GE, The Industrial Internet and the Battle to Come

During Cambridge Service Week this year we heard from Stefan Bungart, Leader of GE Software Europe. Stefan talked about GE's development of a new services platform - Predix. Think of the iStore, but for the industrial internet. GE's position is that it wants Predix to become an openly available platform that can host apps developed by others - apps that are used to remotely monitor and manage machines and equipment, indeed any device connected to the internet - hence the industrial internet.

You could argue that Apple, Facebook and Google have largely sewn up the business to consumer internet - they are the dominant platforms. Other platforms may emerge, but they face an uphill battle to overcome the incumbent players. The industrial internet, however, is still wide open. We don't yet have any dominant players and they may never emerge. However, manufacturing firms across all sectors recognise the way the world is moving. More and more devices are being connected to the internet. These devices are feeding data back to central control hubs and the best of these are using the data to make predictions about product performance and how this can be optimized, as well as using the data to inform future generations of product design. The question - hence the battle to come - is which firms will dominate the industrial internet.

GE has already declared its intent - in 2015 the Predix platform will be made publicly available. Jeff Immelt, Chairman and CEO of GE is quoted as saying "the more we can connect, monitor, and manage the world’s machines, the more insight and visibility we can give our customers to reduce unplanned downtime and increase predictability. By opening up Predix to the world, companies of any size and in any industry can benefit from the investments GE has made by eliminating the barrier to entry". What he doesn't say is what happens to the data that all of these devices generate as it passes through the GE platform. GE is reported to use 10 million sensors to monitor daily 50 million data points across $1 trillion of managed assets. The businesses order backlog is around $180 billion - a number that continues to grow as the installed base of GE assets and those that GE helps monitor increases in size. Will GE be the equivalent of Apple, Facebook and Google for the industrial internet or will someone else seize this market? Potential competitors from the software, applications and consulting industry might include IBM, Microsoft or Tata. From an industrial perspective the smart money might be on Hitachi, Samsung or Siemens. How about Apple, Google or Facebook? Can and will they make the transition to the industrial internet?

The jury's out on how this opportunity will develop, but one thing is clear. The battle for the industrial internet will heat up in the next few years. The potential for innovation and greater efficiency in product and service design, as well as operation and maintenance is too great. The winners of this race will have access to unparalleled data that if used insightfully will drive significant service efficiency and innovation. Firms will still have to deliver great service - they'll have to get the basics right - but they'll do so from a rich and data-informed position that will put them ahead of the rest of the pack and so confer significant competitive advantage.

Trading off service experience and service efficiency

I have just been at the QUIS2013 conference in Karlstad, Sweden. The final session was a panel discussion with the title "Small Details: What They Are and Why They Matter". The panels, which included a mix of academics and managers, started by providing a few examples of small details and why they matter is services. Janet McColl Kennedy, a Professor from the University of Queensland, told a story about an elderly person in  a care home, who when asked "what could we do to make your life better", replied "let me have a real cup of coffee in the morning". The chair of the panel (Professor Ruth Bolton from Arizona State University) countered, by saying "I'm Canadian, so for me it would have to be tea". This short exchange illustrated the importance of small details. To get the service right you have to deliver services which are personalised, contextualised and time dependent. Manfred Dasselaar, a service manager at Ericsson, built of this theme, describing the challenge of getting customers to understand how hard Finnish engineers were working on solving their queries when the customers and engineers were not co-located. He gave the example of a conference call involving Finnish engineers and their external customers. The customers were getting frustrated that they were not getting much from the engineers (they weren't communicating much, but then they were Finnish). Because the customers were not in the room they could not see that although they were not talking much, the Finnish engineers were sharing images and data on their computer screens - screens that the customers could not see. Once the customers understood how the engineers were communicating and how hard they were working on solving the problems, they became much happier. Again an illustration of how small details can influence the service experience.

So who delivers these small details and how do they make sure they are time, person and context dependent? Often the staff at the front line - often the least trained and least well paid people in the organisation, but those closest to the customers. Should we give these staff more autonomy? Should we train them and seek to create an organisational culture which allows service providers to personalise the service experience? At first blush the obvious answer is yes - this might provide a new and sustainable way of competing. But there are three issues with this first blush response. First, by personalising the service we can increase the cost and complexity of the service - we lose the efficiency gains that can be driven by standardisation and commoditisation. Second, by personalising the service we can create inconsistencies in customer experiences. If every time you are served by a different server and you get a slightly different personalised version of the service then how frustrated do you become when one server fails to do that special thing for you that the previous server did. Third, the more we use technology in services, the more we end up standardising the service. This drives efficiency, but does it deliver the best customer experience? Are automated voice systems better than talking to real people?

It seems to me there's a careful service design and delivery tradeoff to be understood here. Clearly personalising services and tailoring them to individuals in time and context can enhance the service experience, but at what cost in terms of efficiency and consistency? In designing and delivery services we need to be clear about the boundaries - where the scope for personalisation lies and where we should standardise and control. Going too far in either direction is going to result in disaster.