As part of my ongoing odyssey within the blogosphere I spent some time catching up on Nick Carr’s blog – someone whose ideas always fascinate me. The most active article over the last few weeks appeared to be one entitled "The iPad Luddites" about the wide range of emotions that this seemingly simple device has evoked. I wanted to comment on this from three perspectives really: firstly about the notion of computing devices like the iPad in general, secondly about the issues of ‘generativity’ that have been sparked by the release of such a device and then lastly about what the emotions around the iPad might tell us about cloud platforms (or service delivery platforms) in general.
The iPad and The (New) Third Way
At a macro level the amount of discomfort felt by technologists seems to depend in direct proportion on whether you view the iPad as a new kind of window into the web (i.e. primarily a highly specialised, beautifully packaged and superbly usable content delivery mechanism for the mass market) or as a replacement for general purpose computing devices. To clear that point initially – as the iPad is not the point of this post per se – I am in the former camp and therefore see such devices as a ‘democratisation’ of computing in much the same way as other mobile computing devices. I think it is interesting and good that people have more choice in how they access content and applications and I feel that such devices – in a general sense – create opportunities for both new consumers (so people who would previously not have been able to confidently access computing services or the web) and producers (so people who can now create content and applications that deliver new services to this newly empowered group). In this sense such devices can lower the barrier of entry and therefore reach wider groups and this in turn can enable long tail business models to flourish by enabling broad access to hitherto inaccessible niches. Even for technology literate people like me I can see a perfectly reasonable desire to not be ‘on duty’ 100% of the time and to feel the relief of falling back into a consumer role for a while. I don’t see tablets as a replacement for general purpose computing devices, therefore, but rather as an alternative way of accessing services and content. As a concrete – and personal – illustration of this I have an 85 year old grandfather who – about 10 years ago – decided that he should become ‘computer literate’. He now has three PCs in his house – one connected to an amateur radio set, another that he uses for ‘tinkering and learning’ (so mild "generative" activities) and one that he ring-fences from his mildly ham-fisted experimentation (he once cut a screen off a laptop and connected the base to a monitor in order to use it purely as a desktop). This third PC is left alone so that he always has simple and secure access to the web and his email. I can see that this third PC could easily be replaced by a device like the iPad and that this would be both a more pleasurable experience for him and potentially open up more opportunities on the web by simplifying the experience and making it more easily consumed. In no way would this reduce his desire to dismember and otherwise experiment on more general purpose computing devices, however. So far so what.
The ‘what’ at this point is the fact that Apple has created not just a new device – which in and of itself is pretty standards compliant from an interoperability perspective – but that they have also created a platform to service it. The nature of this platform – and its tight packaging with Apple products – leads to “generativity” concerns due to a perceived lack of ‘openness’.
The iPad and "Generativity"
One of the key worries that people have about the iPad relates to its closed nature – both from a hardware and software perspective – and Apple’s desire to control access to the consumer base in order to ‘manage’ the experience. There is a perception that controlled platforms like this can have a negative impact on both "generativity" – through heavy-handed governance – and consumer freedom.
In general I support open hardware, platforms and business models and each of these perspectives has different ‘contexts’ for "generativity" that need to be separated if we are to get to the bottom of the issue.
Whilst many people mourn a past where machines could be opened, modified and enhanced in order to tinker and see what happens (i.e. one form of "generativity") I believe that the increasing internal complexity of devices coupled with their decreasing external complexity – i.e. ensuring simple operation for the majority – is an inevitable marker of the increasing maturity of a product category for three synergistic reasons:
- As devices increase in complexity so it becomes both undesirable – and indeed practically impossible – to allow even competent people to open up a device and work on the internals without compromising it;
- This same cohesion is also a key enabler for mass market adoption given that the majority of people just want devices that work; as a result those companies who can attractively package technology for consumption are rewarded financially by the broadest market. This is also understandable from the perspective that once things work well enough and are mature enough messing around with them is more trouble than it’s worth; and
- As a device becomes increasingly consumerised in pursuit of this mass market so the need to produce them at scale also kicks in; this in turn promotes tight integration of components for manufacturing optimisation and also reduces the cost of the device – as a result opening up a device and replacing individual components becomes increasingly difficult and decreasingly cost effective at the same time (rapidly to a point at which it starts to become cheaper to replace the whole device from a consumer electronics perspective).
Whilst it is therefore sad that a tradition of hobbyist hardware tinkering has become obsolete this aspect of ‘generativity’ is probably now the least valuable in any case, as hardware has long since become a way of delivering software rather than an end in itself (i.e. most hardware works sufficiently well and is so complex that the value of “generative” activities has diminished almost to the point of zero; at the same time what can be done with software has increased hugely and hence become the new sweet spot for “generative” activity). There are emerging ‘open hardware’ models that reflect the work that has been done in the software community but this is still a pretty niche activity; for those who really want to experiment with hardware design the complexity of devices now means that most of their work must be done in software anyway.
What is still a critical aspect of ‘openness’ from a device perspective, however – much like in software – is interoperability (so use of standard ports and connectors to enable the integration of your various gadgets). Given that most devices these days are bluetooth, wifi, 3g and usb enabled, however, interoperability to allow different devices to play nice together is now generally a hygiene factor.
“Generativity” and Hardware
Looking more broadly at the question of whether closed hardware decreases “generativity” within society overall we can consider the question from two perspectives. From a strictly IT-centric view we may feel that the increasing penetration of consumer devices means less people who are ‘pc-literate’ and thus able to ‘create’ IT-centric systems (i.e. less people tinkering with hardware or general purpose programming languages to create value through IT). This is to miss the point in my view. In reality the majority of people who will pick up devices like the iPad would never have been “generative” through traditional IT in any form at all – if we give them access to new tools and software that are easily consumable, however, they will be able to use these tools to become vastly more “generative” within their own spheres of expertise. As a result we would have to say that putting better packaged, more accessible and user friendly devices into the hands of the largest number of people increases the overall ability of society to be “generative” in the broadest sense.
Given the maturity of the IT industry as a whole, platforms – from mainframes to windows to cloud – are still a major area of contention and competition despite their essentially infrastructural nature. In much the same way as hardware went through a golden age of tinkering and “generative” activities so the last decade has been the golden age of platforms with everyone wanting to develop a different way of implementing software and controlling hardware; as a result I guess the next level of discomfort people feel with the iPad is the closed nature of the ecosystem in which it exists.
In considering the “openness” of platforms, however, I would look at two key points: how innovation often leads to proprietary platforms in immature markets and whether we should even care about proprietary implementation platforms in a web age.
Innovation and proprietary platforms
From my perspective the rise of proprietary platforms is all part of the natural cycle of any product category that takes a sudden leap into consumerisation and mass market adoption. For a long time people tinker with various apparently disconnected concepts but then at some point someone decides to take a whole bunch of these innovations and collapse them all together into highly specific offerings that have only the 20% of functionality that is really required to enable 80% of the value (i.e. they create a new and much simplified way of doing things within the context of the problem domain – in this case touch-enabled web applications). In order to do this they often have to create proprietary data formats, protocols and APIs – or some subset of these to fill gaps in existing standards – and then orchestrate them all together in a consumer centric way; such combinations become a new platform. At that point they have recognised the needs of their consumers, delivered something ‘new’ that enables them to do things they haven’t been able to do in the past and thereby created a new market via a proprietary platform. All sorts of other people then decide that they can develop a better platform and join in. Apple (and others) are still at this stage since their platform is the proprietary integration of the formats, APIs and standards required to build and deliver applications through their ecosystem. Successful first movers in these spaces often appear ‘magical’ as for the consumer the experience of controlling all that power through simple tools is nothing short of revelationary.
Looking at this process, however, there is nothing specific that impacts “generativity” in the broadest sense; whilst the reach of things that result from “generative” activities might be less because applications and services are tied to a specific platform they do not stifle “generativity” per se.
What is openness in a web age?
Taking this further, in the new business ecosystem do we really care that there are proprietary platforms? That may sound like a strange question but if we step away from technology and concentrate on outcomes for a moment then we can think about the key points at which openness is really important rather than just use it as a mantra.
At the end of the day we’re all just trying to get something done and IT is mostly a pain in the ass that gets in the way. With increasingly sophisticated ways of describing the outcomes we need, however, the key elements of any platform become how effectively they can support us in realising our intent rather than on how they work internally. Two often used measures of ‘openness’ are interoperability and portability; whilst they are not mutually exclusive, preference for interoperability tends to favour implementation diversity and specialisation as a goal whereas preference for portability tends to favour lowest common denominator single implementations and general purpose use cases.
One of the implications of a wholesale shift to web based platforms is that technology is no longer a) monolithic in its delivery and b) an all or nothing costly purchase decision. Web architectures encourage the creation of lightweight components and integration, whilst cloud platforms allow easy access and exit. Given that different kinds of components will be best optimised on different kinds of platforms and with different processes and tooling we can also start to think about the best way to realise them individually; cloud platforms basically allow us to choose the best implementation vehicle whilst still allowing us to integrate the parts back together. If a proprietary platform – because of the degree to which it is specialised and optimised to realise a specific task as effectively as possible – allows us to deliver value far more quickly and cost effectively than an ‘open’ platform (i.e. one which is optimised for portability and hence more general purpose) then which would we pragmatically choose (especially given that the entry and exit costs for delivering on a platform are much lower than in previous models of IT and so we could realistically redevelop that component rapidly somewhere else in future if necessary)? In any other industry we see people building around the components of partners whose design and manufacture is specialised (and hence their internals can be considered proprietary) and all we care about is a) how well those components work within our broader use case and b) whether they fit together with the other components we have. Coming at ‘openness’ from this perspective we see that from a consumers viewpoint (whether that is an end consumer or a partner in a value chain) all that truly matters are that the component in question performs as required, is able to be integrated and has robust processes in use in its creation and maintenance; as a result ‘interoperability’ and measures around process quality become far more important than the technology being used to implement it (and hence portability).
To bring this back to the discussion of the iPad and whether the lack of portability in the underlying platform impacts “generativity” or “open access”, however, we must step back and consider the platform in the light of the macro level definition of “openness” explored above:
- Apple enables anyone to develop an application to execute on their platform subject to certain quality criteria; as a result application providers have access to a huge market of consumers via Apple’s marketplace;
- Application providers can integrate with external services not provided by Apple or any of its other partners in order to enrich the experience delivered by the application on the end device;
- Consumers have access to many different application providers – as well as web access via a browser – and could hence be argued to have complete freedom of choice.
- Consumers can configure many of the applications to integrate data from other web services that they use and hence are not locked into having all of their data held by Apple or its chosen providers.
As a result of these observations it is fair to say that Apple’s platform is fairly open at a macro level – anyone can consume services, provided by anyone using data and services that are part of other ecosystems.
“Generativity” and Platform
So looking at the question purely from a platform perspective we would have to say that the Apple platform does not impact “generativity”, only the ability of the application implementer to define the platform they wish to use according to their preference (so to perform tinkering and “generative” activities in the platform space). For the vast majority of people who just want to deliver services quickly and cheaply to make money or for those who just wish to have a pleasant experience in consuming such services the platform is only of interest in-so-much as it helps or hinders them (as long as they can connect their various services together – so interoperability remains a key requirement). If we relate platforms, therefore – as a necessary but essentially hygiene related infrastructural capability – to hardware then we could consider that the discomfort people feel with proprietary platforms is not that they limit “generativity” overall (so both Microsoft and Apple platforms have unleashed a huge amount of higher level “generative” activity by removing complexity at the platform level) but that they limit “generativity” within a technical space that technology-literate people have been used to controlling and tinkering within. In this context one could propose that decreasing choices in how we implement software that actually does useful stuff is merely the consumerisation of software platforms (in this case the consumer being software developers who have good ideas and just want to get stuff done or people who want to easily consume their good work) and hence the next logical step in the commoditisation of technology markets.
Does this mean that we are doomed to a future of competing proprietary platforms? It is one possible outcome (i.e. the dominance of mega-platform providers) but not the only one. Given that platforms are infrastructural in nature – and hence tend towards commoditisation – it is feasible that there will be open source alternatives which can compete against the big companies. Furthermore it is likely that all platforms will tend towards broad functional equivalence over time – as people adopt useful ideas from competitors – and thus the question will be whether open source or proprietary models deliver the ‘killer’ offering that scales into a de-facto standard. As always the market will decide; the key message, however, is that how we implement the internals of things hardly matters anymore – beyond a few interoperability caveats; what matters is how well we can realise higher level outcomes for people (both economic and social).
The final level at which we need to examine the impact of the iPad is at the level of business models. Whilst we have examined the implications for hardware and software platforms as individual elements, we also have to consider the broader question of how these components are integrated into a business proposition. This is probably the area where most concern justifiable exists, as whilst technologies in the hardware and platform space are neither inherently good or bad, Apple’s current business model is a top-to-bottom walled garden of hardware, platform and delivery channel within a single brand experience. This model delivers a tight integration between all elements of the offer that prevents other companies competing against Apple for specific components within the overall ecosystem of business types they have assembled. To know whether this is a problem – and for whom – we need to consider the different kinds of businesses at play and the relationship between them.
The Apple Business Stovepipe
Currently the whole Apple ecosystem is a tightly coupled, single business model stovepipe that you take or leave as a package.
In this context all of Apple’s components live and die as a package and cannot be independently optimised. On the reverse side consumers cannot choose to only take those elements of the Apple product set that they really want and have to take all elements together (they cannot choose to have an iPad but source applications from elsewhere, for instance, or choose to make use of Apple’s platform on another manufacturer’s device).
These properties of the Apple business model are far more troubling than the individual technologies themselves – which as we have seen are as open as necessary and clearly supportive of “generativity” – because they limit competition within the layers of the ecosystem and prevent “generativity” in the business model space. The iron control exerted by Apple on the people who want to live within their top-to-bottom closed ecosystem means that there are no opportunities for “generative tinkering” from a business perspective (different devices, different stores, different payment models, different brand experiences etc. etc.).
This stifling of competition has some serious consequences for both the consumer – as it limits competition, choice and innovation, allowing Apple to drip feed innovation at their own pace to maximise their revenue whilst simultaneously keeping prices artificially high – and – perhaps more surprisingly – for Apple themselves.
Apple’s Business Components
In my previous post about divergent forms of differentiation I touched on the fact that there are four broad kinds of businesses – culturally and economically – that need to be optimised in different ways. To recap these were:
- Relationship businesses: These businesses are essentially business capabilities that leverage trust relationships to bring together different parties for mutual gain;
- Innovation and commercialisation businesses: These businesses are essentially small, innovation focused capabilities who specialise in IP generation and product and service commercialisation;
- Infrastructure businesses: These businesses are essentially business capabilities that respond to economies of scale; and
- Portfolio businesses: These businesses own and manage brands and invest their capital across a range of other business types to maximise asset growth.
If we look at Apple then we could propose a neat division of the elements of their ecosystem into this categorisation:
- Relationship business: App store. In this business Apple wants to maintain the relationship to the end customer and ‘mediate’ access to other providers to leverage their brand loyalty and the resulting trust people place in their offers and recommendations. This is one of the reasons Apple are so keen to ensure the quality of applications distributed through this channel;
- Innovation and commercialisation business: e.g. iPad. Apple does a great job of bringing innovative and beautifully designed products to market and both the iPhone and the iPad are great examples of this. In this context Apple hold the IP for these products but they do not carry out any of the actual manufacturing etc. themselves. As a result in the device space Apple are concentrating on IP generation;
- Infrastructure business: iPhone OS. In this context Apple are absolutely a platform business and such platforms are infrastructural in nature. The value of a platform is in the breadth and depth of its ecosystem and hence the economies of scale it can generate;
- Portfolio business: The Apple brand. Apple itself is clearly a strong global brand which is loved by its many fans and equally obsessed over by its detractors. This element of the business needs to have a balanced strategy and set of resulting investments across the other business types to deliver growth and ensure the long term health of Apple.
Optimise Apple, Unleash Business “Generativity”, Treat the Consumer Fairly
If we take this view of the components that make up Apple’s ecosystem then we can see that the major elements fall into different business types and hence need to be optimised differently (as per my longer post on this subject generally). As a result rather than sub-optimising all of the capabilities it holds to maintain end to end control Apple could alternatively hold these components as part of a brand portfolio that gets optimised in different ways.
In this instance the optimisation would be (starting from the bottom):
- Infrastructure business (iPhone OS): Allow other manufacturers to license and use the iPhone OS. Platform businesses are all about economies of scale and restricting penetration of your platform is a long term bad bet. Opening up the platform to the broadest market would allow Apple to monetise their IP, create more choice for the consumer and facilitate far greater “generativity” in the business space. Apple would benefit from this “generativity” at the macro level as the larger its ecosystem the greater the penetration of the platform and hence the greater the revenue for Apple. There is a direct parallel here to what happened between Macs and PCs in the last major platform wars; Apple tried to retain soup-to-nuts control whilst Microsoft concentrated on getting their platform into as many computers as possible. Only one of these companies is the largest software vendor on the planet.
- Innovation and Commercialisation business (e.g. iPad): There are three separate ways that Apple could monetise the iPad (or other hardware devices); firstly it could take a decision to retain sole control of the iPad as its own brand device for accessing the broader market that it has enabled through its platform. This would be a sensible position for them given the quality and desirability of their devices, enabling them to be a top-tier producer of devices for the iPhone OS platform that they have created. Secondly it could take the IP that it develops and license it to other manufacturers – it is possible that it could make more money from licensing IP to a broader range of device producers than from selling hardware in isolation. The third way is a combination of both of these approaches, licensing technical IP but also continuing to use this IP themselves to manufacture design-led hardware as they do today. All of these options would help to enable broader business “generativity” by creating competition in the device market around the core iPhone OS platform; from Apple’s perspective, the additional devices resulting from such “generativity” would legitimise the platform as a de facto standard, encourage people to buy their high-end devices (as they won’t feel ‘trapped’) and enable Apple to gain revenue from its other business lines (e.g. platform and apps sales) from a broader base of users;
- Relationship business (App Store): Although Apple would now have to allow other people to deliver applications to devices – as a result of losing absolute control over the platform – an own brand App Store proposition could still hold massive appeal due to the Apple brand and the trust it has amongst consumers (especially those who value simplicity and function). Apple could continue to position itself as a ‘high end’ App Store that has strong quality constraints for offered applications and which ‘guarantees’ successful execution (as today). This opening up of the applications ecosystem to other sellers would enable business “generativity” within the relationship space as consumers could choose the level of cost, support and quality they were happy to live with. Apple could benefit however the market evolved, though, by firstly having a differentiating proposition from a relationship (i.e. app store) perspective based around both their brand and their ease of use (including guarantees). From a broader perspective more competition in the app store space would encourage greater adoption of the iPhone OS platform and the devices required to access it.
- Portfolio business (Apple): There would be a number of implications overall for Apple if they decided to follow an ‘open’ business model, split their capabilities along economic optimisation lines and allow competition into their ecosystem. The primary implication would be that Apple would now be able to optimise all of their business types independently creating more value overall for the Apple brand. An obvious example of such an optimisation would be the licensing of their platform to other companies. More subtly, however , there would be nothing to stop Apple simultaneously pursuing a strategy of openness and individual business optimisation whilst also delivering an integrated end-to-end experience for their consumers as they do today. Each of their ‘open’ business components could still be wrapped in the next layer of business in order to deliver the same simple and integrated customer experience with the same level of guarantees that they do now; this in itself could be the overall brand proposition even as the individual elements are available to other people to maximise Apple revenue, business “generativity” (and hence secondary revenue) and consumer choice (and hence trust). As a result the irony is that an open strategy would benefit Apple as much as the other participants.
“Generativity” and Business
Overall, then, we would have to say that the current Apple business model prevents “generativity” in the business spaces around which they have chosen to build their ecosystem (i.e. hardware, platform and service distribution). This position is probably equally bad for the consumer, for potential Apple partners and – most strangely – for Apple itself. Google’s foray into the mobile market is far more dynamic at the moment primarily due to the openness of its business model and the resulting number of companies involved; it does, however, suffer from the kinds of fragmentation and lack of trust that Apple has managed to avoid through its tight control over its ecosystem. Like with platforms, however, it is likely that such a strategy – whilst probably unavoidable initially – will only work for a short period of time – essentially until the market matures sufficiently for people to understand its basic operation; at that point commoditisation of function – coupled with providers who will take on the role of integrators and trusted brands for consumers in the way that Apple has in its closed system – will require companies to optimise the component parts of the value chain as described above in order to remain competitive.
The iPad and Lessons for Cloud Computing
Given that this post has already extended way beyond my original intention I will keep this last set of observations to a set of bullet points:
- Cloud platforms are increasingly driving people’s understanding that infrastructure is pointless and (mostly) worthless. In much the same way as devices are becoming commoditised and hence packaged so traditional infrastructure is disappearing from our purview as it becomes commoditised. Strangely, however, this doesn’t stop many IT departments and architects continuing to think that infrastructure design and support is a primary and important activity. These activities represent continued “generative” activities in the hardware space, however, and thus provide minimal value for maximal effort given that software is now the primary medium for value adding, “generative” activities. Even infrastructure-as-a-service offerings are only postponing the point at which infrastructure disappears completely behind software platforms;
- Cloud platforms support “generativity” in much the same way as the Apple platform. They enable much greater and more rapid “generative” activities in delivering applications and services to fulfil higher level functions but you have to accept that you can no longer perform “generative” activities in the platform space itself – these are now inherently the domain of platform providers. Experience from the angst of the iPad launch suggests that this will not be an easy transition for the hundreds of thousands of geeks whose primary function is to re-create platforms and frameworks over and over again. In terms of openness, interoperability is now far more important than portability as systems will be constructed from services that potentially span many platforms. The risks inherent in these services being implemented on proprietary platforms diminishes both as a result of the broader portfolio of platforms in play but also due to the speed at which services can be redeveloped elsewhere if necessary. Interoperability also enables “generativity” for platform providers, enabling them to innovate far more rapidly and find ways of helping us to realise our own services much more cheaply and quickly than ever.
- Cloud platforms will operate within the same basic business models as discussed above in the context of Apple; as a result organisations that wish to operate within the IT market of the future will need to decide what ‘type’ of business they are and optimise accordingly. IT companies may hold a portfolio of individually optimised relationship, platform and innovation businesses but the idea of an ‘integrated’ (i.e. stovepiped) IT service provider is a dangerous fallacy. Looking specifically at internal IT departments they will increasingly need to play the role of ‘relationship manager’ and integrate many external services together to realise business aims; in this context they should be looking for exit strategies for the majority of the IT systems and platforms they operate today.