“Make or Buy” Reloaded – A Business View on Cloud Computing
I remember 14 years ago – my early IT days, I was at IBM. We built CRM solutions based on the newly released Siebel 7 Web platform. These were solutions suddenly available anytime, anywhere, and easily on demand for new businesses.
10 years later the same buzz was called ‘Software as a Service’ or ‘private Cloud’. I was glad it finally got a name. In the meantime there has been a lot of progress on virtualization, automation & the ‘Internet anywhere’. The cloud will soon reach it’s plateau of productivity. Also – after first ‘cloudiness’ – we have now a precise and common understanding what the cloud actually is (with all it’s various flavors: Private, Hybrid, Public, SaaS, IaaS and so on). We start to get an understanding on how to manage it and what the cloud means to us. No doubt, the cloud will change the way IT will deliver its services: Increasingly, we will source these services out of the cloud, just think about the simple example ‚Google Mail for corporations‘. But I will not bore you with that, there has been written a lot about this topic – after all, certain things have become a no-brainer, isn’t it? Looking at the cloud scene, I think one issue is that the cloud discussion is very much driven by an ‘inside-out’ view (the view of the cloud providers and IT-departments). But what does the cloud actually mean from a pure company perspective? What does it matter for them? Putting the security & compliance arguments aside for the moment, my hypothesis is that for the business, cost & time to market is at the end the driver to go for the cloud. Secondly, I think companies should be careful with the core processes, which differentiate them in the market – in this area standardization can completely kill their business.
So I hope I can open some new perspectives in my blog and hope you spend your valuable time reading them.
Cloud: the Inside-out view
Recently I was lucky to be part of two solution implementations, both in the ITSM tool area and with very similar scope, but one of them the traditional way and the other one with a SaaS/PaaS Cloud solution. Both I took care end-to-end, from evaluation, to design, until making sure that it is run as a Service. The perfect case for comparison, here is what I experienced:
- The first difference was that in the cloud project, the whole operation part was gone. The solution was developed & run completely on a remote web platform, delivering out-of-the box 95% of the functionality we needed. No thinking needed on how to host the platform. Also, traditional Service Design warranty aspects like capacity & availability have become obsolete, they were delivered by the cloud solution already on a very high level (with actually no ability to influence it).
- What was the same? Both, cloud and traditional solution needed a proper governance & project around it, including requirements engineering, build and test.
- However, in the cloud project, we were very much driven by the standard, that was delivered by the solution out-of the box. This of course had an influence on how we defined requirements
So out of practice I can only agree what has been written about the cloud. From an IT department view many things have become really easy and from a motivational point of view I had the feeling to focus myself on the more ’interesting’ stuff. The example I showed above is rather typical for PaaS (Plafform as a Service) environment, offering a high degree of flexibility for the solution that is actually built on it. Today, a lot of the business is around SaaS (Software as a Service), the solutions delivered to the business are ‘take it or leave it’ – this is a different flavor of cloud: standardized, industrialized solutions with not much customization possible. The perfect example is the e-mail service. Just to show you again the differences between the delivery models, there are 3 ways to deliver eMail Service:
- The ‘traditional’ way: Completely run the e-mail service on your own in own premises
- Classic outsourcing: An outsourcing provider runs the service for you
- Cloud sourcing: You just use e.g. Google Mail and pay a monthly fee, including everything
So what are the consequences of the cloud from an inside-out perspective? IT departments will ‘shrink’ very much, there will be a general shift of Service Operation – and partly Service Transition staff – from the IT departments to the cloud providers (as they have to run everything in the future). The role of the IT department will be very much focused on Cloud Brokerage – sourcing the right IT Services from the cloud and making sure the Governance around it. Having the right Target Operating Model for that is key. The blogs of my colleagues Martin Andenmatten and Ben Martin give a very good insight into that.
Cloud: the outside-in view
Let’s now change the perspective and take a pure coprorate view. Simply speaking, the role of IT is to support business processes by delivering IT services that suit their needs. Beyond that, IT is (and should be) a black box for the business. So what’s the difference for the business having an in-sourced vs a cloud-sourced service? Well, nothing, at least from a quality point of view. With an insourced IT Service you can practically deliver towards the business the same service as in the cloud-based service (and the same time to market if you put enough money in it). For the business it’s actually not transparent where the IT Service comes from. Just think about the e-mail example. However, the big differentiator is cost. In most cases, the equivalent cloud service is much cheaper. So purely outside-in speaking the argument is cost.
I recently came across an impressive graph, which showed me that even in the vertical markets (industries) you can get a cloud solution for almost everything:
And above graph even does not show the horizontal solutions, which cover industry independent areas like HR, Finance and collaboration (That one would be probably twice as big).
So going back to the cost argument and looking at above graph– one logical conclusion would be that all IT Services go into the cloud. Maybe – but there is one additional thing to consider: The problem of standardization. For all processes, which look very similar across all companies and which do not play a differentiator (typically the support processes), going to standard is fine, going to cloud is perfect. You can gain a lot of efficiency there and align to Best Practice. However, for core processes, actually making up your business model, you shouldn’t do this, you need to make the difference. There might be an underlying IaaS or PaaS, but I never would use a standard Software as a Service (SaaS). This also reveals the double role of IT: On the one hand the traditional CIO-role, which is transforming to the Cloud Service Broker, on the other hand the emerging Chief Digital Officer, which acts as an enabler for the business. Both together will be the IT dream team of the future. I used the well-known ‘make or buy’ matrix and added these two roles (red) to illustrate this: