How big should your Capacity team be?

Posted on 24/09/2011 by James Collings

It sounds like the start of a particularly poor joke, but IS there an optimum size for a Capacity Planning team?

This might come down to the heart of the issue for many Capacity experts. Do you want your team to be 100% occupied, or do you want them to be 80% (for the sake of argument) busy so that they can respond to ad-hoc project requests without delays? And is there a 'golden ratio' between the number of services to be managed and the number of planners required?

Sadly, in the author's opinion... No!

Some services are very complex and so crucial to the success of an organisation that they require the dedicated and undiverted attention of a single person. Other services might be so trivial and unchanging that a single person could do 10 or more. So if you can't set the number of people in a team using some global equation, are there at least a set of roles that every team should have?

Yes! (phew... Otherwise what would be the point of this article?)

Every capacity team should have a manager that actually understands the subject matter. This person might be a 'team leader' if that fits into the local organisation tree more appropriately, however the fundamental role still exists. A manager that does not understand the intricacies of what a team of highly specialised analysts are doing and the challenges they face is not going to be able to properly represent them higher up the organisation. Also, many technical analysts are not the most skilled presenters, and therefore a manager that can present this technical work at IT Board level is an absolute must. Having an experienced manager that can be 'hands on' also means that they can cover for periods of sickness, holiday, or unexpected peaks of the team's workload. Any organisation that relies on having a 'generic' manager heading up the capacity team will have to employ an additional 'analyst' to provide this cover and will therefore be wasting money.

Each capacity team will also benefit from have a data analyst. This role can typically be undertaken by a junior member of staff, or by someone that is just starting their career in capacity management (having transferred from a help desk role or similar). This team member will be responsible for ensuring that the capacity toolset is working correctly, and also for ensuring that any regular reporting is delivered according to schedule. They may also extract system, application, and business data from the Capacity Management Information System for use by the technical analysts within the team. In this way, they will gain an appreciation for the requirements of the more experienced analysts within the team, and also free up those analysts time so that they can concentrate on the 'dark arts' of modelling and planning.

The only other essential role within the team will be the capacity analysts/planners/managers. They are the heart of a capacity team, the number of which is completely organisation dependant. As mentioned at the beginning of this article, there is no magic formula for working out how many analysts you need. The only truism is that the more that you have, the more responsive they can be to fluctuations in workload and when not responding to new requests the more detailed their work can be on their core services.

So, in summary. A team needs a 'hands on' leader, a 'number crunching' data analyst, and then a number of experts. I have set up capacity management functions in which I have filled all three positions, but once a team grows I have found that this separation of roles is the most effective.

Posted in Blog Posts, Capacity Management | Tagged , , , , , , , , , , , , | Leave a comment

Is Freelancing the solution to the recession?

Posted on 30/08/2011 by James Collings

Startup Britain

At the start of this week, I was able to join the Startup Britain tour of England and Wales at the Bristol leg of its journey.  The initiative is a new campaign by entrepreneurs for entrepreneurs, launched on 28th March 2011. Designed to celebrate, inspire and accelerate enterprise in the UK.  Will it work?  I don't know, but anything that has a chance of improving the economy of this country deserves support.

I was there to answer any questions that attendees might have had regarding freelancing.  From the description above, one might be tempted to think that the whole initiative is aimed at getting as many people as possible to be freelancers as possible.  Nothing could be further from the truth.  Entrepreneurs come in all shapes and sizes... and so do their business models!  Most businesses begin with a good idea.  This is fundamental to the success of that business, because if you start with a BAD idea (or even a mediocre one) then getting customers is going to be an uphill struggle.

But after the initial idea... what happens next?  Regardless of whether the business is selling a product or a service, typically one or two people try to make it a success.  Some businesses grow quite quickly, gaining turnover and staff as demand for their product/service increases.  Startup Britain was there to give advice to businesses in exactly this position.  Let's face it.  Going from one employee to two is a doubling of the workforce.  It is a major step.

However, some businesses stay as a single (or few) person enterprise.  Their way of growing is to increase profit margins and turnover.  Once again, there were people with me on the Startup Britain tour who could advise entrepreneurs how best to achieve this growth using their own, first-hand, experience.

And then there were the attendees that are still at the "idea stage".  They haven't taken the step to turning that idea into a business, and wanted to get as much advice as possible before taking the plunge.  Those that were considering the freelance model for their business were the ones that I wanted to help.  But does that mean that I wanted them all to become freelancers?  No.

Horses for Courses

Freelancing can be a difficult way of life.  Certainly, there are advantages.  The ability to pick and choose the projects that you work on, the escape from office politics, and the work-life balance are 3 of the most common reasons that freelancers give for their chosen way of working.  There are others who cite the ability to get paid exactly what they’re worth, rather than conforming to an HR dictated salary policy.  All of these things sound very attractive.  Which begs the question, why isn’t everyone a freelancer, and why do many return to permanent employment?

There is no job security for a freelancer.  It is not possible to “keep your head down” and “avoid the chop”.  If a freelancer does not deliver, they will lose the work that they are doing or not get a renewal for future work.  This is compounded by the difficulty in securing another contract with the spectre of a failure hanging over you.  And while the freelancer is trying to find another piece of work, there is no way of paying the bills.

So, if you have decided that being a freelancer is for you, then talk to the PCG about what it really entails.  However if you merely fancy a change from your current permanent job, then think this through carefully… again.  Some employers are wary of taking freelancers back as salaried employees.  Having jumped aboard the “freelance train” you must be prepared for the long journey.  The PCG can help you with this, but when the chips are down you will have to rely on your own strength of character to see you through.

Posted in Blog Posts, Freelancing | Tagged , , , , , , , , , | 1 Comment

Can Tools ever replace Analysts?

Posted on 04/08/2011 by James Collings

Over the years, the quality of tools available to the Capacity Manager have improved immensely. Little's Law may have been written before the advent of mass computing, but its application is crucial to predicting computing capacity. One of the earlier applications of the Law was to predict queuing on telephone exchanges. Whilst I am too 'young' to remember those days, I am reliably informed that it was common practice to employ teams of people to watching a bank of lights and manually count the number of flashes occurring within a defined period. The Capacity Manager would then use these observations in their analysis and predictions.

Things have moved on immeasurably since then. Not only has the use of computers increased, but the automatic collation of performance and capacity data has meant that the Capacity Manager can effectively define the metrics that they want recorded, go home for a day or two, and then return to analyse the output.

Capacity Toolsets (even simple spreadsheet programs) make the effort of analysing, modelling, and reporting so simple that many Capacity Managers in the industry today do not need to know what Little's Law is or how to apply it using pencil and paper. But can the tools do even more? Are we able to do away with the trouble of having a specialist Capacity Manager completely? Surely with the intelligence and capabilities of the Toolsets today, anybody that can point a mouse and click can be a Capacity Manager?

A major step in the work of a Capacity Manager is to identify the cause and effect. What business activity is causing the profile of resource utilisation, and what will be the effect of a change in that activity? To do this effectively, the Capacity Manager needs to understand the business processes that the IT service is supporting. They needed to ensure that they are measuring application transactions that relate to the business processes, and then they need to find a robust link between the application transaction and a measure of resource utilisation. Throw in a business forecast (which will need manipulation to turn it into an effective forecast of application transactions) and you have a recipe for quite a powerful artificial intelligence neural network.

Certainly, a Capacity Toolset can try to correlate EVERY resource metric against EVERY application transaction, and report the closest fit. But what good is a correlation between the count of inodes in use and the number of new customer records, if the system can only store a limited number of customers due to licence restrictions.

So, to answer the original question... No. A tool will never be able to replace an analyst. I would suggest the reader considers the analogy of Word Processing. These days, anyone can produce a newsletter or simple flyer for themselves... however Word Processing tools haven't turned us all into publishers able to produce a newspaper or magazine!

Posted in Blog Posts, Capacity Management | Tagged , , , , , , , , , , | Leave a comment