Smarter computing isn't an all or nothing proposition. In this note we explore options for prioritizing investments along the path to smarter computing enlightenment.
As I talk to clients in a variety of industries about smarter computing, some variation of this question often comes up, "If we're not doing everything you include in your definition of smarter computing are we at least doing smart computing?" Although posed humorously, it is a fair question, and one that deserves a serious answer.
How we label something reflects the way we perceive or frame it, so I am happy that IBM chose to call the combination of cloud computing, big data, and workload optimization "smarter computing" rather than "smart computing." To me, it's a reminder that we are on a journey, always attempting to improve. No matter how smart we are today, we want to be smarter tomorrow. That said, there is no "right" path to smarter computing that will be appropriate for all enterprises, but there are some general guidelines that may help an enterprise manage its IT investments more effectively.
First, it's important to identify where you are, and where you want to be - or can be - in a reasonable time given your current infrastructure, budget, and constraints including culture (resistance to change is a bigger barrier to smarter computing than old technology or small budgets). In the figure here, we present a simple model of smarter computing adoption to guide or at least inform your process.
A checkmark in a cell indicates an investment in the associated technology with some commensurate level of adoption within the enterprise. The labels on the left identify levels of enterprise sophistication or maturity that we associate with effectiveness (higher levels, in general, equate to more effective use of IT).
Absent simply indicates that none of these technologies are yet in place. In many organizations, business units begin to experiment with cloud computing and even analytics but don't coordinate these exploratory efforts with an IT function. We would still classify these enterprises as absent until the adoption of cloud computing or analytics is part of an overall strategy rather than an individual project with plausible deniability for senior management.
The novice level indicates that exactly one of the technologies is in use and supported at least tacitly by IT, intermediate is any two, and advanced indicates an enterprise that has adopted all three. Within the novice and intermediate levels, relative position has no meaning (a higher position within a level does not indicate a more effective or mature IT environment).
We are not aware of any major enterprise moving down levels, except perhaps due to divestiture or downsizing. Outsourcing any of these technologies would not impact our classification of the enterprise that ultimately derives benefits from the technology. As an enterprise moves to higher levels, it can get economies of scale and synergies between the technologies. For example, workload optimization enables more efficient cloud-based applications and cloud computing enables scalable big data applications.
Unlike traditional maturity models, which borrow from Abraham Maslow's hierarchy of needs, the model we present here does not imply that a firm must go from gray to red to white to blue. It is possible to invest in these technologies independently and in any order or combination, with adoption driven by business requirements.
Firms with no current smarter computing investments and no established demand for exploiting analytics will likely choose cloud computing as a starting point. It is the least expensive place to get started, as one can experiment with cloud deployment on pilot projects and proofs of concept for the price of a daily latte. If budgets are tight, this is also a smart move because out of these three technologies, cloud computing may generate the fastest return on investment based on reduced costs and increased flexibility, including a decrease or deferment in capital expenditures.
Adoption of big data with analytics typically comes next. Although these concepts are not "new" (firms from American Express to WalMart have been used as business school case studies for these topics for many years) what is different now is the scale and predictive power of the technology and the availability of tools that make it practical for even mid-size businesses to leverage this newfound intelligence. Businesses in all industries can deliver new, custom offerings to their clients based on behavior and market conditions in time to have a business impact and at costs that were unheard of just a few years ago.
Finally, in a typical environment, comes workload optimization. Wringing the last ounce of performance out of IT investments demands this optimization. Of course some tasks may inherently require it and make it a first order priority (such as answering Jeopardy questions in 2-3 seconds with no Internet access).
We recommend that firms identify their target maturity level 24-36 months from now by asking a few potentially tough questions, and then work backwards to their current position to find an appropriate path. This requires strategic thinking and should involve decision makers within and outside the IT function. Sample questions would include:
How important are issues like flexibility, deployment time, capital expenses vs operating expenses, and physical control over one's data? (Asked to help determine priorities for cloud computing)
How might we benefit from the development of applications that leverage insights from customer behavior and market conditions to provide more responsive or innovative products, to deliver them more efficiently or effectively? (Asked to help estimate the likely payback from big data/analytics investments, which do more than the others to facilitate growth)
How often do we find ourselves capable of solving our IT problems, but not in a reasonable time or within our budget? (This will help management prioritize investments in workload optimization, which makes the best use of available resources)
Returning to our original question, if your enterprise is at the absent through intermediate levels, just how smart are you, and should you care? If your enterprise runs on information, and you're not in the advanced level and have no plans to be there, it might be a good time to look at your competitors to assess their smarter computing investments and results. If your competition makes these investments before you, they will likely reap the benefits first, too. They may not be smart yet, but they just might be smarter than you.