Pilot Projects

I spent the first eight years of my career in enterprise IT for a Fortune 15 company. Back then IT was a vibrant work environment. Money was flowing into IT to fund resources, software, hardware and vendors. Today it’s a different story. The mantra for the past 11 years has been “do more with less”. It’s good business sense to tighten budgets when the business starts to see warning signs of market stagnation. However, if the business does not begin to fund new projects to look at new technologies it will find itself in a difficult situation.

The first problem is being on outdated hardware and software. This doesn’t just exist on the desktop but also in the data center. There are still an incredible number of system still running Windows XP (I don’t have any specific numbers so this is anecdotal evidence). Windows XP was released at the end of 2001. I also talk to a number of people who have laptops or desktops that are four or more years old. Back in the data center the story is a bit better. Again this is anecdotal evidence but I’m seeing the instances of Windows Server 2003 dwindle and I see almost no Windows Server 2000. As for SQL Server, while there is still a somewhat hefty number of databases on SQL Server 2000 that number it dropping pretty rapidly, though not as fast as I’d like to see. I’m not saying that companies have to be on the bleeding edge of technology, but being on any technology that’s more than six or seven years old is plain negligent.

The second problem is outdated applications. When I was in IT our primary mission was to support the business. This entailed updating existing applications to support changing business requirements and writing new applications in support of new business opportunities and needs. There’s been plenty of articles and blog posts on the consumerization of IT. This effectively means for the first time in the history of IT consumers have access to more cutting edge technology than what IT is making available to solve business problems. I blame this back on the extremely tight IT budgets and the lack of vision by company leaders to see the value of funding IT.

So far I haven’t said a word about pilot projects – the title of this posting. Before 2000 my company used pilot projects to try out new technologies (hardware and software). We would start with small targeted projects using well defined metrics to track success. Some of them were successful and resulted in broader rollout and usage. Others, the majority, were harvested for learning and then shelved. The point isn’t to track your success rate but rather to keep trying out new things. Almost every consumer products company (think P&G) does this with new products and they’re not afraid to shoot poor performing products in the head, quickly.

I like the mantra of “fail fast”. It sets the right mindset that failures are going to happen but without them change and innovation doesn’t happen. In IT this is equivalent to running pilot projects. Take Windows Azure, for example. Many of the survey results I see ask the wrong sets of questions which I believe lead ITDMs (IT Decision Makers) down the wrong path and cloud their thinking (pun intended). Every company should either have Cloud Computing already baked into their IT strategy or at the very least have a handful of pilot projects trying out different cloud providers to solve various business problems.

Today I wouldn’t go work for a company that had its head in the sand regarding cloud. And I certainly wouldn’t work for a company that was still running Windows XP. Both speak volumes about the IT leadership and the value (or lack there of) the company places on technology.

Comments

• Anonymous
  June 24, 2011
  Excellent thought Dan! I totally agree with your points. specially, "Fail Fast". Every company IT leaders should have a special budget for the pilot project. It does help both employer and employee in long run of their business. I think, this is the reason MSFT is giving trial versions of each product. We should make use or it for the pilot project. sqlserverrider.blogspot.com

• Anonymous
  September 25, 2011
  The comment has been removed