The idea for this blog post came while listening to Mark Graban’s excellent podcast on Lean Thinking. I learned a couple of things that got my thinking about IT operations going. What particularly caught my attention were two still very common underlying methodologies: ITIL and the good old waterfall model. Although both are around for a longer while, I’m seriously wondering why. Are these methods good enough to address today’s challenges of IT operations?
Today, delivering 99.9% service quality is already a very good result. Actually, that seems to be a new gold standard for cloud SLAs. But is this really enough?
Well, this 0.1% of service unavailability doesn’t sound like a serious issue that would make much of a difference. But in fact, it translates into:
- 2.15 million of emails per day not reaching the recipient
- 3.5 million of Google searches per day failing
- A system downtime of 8h 45min per year
And these are already impressive numbers with millions of unhappy users behind them. So, rather than lean back and relax, IT business needs to set new goals and integrate lean management into the current process frameworks.
As explained earlier in my post on 5S in Data Management, I’m a fan of Lean (and Lean Startup) and Six Sigma. I encourage you to look at these methodologies and understand them as a very valuable complement to ITIL based operations. Why is that?
- First and foremost, both methodologies date back to the 80s. This timeless success caught my attention.
- The underlying concept is that perfection – if at all – can only be achieved through countless iterations. I love that relaxing thought.
- Iterations add flexibility and agility to operations which I consider as essential. Reality tends to find its own ways of proving our assumptions wrong.
- ITIL is an excellent repository, but the continuous improvement is somewhat “underrepresented”.
- Lean Thinking puts customer / user value into the spotlight. ITIL is prone to distract people from that because of its complexity. Many organizations I worked with were faced with the risk to turn their focus too much inward and make themselves occupied with managing the complexity rather than contributing to the business outcome.
Now getting back to my example of system uptime. Six Sigma targets 99.9997% quality of service (actually Motorola targeted 3.4 defective parts per million when inventing SixSigma) which amounts to an astonishing 1min 30sec downtime per year. SAP and the Standish Group assume that 67% of SAP customers have an average downtime of 9 hours per month.
I’m convinced that we should roll up our sleeves and learn as much as we can from the manufacturing community. Lean & SixSigma tools such as DMAIC, Kaizen, 7 wastes are extremely valuable and have proven higher business outcome for decades.