I’ve gathered these 10 stats by scouring the web to find 10 interesting IT stats. I found 10 without having to resort to a stat about cloud computing.
- Global IT spending estimated to fall by 9% in 2009. (Goldman Sachs) This estimate is a decline of 5% from Goldman’s original estimate of a fall of 4%. It has been “corrected ” on the basis of recent figures for GDP, capital spending an corporate profits. Goldman also estimates that the decline will be greatest in developed countries (averaging 12%.) The hardest hit part of the IT market at the moment is hardware, where server and desktop PC revenues are down by more than 20%. In November 2008, IDC was estimating that IT spending would grow by 2.6%. Those were the days…
- Worldwide semiconductor revenue to decline 24% in 2009 (Gartner) It’s brutal for the semiconductor industry. Revenue is forecast to decline to $194.5 billion in 2009, from about $256.4 billion. In mid-December of 2008, Gartner was forecasting a mere 16 percent decline. Gartner’s projections are that it will take 3 years to get back to 2008 levels. That’s how long it took after the dot com collapse.
- IT outsourcing accounts for 7% of India’s GDP (Nasscom). It also accounts for 33% of India’s foreign-exchange inflows and employed about 4 million people in 2008. There are varying opinions about what will now happen. The cost advantage of outsourcing to India that used to be at least one sixth has now fallen to one third and other developing economies across the world are muscling in to this market. Some commentators expect India’s IT economy to grow while others believe that much of the outsourcing will be replaced by automation. The recession may lead to some economies raising barriers to international outsourcing.
- 43% of data centres are running out of physical space (Aperture Research Institute, 2009) Maybe the recession will help on this one or maybe not.) Nearly 90% of those surveyed said that 75% or more of the data center space was already used and more than 43% of respondents reported that usage was above 90%. The situation is exacerbated by the fact that servers use more power than ever before putting pressure on the power supply infrastructure – and in some case on the capability of the local grid to provide more power.
- 40% of all application downtime is configuration-related. (Forrester) You have a right to be skeptical about any statistic that is a big round number. here’s another one on the same topic. “Application downtime is caused almost 60% of the time by application configuration problems.” (Enterprise Management Associates, 2008). I’m skeptical that the exact figure is 40% (especially as the other survey says almost 60%) but the order of magnitude is no surprise to me at all. The truth is that most configuration is unautomated. It’s a manual process done by editing parameters and writing scripts and such activities are error prone. I know this from my configuration of this web site which is not exactly error free. If only there was a nice open source tool I could use….
- The average cost of a single hour of application downtime is about $45,000. (Enterprise Management Associates, 2008) The figures are an average that spans low mid-tier companies to large corporates. There is a tyranny in averages of course. For transaction intensive organizations the cost has to be at a much higher level – possibly up to millions of dollars – and, at the other end of the normal distribution, the costs will be much lower. For a different opinion, try Forrester’s survey. They suggests that, for 42% of organizations, the cost of downtime is above $100,000 an hour. The spread was like this: 33% estimated the hourly cost to be between $10,000 and $100,000; 25% estimated between $100,000 and $500,000; 13% estimated costs to be between $500,000 and $1 million; and 4% estimated the cost to be greater than $1 million per hour. Another survey (Infonetics Research, 2004) suggests that overall downtime costs for the average organization come in at 3.6% of revenue. Looks a bit high to me, but who knows…?
- 10% of IT budget devoted to IT security (Forrester, 2008) Depressing isn’t it? I think so. And 20% of those surveyed believed that IT security spend would rise in 2009. I suspect that’s not now likely to happen, but in a room where people are arguing for budget, who’s going to suggest cuts in the IT security budget? The usual stats that never seem to head south about the number of companies suffering security breaches of one kind or another continue to appear. There’s reason to believe that the recession will simply swell the ranks of the bad guys.
- 30% of software projects fail outright (Standish Group, 2008). The full figures from the Standish Group’s report are: 90% of software projects are late, 66% are deemed failures and 30% are scrapped. More remarkable is that this set of stats has been pretty muchteh same in every Standish Report on this topic that I’ve ever read (since God was a boy.) The fix is to automate the development process better. But every time we get close to doing that technology changes and we run off to develop something quite different using different languages and different methods. It may be a law of the universe that 30% of software projects have to fail.
- Worldwide virtualization software revenue to increase 43% in 2009 (Gartner, 2009) This is no surprise as a projection. If you didn’t know, virtualization revenue was $1.9 billion in 2008 sand 43% growth will put it at to $2.7 billion in 2009. This is a mature technology on the server. What’s less certain is the take up on client virtualization. Gartner estimates that revenue from client virtualization will go ballistic, tripling from $74.1 million to $298.6 million in 2009.
- 70% of typical IT budgets go to running existing IT applications and infrastructure (IBM Global CEO study, 2008) I thought it was worse than that – more like 80%. I’ve seen stats which say that – from old surveys. They were describing ratios of that ilk 20 years ago. So maybe this is another “law of physics” type of stat. Consider this. If you look at an application from a life-cycle perspective then you normally discover that only 10-20% of the costs are in system development (when you do all the development yourself). This means that the rest of the costs for the life time of that app will be in infrastructure and running the app. So even if you start with a “green field” you should get to a 80:20 or 70:30 ratio after a few years of building and burning in new apps. It’s a dubious stat imho.
