In my last post I commented on the disruptive technology elements, some possible new entrants in to the market, and I suggested that a different approach could/would be taken to assembling and deploying. I ended with the following questions; What would the delivery elements, both products and services, be if my thoughts on this disruptive approach are right? I also included the diagram below to get some ideas and thought back as comments started from the previous post and I promised more detail in my next post so here it is.
If we start on the left then we know that at a personal level of use we increasingly are moving to wanting to have it our way. Think Apple iPhone for the selection of those services we want to use, and cross it with a corporate thin client for the core enterprise services as refaced versions of existing applications. The first change is that we will have to get out of calling people ‘users’, they will literally be consumers selecting the services that are the best fit to enable them to perform their role. So there will be competition and dynamics in what will be consumed, and from where, rather than say an imposed Enterprise Business Intelligence package. However that does mean that the existing enterprise applications rendered as ‘services’ die away, definitely not in my view, that layer of enterprise transaction recording, procedures, shared and common data etc, has to remain in place. Once again, as before with each new generation of technology, it is just that a new layer with different capabilities getting added. In this case the different capabilities focus on maximising external business options around events and people activities.
Just a footnote to the above as it might seem strange not to mention the acquisition of Sun by Oracle, but in practice i don’t think i have anything to add as it just looks like a pretty good example of my point in this and the previous blog!
The diagram shows this as a narrow point because literally every consumer (user) will be different in their selection and use of services to augment the enterprise defined services; and the differentiation in abilities is what creates the ‘value’. However it won’t all be on a PC, for a start the number of smart phones in the world is moving rapidly to equate to the number of PCs, throw in Web Books, Media Devices and home systems and the question for the next layer is not only how to deliver to these different forms of devices, but how the enterprise, or any other supplier of services for that matter, can provide the controls on what can be delivered to what device, and what can be actively done from each device. That conveniently leads to the third layer on connectivity and policy management which I think is the real challenge and where a whole new market for the ‘system integrators cum outsourcers’ lies, but it is easier if I leave this layer and consider the last two on the right.
The range of services available to be consumed is pretty self explanatory, but where they come from is more the issue. Some may be from secure ‘internal’ sources running on so called ‘private’ Clouds, others may be from available as services that you consume for payment, the equivalent of say Salesforce.com today, except I suspect that all major software vendors will be using the same model to delver a least some part of their newer capabilities for the new business consumers/users market. Other services might be free, or at least supported by advertising, Google Maps as an example. Tied to that is the last layer of resources to support the services, and at this end of the diagram standardisation is all, with huge volumes at large margins. There are a number of obvious points about what is running on which resources, and if they are internal and secure, or external and shared, or even external and public, but the decision criteria are reasonably clear, and related to the type of service.
That clarity is not present in the middle layer, exactly how the two ends; one focussed on massive numbers of individuals plus workgroup, departmental and enterprise differentiation as value creators; the other around totally standardised but huge in capacity resources based on cost as the objective, plus the spread of public, and private, provisioned services, plus some enterprise applications are connected is the big new ‘integration’ issue. Its not just the ability to provide a connection, there may be rules as to what services can be run on what devices, as an example, then there is a whole question of who did what with whom, or using what service, for governance and compliance, and thats before we get to who pays whom for what questions.
So the theory that open standards and APIs reduce integration works as long as we are talking about a single dimensional IT model based on simple technology integration, but the integration market for Services from Clouds is only just beginning to become clear. I can’t see enterprises wanting to diversify from the challenges in their core businesses as a combination of global competition and a downturn in the market continues to occupy their prime management time, so this looks to be a new market for the SIs, who I would now christen as Services Integrators, rather than their old role as System Integrators.
A word of caution on this post, please remember its speculation, hopefully informed and thoughtful speculation but all the same it is speculation. This is my personal attempt to draw together the pieces and try to share some conclusions with others who are attempting to do the same, so thoughts, ideas and your own views on the topic are welcomed!