Monday, 23 January 2012

Availability Management: Get rid of the 'nines' and bring in the $$$s.

Availability management is at the core of customer satisfaction.  Without availability, no service is delivered and there is no service revenue.  However, due to the continuous changes in patterns of service consumption, the bottom-line business impact of service downtime ('unavailability' is not a word) will vary from one instance to the next, according to the volume of demand that has not been met, and the value of that demand - after all, not all customers are equal.

IT organizations need to define how service availability translates into revenue.  Availability has become a critical business metric that IT can use to improve service provision and drive increased revenue, particularly for ‘clicks-and-mortar’ e-commerce organizations where downtime equals lots of lost revenue.  Although availability management isn’t the most glamorous corner of IT, from the user perspective it is a key influence on customer satisfaction and thus business success.  If a service is not available, revenue will be lost, the brand will sustain damage and customers will look for other service providers.  Your competitors are only a mouse-click away, so customer loyalty is lost far more easily than ever before.

Why availability is a problem
Many IT organizations are still device-oriented departments, fixated with system uptime as the key availability metric.  These have a grand total of zero relevance to the business. For IT, a ‘five nines’ availability metric (99.999% uptime) may be seen as quite an achievement (about 5 minutes downtime a year), but if that 5 minutes of downtime falls right in the middle of the busiest business period of the year, the direct impact (lost revenue) and indirect impact (brand damage and customer attrition) is likely to be job-threatening.  IT managers who let this happen will have to answer tough questions on why IT wasn’t ready.  IT managers need to understand that downtime is not random, nor does it have a uniform cost every time it strikes.

The Nines - Crap film, crap IT metric

Work from the top down
In the ITIL 3 process map, Availability Management quite rightly sits in the Service Design Processes group, indicating that it is a critical part of the process of designing a service.  Designing high availability into a service is a good start, but there is always a threat of downtime during the operation of the service coming from external factors.  Downtime is always caused by a change of some sort - either a haphazard change applied by the service provider which has impacted the service in an unforeseen manner, or a change in the pattern of service consumption that has resulted in an overload.  Thus, the business needs to be aware of IT changes so they can veto badly timed changes, and IT needs to be made aware of marketing activity that will drive spikes in demand so that they can plan for capacity to meet demand.

Thursday, 19 January 2012

Take a holistic approach to ITSM and avoid a world of pain

The Incremental Approach to ITIL

Many organizations want to improve on the cost, quality and agility of their IT service provision by implementing ITIL best practices, but this often seems like a daunting prospect.  As is often the case, the burning question is – how do we start?

You can’t boil the ocean.  Tackling 10 processes simultaneously is prohibitively complex and almost certainly doomed to failure, so the default roadmap to ITIL nirvana is to slice it up process-by-process, selecting the most urgent process (or couple of processes) for improvement first and finding a tool to that supports.  This strategy makes good sense in principle and outputs a rather smart roadmap chopped up into manageable, bite-size chunks. However, there are two major pitfalls with this process-based approach:

  • It draws organizations into a web of integrations and upgrade problems as an incremental approach to procurement is assumed by default.
  • It ignores the fact that it isn’t 100% of a single process that would bring the most immediate value, but selected elements of multiple processes – to create a ‘thread’ of value across ITIL.
Turn to page one of your ITIL roadmap and the focus may be on applying service desk automation or a solid change management process, but are you giving consideration to the last item in the roadmap?  Vendor evaluation tends to work on a very narrow “what we need now” basis.  If you need a service desk tool, you need a service desk tool, so go out and buy one that fits your needs.  Makes sense so far.  But turn to page two in the roadmap and change management is the next priority.  Does your service desk vendor supply an adequate change management tool?  No?  Then you’re looking at another vendor evaluation cycle.

The vendor evaluation cycle

Integrating two incident and change management tools may not involve a great deal of effort, but each time an additional module is added in line with the roadmap, the cost of integration is multiplied and the pitfalls of integration become all too apparent.  Each stage in the roadmap will be affected by delays, costs and risks.  Implementation is delayed by vendor evaluation.  Integration is often complex and requires expensive consultant time.  Risk of failure is high.

Oops! You failed to plan and fell into the Incremental ITIL trap.

The Holistic Approach

Individual ITIL processes don’t exist in a vacuum.  The greatest value to be gained from the ITIL framework comes from the interconnected nature of the processes.  Much of this comes from the sharing of data across processes (e.g. CMDB).  Implement one process in full and there is no data sharing.  Implement elements of Incident, Problem, Change, Release and Config and you’re leveraging that data five times over.
By slicing ITIL horizontally and tackling the must-have elements, irrespective of which ITIL guidelines they are associated with, you can create a thread of value that spans across ITIL and takes advantage of this leveraging.  Instead of taking a short-term view and putting 100% of your resources into an ‘ideal world’ Incident Management strategy, there is more to be gained by distributing efforts and tackling the must-haves for quick wins, wherever they might lie. This ‘long thin thread’ forms a solid starting platform from which you can get quick results, demonstrate the value of ITIL in action, justify further investment and develop your roadmap in a safe, stable manner.

At a lower level, there are strategies for ensuring a smooth roll-out of your ITIL roadmap.  For example, look at the information you will need to capture to run process.  By standardizing the information stored at an early stage you can avoid administrative issues later on.  A broader view of the specific data requirements can help avoid later projects to realign.  It’s a lesson that organizations often learn in hindsight and one of the areas where a consultant can really help out.

Holistic ITIL gets the "Pareto Middle-Distance Stare of Approval"

Planning is, as ever, critical to enabling this holistic path to value.  Work out what you need now, but also look at what you want in the long term from your ITSM strategy.  Having a 3-5 year view may not be of great use to you now, but it can certainly help avoid some of the many pitfalls in the future.  Factor this into your vendor selection and you stand a better chance of supporting your long term goals without getting stuck in a quagmire of integrations projects, upgrade paths and administration overheads.  Evaluating toolsets against how they can support your extended roadmap is the key to avoiding many of these problems in the future.  How many vendors do you need to work with to enable your plan?  Fewer vendors means fewer of the issues that plague enterprise IT solutions initiatives.  Selecting a vendor with comprehensive support for ITIL processes can mean that you don’t get the best-fit for each one, but it’s unlikely that your ITSM needs vary more than 20% from any other organization in the same industry vertical and gaining 80% of the value from 6 processes is better than 99% from just one.

Remember, ITIL is a means to an end.  Don’t just focus on applying ITIL.  Go for breadth, instead of depth, to leverage process overlap and interconnectivity.  Focus on the parts of ITIL that are of value to you and prioritize them in terms of value to the business as a whole.  In these days of economic flux, IT needs to be seen as an asset to the business.

Friday, 13 January 2012

Cloud Nonsense - A Collection of Cloud Computing Faff

The trend for bad cloud metaphors has been going on for some time now.  Thankfully it seems to be running out of steam and the wave of cloudmania has hopefully broken.

Bad Cloud Metaphors

 "Don't you dare associate me with this nonsense!"

Winner of the Cloud Nonsense Award 2011

Cloud computing can cause rain and land the poor villager in jail

This is a direct quote:
"Cloud computing is unreliable in a storm; if it rains, the data might get washed away. Information stored in mobile phones leaches into the phone’s battery and stays there even if the SIM is destroyed. Whatever you say hangs in the air around you for an hour, and can be sucked into recording machines and played back."