms,
languages and development tools – sometimes these even differed between products made
within the same business unit.

In addition, there was a team of engineers who developed software that GE sold to customers
independent of GE products. In 2010, this group generated $2.5 billion (1.67% of total
revenue). 7 Spread over two centres, (one in Michigan, the other in Virginia) even these did
not share a common platform, language or tools, though the differences were less than those
across business units.

The Industrial Internet, Internet of Things, the Smart Factory,


Industry 4.0
While GE’s industrial transformation was playing out centre stage, in the background a digital
revolution was taking shape for which various terms had been invented: the ‘Industrial
Internet’, the ‘Internet of Things’, the ‘smart factory’, and ‘Industry 4.0’. Whatever the name,
the basic notion was the same and covered five steps:
x Step 1: Gather data via sensors from operational equipment; be it jet engines, pumps, wind
turbines, etc.
x Step 2: Collate and store that data on a common platform and in a common format to
allow for rapid, relational analysis.

6 “Bizjournals.com.” Bizjournals.com, www.bizjournals.com/albany/news/2012/01/27/ge-plugs-in-big-


data.html , accessed 16 Mar. 2017.
7 “Bizjournals.com.”

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 3
For the exclusive use of U. Putro, 2017.

x Step 3: Create and use analytic tools to spot patterns, anomalies and events across data
pools.
x Step 4: Create and use tools to project or predict outcomes based on the data collected and
analysed.
x Step 5: Determine actions that optimize process functionality and/or reduce costs, not just
for individual components but also for the operational value chain; create feedback loops
that enable learning (including autonomous learning), refinement of analysis and future
predictions.

How the Industrial Internet and the related industrial digitization and analytics would develop,
or even how big it might be was not yet clear in 2010. Nevertheless, GE was committed to
pursuing the opportunity. Its pursuit was validated by a Boston Consulting Group (BCG)
study 8 that estimated that by 2020 companies would spend €250 billion adopting different
aspects of the Industrial Internet. Specifically, the study identified predictive maintenance,
self-optimizing production and automated inventory management as the greatest areas of
spend and opportunity.

Creating Value
GE’s own analysis produced similar conclusions to those identified in the BCG study.
Specifically, GE determined that opportunities for predictive maintenance and optimized
production were the biggest ones to pursue.

Of the two, predictive maintenance offered the most immediate opportunity. For many of
GE’s customers, unplanned maintenance and repair had significant economic consequences.
If a plane’s engine, for example, needed unscheduled maintenance or repair, this involved not
only the direct cost of fixing the engine but also a major opportunity cost because while the
plane was in the repair shop it was not generating revenue from passengers. Given the high
fixed cost of a plane (and its engines), any loss in revenue meant that all the costs associated
with that plane had to be amortized over a lower revenue base, with resulting higher costs per
unit and lower operating margins – not good for an industry where margins were already thin
and when firms had little pricing power. If analytics could determine the optimal type and
schedule of engine maintenance and keep the plane safely in the air for longer and avoid
unanticipated repairs and revenue disruptions, this would (a) bring in more revenue, (b)
increase output and thereby lower unit costs, and (c) potentially extend the useful life of the
engines. All three outcomes had significant economic value to airlines. More importantly to
GE, these benefits could apply to the majority of its other businesses and customers.

The second, and potentially more valuable, opportunity was process and output optimization.
Many of GE’s customers had operations with high fixed costs, so optimization that increased
throughput or output (independent of that occurring from predictive maintenance) was
extremely valuable. For example, on average 65% of the oil in a well was left in the ground

8 Nicolas Hunke, Zia Yusuf, Michael Rüßmann, Florian Schmieg, Akash Bhatia, and . “Winning in IoT: It’s
All About the Business Processes.” Www.bcgperspectives.com, 5 Jan. 2017,
www.bcgperspectives.com/content/articles/hardware-software-energy-environment-winning-in-iot-all-
about-winning-processes/ , accessed 16 Mar. 2017.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 4
For the exclusive use of U. Putro, 2017.

because the natural pressure to lift it to the surface was used up in extracting the first 35% and
the cost of forcing the remaining oil to the surface was too high. If pumps, pressure, filters,
etc. could be monitored, analysed and optimized so that “lift” could be increased even by 1
percentage point, that would generate an additional 80 million barrels annually, which was
equivalent to all the oil produced globally in a single day. In industries where companies were
stuck with low or declining pricing power, direct cost savings or lower unit costs due to
higher output represented extraordinary value. To the extent that process and output
optimization could be done in real-time, including automated learning and self-correcting
adjustment, the value to customers could be staggering.

Software Rivals
In creating value for customers via both predictive maintenance and optimized output, GE felt
it had an advantage over rivals such as IBM, SAP, Oracle and the vast number of start-ups in
which over $2 billion of venture capital had been invested by 2017. Since GE made the
important components of many of the systems that would be monitored, analysed and
optimized, it was better placed than the software firms to determine how best to fit the
equipment with monitors and sensors, how to analyse and interpret the data, and how to adjust
and optimize the equipment.

Not surprisingly, GE’s software rivals did not share this view. They maintained that they had
at least five important advantages over equipment makers. First, the software companies
viewed platforms, data formatting, intelligent analytics and predictive modelling as the
essence of the engine of industrial digitization, and these were exactly their areas of special
competence. For them, industrial machines were simply “the ignition” – necessary but not
what powered industrial digitization. What powered industrial analytics were the sophisticated
capabilities that they had spent many years developing.

Second, they excelled at looking across the entire process and value chain (before peering
down into the equipment and components) and were convinced that this “top down” approach
allowed them to optimize the entire system, not just its component pieces. In contrast,
software firms argued that GE’s “bottom up” approach (starting with the equipment) was
overly focused on optimizing component pieces at the expense of the whole.

Third, they were agnostic about the machines that were the ignition for their analytic engines.
Whether a customer’s MRI machine was made by GE or Siemens, IBM didn’t care. It could
gather the necessary data, format it, store it, analyse it, and advise on how to optimize it
within the entire system. In contrast, makers’ analytics would be biased toward or limited by
what they made.

Fourth, they could leverage solid, high-level customer relationships. SAP, Oracle and IBM
provided IT solutions that were often mission-critical to customers, and thus they frequently
interacted with the CEOs and CIOs of their customers. If a customer was buying a new ERP
(Enterprise Resource Planning) system for $311 million, its CEO was almost certainly
involved, whereas he or she might not be in the case of the purchase of a jet engine for $11
million.

Fifth, they thought they were much better able to attract, retain and deploy the human capital
necessary for Industry 4.0, i.e., to win the war for talented software engineers, systems’

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 5
For the exclusive use of U. Putro, 2017.

analysts, mathematicians and statisticians who made digital transformation possible. As


demand exceeded supply in many specializations and geographies, software firms not only
understood the compensation required to attract and retain such talent, but they had more
desirable corporate cultures that were less hierarchical and offered more freedom and
creativity compare to the old “rust-belt” giants like GE.

Among the software firms pursuing Industry 4.0, IBM’s experience shed an interesting light
on GE’s approach of “making what you optimized”. In 2017, IBM, along with SAP, Oracle
and others, insisted that not making the industrial equipment helped them better optimize the
systems. However, IBM had sung a different tune back in the early 1990s when, under Lou
Gerstner’s leadership, it had transformed itself from a ‘hardware supplier to customers’ into a
‘solutions creator for customers’. Back then, IBM argued that because it made the hardware
that would be included in any IT solution, it could produce a better solution than a mere
integrator such as EDS or Accenture. EDS and Accenture argued that their “product agnostic”
position allowed them to provide superior IT solutions because they were not tied to the
hardware loyalties that limited makers such as IBM or HP.

GE was making the same fundamental case in 2017 that IBM had made back in 1993. Perhaps
most informative for GE was the dramatic success in terms of increased revenues, profits and
stock price that IBM had enjoyed with its “hardware + software” strategy back then: from
1993 to 1999, while Accenture’s stock went from $12 per share to $20, IBM’s had soared
from $13 per share to $130.

Capturing Value
But even if GE’s “hardware + analytics” strategy was correct and the estimated market
opportunity of €250 billion by 2020 was reasonable, the burning question was how exactly to
capture that opportunity. GE had set an analytics revenue target of $15 billion by 2020, but
the nature of analytics was different for products and CSAs so the old sales process and value-
capturing mechanisms would not apply in helping them hit their target.

Among the various differences, the nature of contracts was one of the most important. Both
products and CSAs were based on fixed-price contracts, whereas analytics would be priced on
the value created and/or improvement in outcomes produced; GE would only get paid if its
analytics produced value for the customer. Furthermore, that fee in analytics was not fixed but
varied with the outcome. For example, if GE’s analytics helped reduce fuel burn by X per
engine, it would receive $1 per kilometre flown; if GE improved engine efficiency by X+2, it
would get paid $1.50 per kilometre. In addition, whereas the price of products and CSAs were
generally standardised from one customer to another, it was clear that analytic agreements
would vary substantially because the key metrics used to evaluate performance would change
depending on the customer’s business model and strategy.

Another difference was the level of understanding needed of the customer’s strategy, business
and financial model. In the past, GE had needed a reasonable grasp of these things, but the
prime knowledge required was focused on product performance and specifications. In the
future, GE could only determine the key metrics to analyse and appropriate price analytics
contracts if it knew the customer’s business almost as well as the customer did.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 6
For the exclusive use of U. Putro, 2017.

Customer contacts and decision-makers were also likely to be different. The purchasing
departments that had been heavily involved in buying GE products would be less involved
with analytics. In contrast, the CIO (who was often not involved in GE product purchases)
was likely to be heavily involved with analytics sales, as were other C-suite executives,
including the CEO, depending on the nature of the offering and business areas it covered.

GE had to be more proactive about selling analytics compared to products. It could not simply
wait for customers’ RFPs because it might be 5-10 years before customers gained enough
experience with analytics to specify exactly what they required. Not only would GE have to
forego billions in potential revenue in the interim, but if customers could specify their needs
to GE, they could specify them to a host of other competitors.

From its product sales, GE knew how difficult it was to win business when RFPs went out to
multiple providers and how small the margins were on such tenders. In contrast, by
understanding the customer’s analytics needs better than any competitor (or even the
customer itself) and proactively presenting solutions, GE could have the opportunity for
uncontested or less contested opportunities, higher win rates, and better margins.

Although GE determined to focus first on analytics relative to its own products, it had
determined that longer term it would expand beyond this. There were two reasons for
expanding over the longer term. First, since few customers purchased exclusively from GE,
even when its products were clearly superior, it never enjoyed 100% market share. So if GE
limited its analytics offering to its own equipment, its potential would be limited. Second,
even when GE’s equipment was critical to the customer’s overall value chain, GE rarely
produced all the components in that chain. For example, a GE submersible pump might be
critical to a pipeline, but valves that it did not manufacture might also be key to outcomes like
flow-through volume. If GE was to get paid based on the value produced, it could not limit
itself to analysing only the component pieces it made. It could not risk optimizing its
equipment at the expense of the overall result and risk losing out on the rewards as a
consequence.

Although payment based on ‘outcomes produced’ would likely be the norm with analytics,
this model was clearly much riskier to GE than selling products or CSAs. Given the greater
risk, should GE insist on more operational control? And how much control would customers
be willing to give up? For example, if its analytics showed that a valve that GE had not
manufactured should only be opened 80% when a GE-made pump was running at 85%,
would the customer allow GE to adjust the valve accordingly? And further down the road,
would the customer allow the system to autonomously adjust the valve and pump as required
based on analytics over time?

Given these complexities, some GE executives believed that the company should license its
analytic platform and software to customers as a first move (in the same way Microsoft
licensed Microsoft Office). Although the price would be lower, it would transfer the
operational and outcome risks to the customer. Once GE passed the breakeven point on key
platforms or analytic applications, the cost of granting new licenses to existing or new
customers would be close to zero. Like Microsoft, GE could enjoy margins of 70%-90% with
this approach – and with those margins its stock would attract a significantly higher multiple.
In short, even if revenues were lower, the company’s stock price and market cap might be
much higher with a licensing approach compared to a fee-for-performance business model.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 7
For the exclusive use of U. Putro, 2017.

Actions and Progress


Although not all these issues were resolved in 2009 when GE began its digitization journey, it
was clear back then to Immelt then that GE needed to create a systematic approach to
software development (how data was collected, formatted, pooled and analysed). In 2011,
Immelt announced the creation of GE Software, based in San Ramon, California, with Bill
Ruh at its head. 9 Ruh had previously served as EVP (executive vice president) and CTO
(Chief Technology Officer) at Concept Five Technologies, as Vice President at Cisco, and
had held various positions at IBM. One of Ruh’s key tasks was to incorporate the few from
within GE and hire from without the people with the expertise to drive the digital journey
forward. From 2009-2015, he grew GE Software from zero to over 1,200 software experts.

A key accomplishment of this group was the creation of a common platform for analytics –
Predix. This cloud-based platform allowed data to be gathered, stored and analysed regardless
of industry or customer. Just as the Android operating system for smartphones offered a
common platform that enabled the creation of a variety of applications, Predix would ease the
challenge of developing application software and make it more efficient both for GE and
external developers. Predix also provided mobility, security (encryption), and the scale
required for industrial analytics. In particular, achieving the thresholds required for security
and scale were monumental and considerably higher than for the Consumer Internet. While
all consumers wanted their personal information secured, the damage that could be caused if
industrial systems were hacked was almost unthinkable. As a consequence the encryption and
security required were of the highest level. Similarly, while the Consumer Internet generated
large amounts of data, the scale needed for an industrial platform was of another magnitude
altogether. The amounts of data generated by the continuous time-series nature of industrial
machinery would dwarf the discrete transaction data coming from the Consumer Internet
made famous by companies such as Amazon.

Creating a platform that could meet these requirements (and more) came with a high price tag:
the cost of the Predix platform together with other products and capabilities that GE had
developed by the end of 2016 matched and likely exceeded its early investment commitment
of over $1 billion. 10

GE Digital

In September 2015, GE created GE Digital and Ruh was asked to head the unit and promoted
to Chief Digital Officer. GE Software was folded into the new entity. 11 On the announcement,
Immelt commented,

As GE transforms itself to become the world’s premier digital industrial company,


this will provide GE’s customers with the best industrial solutions and the
software needed to solve real world problems. It will make GE a digital show site

9 “Creation of GE Digital.” GE Newsroom, 14 Sept. 2015, www.genewsroom.com/press-releases/creation-


ge-digital-281706 , accessed 17 Mar. 2017.
10 Gertner, John, “Behind GE’s Vision for the Industrial Internet of Things,” Fast Company, June 18, 2014.
11 “Creation of GE Digital.”

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 8
For the exclusive use of U. Putro, 2017.

and grow our software and analytics enterprise from $6B in 2015 to a top 10
software company by 2020. 12

Ruh said he expected digital revenues to increase from $6 billion in 2016 to $15 billion by
2020. 13 He agreed with the $250 billion estimate proposed by BCG years earlier, with $100
billon of that spent on the software platform and $150 billion on software applications.
Concerning GE’s position in that race, he commented,

Our target is to eventually have hundreds of thousands of app developers working


on Predix…We have first-mover advantage. We’re the only ones with a platform
at the edge [of the Internet] all the way to the cloud. 14

With $6 billion in digital revenues, GE had had some early wins. These came partly from
being proactive with its customers and leveraging existing customer relations, and partly from
simplifying for customers the incredibly complex nature of analytics into a three-part
framework: Get Connected, Get Insights, Get Optimized. 15

A deal signed in October 2016 with New York state’s public electrical utility (New York
Power Authority) to provide data analytics across all of its 16 power-generating facilities was
an example of GE’s early success. 16 Although it had previously struck deals with more than
20 utility companies across the US, none had applied to all of the respective utility’s
generating facilities. For the NYPA deal, GE predicted that it would create $1 trillion worth of
value to the economy in the forthcoming decade. Sensors attached to equipment were
expected to provide predictive alerts of “possible failures weeks before they might occur”.
While details of the deal were not disclosed, the revenue to GE was believed to be outcome-
based, with performance incentives.

Challenges Ahead
While not all competitors would agree with Ruh’s conclusion that GE had a first-mover
advantage and was the only company with a digital IT stack from the edge of the Internet to
the cloud, no other company could boast $6 billion (4.9% of revenues) from Industry 4.0
services. Nonetheless, GE faced a number of challenges.

Chief among them was ensuring it had the human capital to capture the potential
opportunities. Even though GE had made progress, it was clear that the sales process and
capabilities required for analytics were different from those that had led to success in product
and CSA sales. Normal sales training was unlikely to take GE from the “low-hanging fruit”

12 Ibid.
13 Detar, James. “GE Sees Digital Revenue More Than Doubling To $15 Billion By 2020.” Investor’s
Business Daily, Investor’s Business Daily, 23 June 2016, www.investors.com/news/ge-courts-silicon-
valley-investors-for-digital-industrial-push/ , accessed 17 Mar. 2017.
14 Detar, James. “GE Sees Digital Revenue More Than Doubling To $15 Billion By 2020.”
15 “Predix.” Start Your Industrial Internet Journey | Predix.com, www.predix.com/get-started ,
accessed 17 Mar. 2017.
16 Rulison, Larry. “GE Power Wins NYPA Grid Software Contract.” Times Union,
www.timesunion.com/business/article/GE-Power-wins-NYPA-grid-software-contract-9869789.php,
accessed 17 Mar. 2017.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 9
For the exclusive use of U. Putro, 2017.

that had generated the $6 billion in initial revenue to the more challenging loftier fruit
associated with the goal of $15 billion. Being proactive and understanding the customer’s
digital needs better than customer, as well as knowing the customer’s business well enough to
propose solutions, measurements and compensation payments, required a level of strategic
thinking and business acumen that its sales people had not needed before. To go from $6
billion to $15 billion meant that GE would need to determine what percentage of the current
sales force could be retrained and how many new people it would need to hire. It was clear
that some existing sales people had neither the talent nor the motivation to make the
transition.

Another challenge was the recruitment and retention of software experts. By 2017 GE had
recruited 1,500 experts, but hiring, retaining and motivating the talent needed (given the
limited supply and growing demand) was daunting. Basing GE Digital in the heart of Silicon
Valley was essential to its success. Most executives agreed that if the digital operations had
been situated near global headquarters in Fairfield, Connecticut, the task would have been
much harder.

There was also the issue of cannibalisation of GE’s product and even CSA revenues. Digital
services would likely extend the life of GE products. For example, jet engines operating in
dustier climates such as China and the Middle East wore out faster, but analytics revealed that
by rinsing the engines more frequently than in other climates, performance (especially
regarding fuel burn) could be improved and the useful life of the engines extended.17
Similarly, some executives saw a potential for digital engagement to impact some of GE’s
“break-fix” CSAs. If analytic services reduced the failure rate or prolonged maintenance
requirements, it would lower the fees charged in a typical CSA. Because CSAs accounted for
75% of GE backlog, any reductions that were not offset by increased analytic revenues and
profit contributions would result in a net fall in earnings.

It was also clear that GE could not “go it alone” – it would need to open its Predix platform
and tools to outside developers. As an analogy, when Apple constructed its closed mobile
phone operating system (iOS) it recognized that it could never employ enough developers to
anticipate consumers’ needs – hence Apple not only opened iOS to tens of thousands of
independent developers but also created development tools that made their efforts to create
applications for iPhone customers even easier. Ruh was sure that GE would need to do the
same.

However, external collaboration went far beyond independent developers. GE had already
discovered the benefit and challenges of enterprise-level collaboration, forming a joint
venture to focus specifically on the digitization and optimization of entire airlines (not just
engines) with Accenture:

Taleris – a joint venture between GE Aviation and Accenture [is] dedicated to


providing airlines and cargo carriers around the world with Intelligent
Operations services to predict, prevent and recover from operational disruptions.
Intelligent Operations services use integrated tools such as performance data,
prognostics, recovery and planning optimization solutions to improve operational

17 Hatch, John. “Can GE Innovate Innovation with the Predix Platform?” New Equipment Digest, June 2,
2016.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 10
For the exclusive use of U. Putro, 2017.

efficiency. Taleris’ technologies and services are OEM and equipment agnostic –
able to provide capabilities across all airframes and systems. 18

The benefits of Taleris included reduction of unscheduled maintenance, fewer delays and
cancellations, increased aircraft availability, enhanced on-time performance, increased
maintenance efficiency, reduction in maintenance costs and lost revenue. 19 Taleris scored a
major contract with Etihad in 2013, although in the three years since the formation of the JV
no other major deals with airlines were signed.

As of 2017, GE had entered into a number of alliances with various partners. Indeed the list
was so long and included firms who were traditional rivals that some analysts questioned the
alliance strategy or wondered whether it was simply a “natural hedging” approach (see
Appendix 1).

Going Forward
How much GE Digital could impact the company’s overall performance was unclear. At $6
billion in 2016, GE Digital was just under 5% of total company revenues. While the ultimate
margins of the unit were expected to be much higher than product units, its margins were not
disclosed. Since the creation of GE Digital, the stock had until very recently outperformed
other industrial companies (see Exhibit 4). Was the market beginning to believe in the
digitization and the analytics strategy and capability of GE? Or did the stock price simply
reflect the improved profit prospects of GE’s core businesses, not the least of which was oil &
gas?

When Immelt ran out of runway in June 2017, one could make the case that with $6 billion in
revenues GE Digital was getting close to the point of “rotation”. 20 However, even if it had
achieved that critical take-off point, what would the new CEO, John Flannery, need to do to
ensure that successful flight could be sustained? Perhaps Ruh was correct and GE Digital still
enjoyed a first mover-advantage. If so, the good news was that it was out in front; the bad
news was that it could not look to others for what to do to stay there.

18 “General Electric - Accenture Alliance.” General Electric - Accenture Alliance. N.p., n.d. Web. 20 Mar.
2017.
19 “Etihad Airways and Taleris Implement New Technology to Predict Aircraft Maintenance Faults, Reduce
Flight Delays.” GE Newsroom. N.p., 29 Sept. 2014. Web. 20 Mar. 2017.
20 Rotation is the speed at which if the pilot pulls back on the stick, the plane has enough lift to break free of
the tarmac.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 11
For the exclusive use of U. Putro, 2017.

Appendix 1
GE Alliances 21

Accenture
The Industrial Internet of Things has been heralded primarily as a way to improve operational
efficiency. But in today’s environment, companies can also benefit greatly by seeing it as a tool for
finding growth in unexpected opportunities. In the future, successful companies will use the Industrial
Internet of Things to capture new growth through three approaches: boosting revenues by increasing
production and creating new hybrid business models, exploiting intelligent technologies to fuel
innovation, and transforming their workforce. Together, Accenture and GE Digital are changing the
game for clients by designing, developing, and implementing Industrial Internet solutions that drive
major improvements in productivity, asset uptime, and business insights.
PwC
The PwC and GE Digital strategic alliance helps clients harness the power of the Industrial Internet to
build thriving digital enterprises. The Alliance combines PwC’s strategy through execution approach
and business transformation services with GE Digital’s intelligent software and sophisticated analytics
capabilities. Together, PwC and GE Digital help companies across the oil & gas, utilities, industrial,
retail, healthcare, and transportation sectors reduce downtime, increase efficiencies, and help drive
profitable growth. PwC adds value to clients through a network of firms in 157 countries and more
than 208,000 people.
Deloitte Digital
Deloitte Digital is a creative digital consultancy that combines innovative technology capabilities,
business acumen, and industry insight to imagine, deliver, and run the Industrial Internet of Things
(IIoT). Our IoT vision focuses on top and bottom line value creation. With our deep industry
knowledge and a broad partner ecosystem, we deliver results—from ideation to pilots to deployments
at scale—with an agile approach that delivers rapid returns.

21 Source: https://www.ge.com/digital/content/meet-our-ecosystem

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 12
For the exclusive use of U. Putro, 2017.

EY
Together, EY and GE form a powerful strategic alliance with a single purpose: to build a world that
works better for companies and their customers. The Digital Industrial Revolution brings together the
cloud, big data, IoT and cyber in the Industrial Internet. Our common goal is to bring significant
positive change to our global clients as they pursue their business strategies in a digital world. EY is a
global leader in assurance, tax, transaction and advisory services. The insights and quality services we
deliver help build trust and confidence in the capital markets and in economies the world over. We
develop outstanding leaders who team to deliver on our promises to all of our stakeholders. In so
doing, we play a critical role in building a better working world for our people, for our clients and for
our communities.
Tata Consultancy Services
TCS and GE are creating innovative digital ecosystems for various industries by leveraging TCS’
comprehensive industry domain expertise and digital assets to enrich Predix. As a Predix early
adopter, TCS has released several digital solutions on the platform, such as Satellite Image Analytics,
Supply Chain Monitoring, Prognostic Maintenance, and Engine Telematics. Moreover, TCS has the
largest pool of Predix trained experts and plans to ramp up this investment to 1,000 developers in
2016. TCS is significantly expanding its global team to help businesses seize opportunities for
unprecedented growth in the Industrial Internet.
Microsoft
GE’s Predix platform on the Microsoft Azure cloud will allow customers to take advantage of the
combined industrial and enterprise software expertise of Microsoft and GE – ensuring greater choice
and flexibility when it comes to harnessing the power of data from machines and systems of
intelligence. Customers worldwide will be able to bridge the divide between the operational and
information technologies that make up the Industrial Internet of Things.
Oracle
Oracle offers industry-leading cloud-based solutions that provide global enterprises complete
flexibility and unmatched customer benefits. Together with GE Digital, we offer the world’s most
innovative, outcomes-driven, products by combining the traditionally disconnected silos of Operations
Technologies (OT) with Information Technologies (IT) delivering cutting edge, wing-to-wing
solutions, faster ROI, and a simplified path to becoming a fully digitally connected business.
Customers gain an unprecedented ability to automate and optimize business processes, enable better
asset visibility, more efficient supply chains, intelligent manufacturing, predictive maintenance and
service, and improve logistics and transportation efficiency with maximum security.
Intel Corporation
Intel pushes the boundaries of smart and connected technology to make amazing experiences possible.
Through Intel’s partnership with GE, a comprehensive portfolio of products from Intel and its
ecosystem is helping spread intelligence throughout our cities, factories, healthcare facilities, and
machinery. Optimized GE software for Intel® products enables faster, more scalable IoT deployments
that deliver higher ROI and comprehensive security from chip to cloud.
Huawei
Huawei and GE Digital are partnering to help customers drive digital industrial transformation with
IIoT solutions.
Cisco
Cisco and our partners work with innovative clients around the world to help them digitize. We are
building and implementing digital roadmaps and transformations in industries ranging from
manufacturing to retail to government. We are helping manufacturers improve their operations, banks
to reimagine the customer experience and energy providers to become more efficient. And we are
doing it today.

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 13
For the exclusive use of U. Putro, 2017.

Exhibit 1
GE Revenue and Profit under Immelt

Source: Company annual reports


Exhibit 2
GE’s Stock Price Plunge vs. Other Industrials

Source: www.ycharts.com, accessed March 17, 2017

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 14
For the exclusive use of U. Putro, 2017.

Exhibit 3
GE’s Stock Performance vs. S&P 500

Source: www.ycharts.com , accessed March 17, 2017

Exhibit 4
GE Stock Performance vs. Peer Industrials

Source: www.ycharts.com , accessed March 17, 2017

Copyright © INSEAD This document is authorized for use only by Utomo Sarjono Putro in 2017. 15