I’ve weathered a bunch of “efficiency” drives, including Total Quality Management, ITIL, Six-Sigma, SEI CMM, and seven different “centralisation” initiatives. While each drive brought us some value, all of them ultimately failed because they misunderstood what it means for an IT shop to be considered efficient.
Random House’s dictionary illustrates the problem: you can either define efficiency as (a) accomplishing a task with a minimum of time and effort, or (b) accomplishing a task with the least wasted time and effort. Those two definitions are not interchangeable in IT service delivery thanks to resiliency, stability, and progression factors.
Resiliency (in this context) represents the organisation’s ability to endure surges (increased workload) and strife (decreased capacity). One quality initiative declared that we’d pare down our staff to the absolute minimum engineers required to provide basic tech support. The plan promised to save a bunch of payroll, but would break the work unit: absolute minimum staffing meant that no one on the team could ever get sick, or have family issues, or take a holiday. Once you removed just one high-demand/low-density engineer from the schedule, support failed. When a new product launched that required extra attention, support failed. You must buffer your staffing to accommodate the ebb and flow of business operations.
Stability represents the organisation’s ability to protect itself from known preventable harm. An executive once asked IT to give every user local admin rights over his or her company PC so that they wouldn’t have to wait on IT to install software, load device drivers, and so on. The executive swore to the heavens that no one would ever abuse their vastly increased power… We politely refused because we’d already caught the louse abusing his tier one augmentee rights to install videogames on his laptop. Letting everyone have admin may decrease helpdesk workload in 80 per cent of cases, seeing as most users are decent, honourable people. It’s the handful of rogues that drastically increase the workload of the much more expensive security department. An efficiency gain that increases vulnerability is a terrible idea.
Progression represents the organisation’s ability to evolve beyond the status quo – to experiment with and deploy new and better solutions for the business. I’ve fought viciously against “right-sizers” who were convinced that two inexpensive tier one engineers were better value than one expensive tier two engineer. Their logic was that we’d get save money by purging experienced senior staff; after all (they said), a junior tech with a solid checklist can solve the same tech problem as a senior tech working from memory, right? Wrong! Seasoned staff provide advanced troubleshooting skills, and also mentor junior staff on both tech work and tribal knowledge. Advanced engineers cost more because they’re worth keeping.
To be clear, I’m a huge fan of making processes lean and consistent. I enjoy eliminating organisational waste. I want my IT departments to run smoothly, reliably, and free of drama. That being said, I also want my IT service capability to be as strong and as flexible as I can make it, so that we can consistently provide high-quality coverage without the risk of being casually degraded by changes in the operating environment. The more efficient an IT organisation becomes on paper, the more brittle it usually becomes in practice.
That’s why I recommend being wary of any new fad or product that claims to deliver massive savings. If one of a solution’s steps involves eliminating your critical personnel, politely decline. The efficiencies gained won’t be worth the harm that’s inflicted. Invest in your people instead.