♦  IT Priorities Research

Cloud Computing

At the core of cloud computing is the premise of computing as a commoditized resource delivered as a service. The commoditization occurs when a technology becomes ubiquitous and well-understood, and therefore a weak source of competitive advantage. Sooner or later, all technologies go through this commoditization process. This typically involves the transition from a crafted object, to an industrial product, to a service. In IT resource terms, such a transition involves the centralization of computing resources and reliance on specialist providers.

Centralization is essential because it engenders both specialization and the scale necessary to substitute products with services. Centralized service providers rely on a range of enabling technologies in order to generate and deliver computing through a services-based utility model. Virtualization, clustering and grids are the three principal enabling technologies behind cloud computing, although others are influential. These technologies decouple IT services from specific devices and allow computing to be sourced from massive pools of shared resources, rather than from specific, identifiable devices. The distinction appears trivial but the implications are profound and, in Datamonitor’s opinion, capable of market disruption.

Learn more about cloud computing in Can Cloud Computing Help Enterprises Weather the Economic Storm? (DM-2267)

IT Spending

Last year, IT capital budget growth flattened while IT operational budgets were still growing. This shows organizations first cut capital spending before operational budgets. This is why the current recession feels worse to technology vendors than it does within many IT organizations. New sales for technology vendors are largely funded through their customers’ capital budgets. Capital spending has not grown for two years and may not recover until next year, at the earliest.

Just as reductions in IT operational budgets may not be over, cuts in IT capital spending may still be in the offing. 43% of IT executives expect that they will not spend all of the money currently allocated in their IT capital budgets for this year. Only 9% expect they will be able to spend more. The good news, however, is that the remaining 48% expect to spend all of the money currently allocated for IT capital projects.

Learn more about IT Spending in 2010 in IT Spending and Staffing Benchmarks 2009/2010 (CE-4201)

IT Headcount

Over the long term, technology has enabled businesses to flatten bureaucracies and push more responsibility down the organizational chart. Nevertheless, the ratio of IT management positions to staff has been stable over the past three years. The median ratio of IT management positions to total IT staff for our sample was 10.4% in 2008, compared to 10.3% in 2006.

Learn more about IT Staffing Levels in IT Staffing Ratios: Benchmarking Metrics and Analysis for 14 Key IT Functions (CE-5003)

What Makes a Good IT Metric?

Good metrics are SMART: specific, measurable, attainable, repeatable, and time-dependent. Metrics must be specific and detailed in order to be useful. For example, "Server Performance" is a bad metric, as it’s unclear what this actually means and how it would be measured. A far more specific and measurable metric is "Peak Daily CPU Utilization % for Server ACCT0101." This definition is clear and the means to measure it are straightforward.

Furthermore, a metric is only attainable if it is realistic. Aiming for "100% Uptime for Server ACCT0101" may be a laudable goal, but even fault-tolerant servers experience the occasional problem, however minor. A more attainable metric might be 98% average uptime over one month. Metrics only have meaning when they are repeatable and can be consistently captured and measured over time. Metrics must be presented within the context of time (e.g. average daily rate, peak value over a month, or a ratio captured every month-end).

Good metrics must be valid, meaning that they indeed measure the activity that they are supposed to measure. Reliability is also key, meaning that metrics are characterized by the quality of their underlying data, and are designed in such a way as to accurately convert that data into a validated and accurate metric. For each metric under consideration, it is crucial to determine if and how the activity associated with that metric can actually be measured. Even where data is measured and captured, there should be a provable relationship between what is measured and what you want to know, which then can be expressed in a model.

from Making IT Metrics Work (IN-6525)

Negotiating Better Contracts to Save IT Money

Distinguish between interests and positions. Opposing parties will take positions that are symbolic representations of underlying interests. Consider this classic example: Two individuals are arguing over an orange. A deadlock seems imminent, as there is only one orange left, and both want it for themselves (the position). However, in examining the underlying interests, we find that one wants the orange for making juice, while the other wishes to use the rind for baking. Thus, a solution is negotiated whereby more value can be created out of a limited resource.

• Asking questions such as, "What is motivating you here?" or "What would you like to accomplish?" will help identify compatible interests.
• Create a prioritized list of items that are of value to both parties and identify areas of compatibility.
• Separate people from problems. Instead of thinking in terms of "Who is right?" think of "What is right?" Control emotions and biases.

from IT Cost Cutting Bundle (IN-6521)

Browse our FREE Top IT Priorities reports