Tom made two observations about the types of applications his company is working with--both of which pose a challenge to the traditional application planning of the past.
The first is the changing nature of an application's user population. Historically, as noted, most applications were focused on automating internal business processes, which meant relatively stable loads and numbers of users. With the current move to making applications accessible to external parties (via distributed co-creation), the load on an application can vary tremendously. In other words, external users are now becoming a significant part of the application user population, and the numbers and patterns of users puts paid to the old expectations of stable user population sizes.
The second was even more striking than the first. Tom said that as business groups within companies get more familiar with the capabilities and opportunities offered by integrated applications and distributed user populations, they start designing more and more innovative business offerings that assume the ability to handle unpredictable loads and manage very large amounts of traffic.
My interpretation of these observations is this: if you accept the validity of McKinsey's distributed co-creation future and that companies will absorb the ability to leverage those capabilities, the established practices and assumptions that underpin most IT organizations are no longer applicable. These practices and assumptions are designed for a world of stability, consistency, and little variation, while future business trends will undermine every one of them. The future demands vastly more scalability, elasticity, and agility.
It's impossible to predict exactly what this future will look like, but for sure, if you're an IT organization, as the saying goes: "what got you here, won't get you there." Here are some thoughts about how to prepare:
1. Find every commodity application you run and create a plan to migrate to a low-cost reliable producer
Email is the poster child for this migration. It's hard to understand why any organization runs its own email infrastructure. Spending any more than the minimum on any non-differentiating application consigns you to the McKinsey also-rans and ripe for outsourcing.
2. Create a cloud action plan, stat
The demand from business units will arrive enormous and insistent. Trying to hold them off while the three-year transition plan comes into play is a recipe for what is termed "shadow IT." When someone responsible for a P&L sees a competitor rolling out an offering like a Denny's, the first question is "why can't we do that?" rapidly followed by "how soon will you be able to do that?"
3. Get ready to improve your organization's skills
I can't tell you how many companies we interact with that don't understand that building scalable, elastic apps requires more than an underlying cloud infrastructure. Both development and operations skills need to be learned, and fast.
Sign up for Computerworld eNewsletters.