I recently ran across an interesting post by Eric Lundquist from Information Week. Eric’s primary premise is that Dell’s disappointing results and HP’s most current woes are in part due to the big “A’s”: Apple and Amazon.
The Apple Effect
Lundquist references Analyst Ben Reitzes of Barclays in his argument. Reitzes says, “results and commentary around tablets suggest to us that Apple continues to disrupt the traditional PC market and gain momentum through tablets and smartphones.” This argument is reinforced by Mary Meeker’s most recent research that shows that nearly 30 percent of U.S. adults own a tablet computer or eReader — amazingly, up from only 2 percent just three years ago. Explosive growth in this segment is an understatement here.
Lundquist’s position is that the Apple effect is having not only an impact on the consumer PC business, but also may be affecting HP’s traditional printer business as well. Eric posits that, “the widespread use of tablets means fewer documents are getting printed.” I fully agree with Lundquist here; in days past I would print stacks of PowerPoints and PDFs before I got on a plane. Today, I move them to the iPad, iPhone or Kindle to read. This is still a somewhat painful process, but it absolutely beats reams of paper filling up precious carry-on space.
Putting this trend into context for outsourcing customers, especially those with existing end-user computing (EUC) agreements, it’s becoming clear that the growth of non-traditional PC segment is a key area to keep an eye on. Given the torrid pace of growth since the first quarter of 2010 of what is essentially a new segment, and the increase in interest in Bring-Your-Own-Device (BYOD) models, customers are increasingly going to be grappling with how to converge traditional outsourcing agreements with emerging multi-device trends.
ISG is currently conducting research on this convergence; we’ll be publishing the results later this year. If you are interested in participating in the survey in order to receive a copy of the full results, please let me know.
The Amazon Effect
Lundquist’s position here is that Amazon’s Web Services (AWS) is working its way into enterprises, which may eventually affect Dell and HP’s ability to capture higher-margin cloud business. It’s clear that Amazon is disrupting the traditional infrastructure space, but the degree to which that’s occurring in large enterprises is still yet to be determined. We’re seeing quite a bit of AWS usage at our clients, but it’s still very much on the periphery, primarily used outside of IT for greenfield projects.
What really piqued my interest was this statement:
“Amazon’s infrastructure-as-a-service is considerably ‘stickier’ than software-as-a-service offerings from Google or Salesforce.com. It is much easier to switch in and out of software applications than it is to start shuttling around your IT infrastructure.”
This is where I disagree with Lundquist. I’d argue the exact opposite: software-as-a-service (SaaS) is much “stickier” than infrastructure-as-a-service (IaaS).
Applications are logical representations of a company’s business processes. When you change an application, you also change a business process, and changing business processes is hard. Attacking entrenched business processes does not happen often, and when it does happen, it generally takes a long time to fully execute. So companies don’t do it very often.
This is not to say that changing the underlying infrastructure is not hard as well, because it is. But it’s hard in a different way, because it’s almost always an IT-focused event. It requires close coordination between the infrastructure and apps teams, but not nearly as much interaction and buy-in from business users. Yes, business users need to test, but they don’t need to change.
Having been both a developer and a project manager over many types of these changes, I can say without a doubt that I’ll take an infrastructure change over an application change every time. I’ll bet many of the readers of this blog would say the same.
Putting this argument into context for outsourcing customers, I think “as-a-service” application platforms, once implemented, will have a longer shelf life than “as-a-service” infrastructure platforms, and will therefore be much ”stickier” when it comes to commitments that companies make around these platforms and associated services. This is for two reasons: First, the reasons discussed above — it’s harder to change at the application level than it is to change at the infrastructure level, primarily due to organizational inertia. Second, the level and volume of industry thought leadership around cloud interoperability is heavily weighted to the infrastructure side of the house. My prediction is that we’ll see IaaS interoperability standards and frameworks long before we see the equivalent on the SaaS side.
In theory, there will be a clearer path to migrate between IaaS providers than between SaaS providers, which should make IaaS less sticky than SaaS. This is still theory, of course. Time will tell how much progress is made in this important area, especially among the traditional IT outsourcing providers.
Let me know what you think — agree or disagree?