As an IT project manager, one of my key goals is to determine how long a project or programme will take. The problems faced when estimating timescales relate to the areas which are difficult to quantify, for example, unfamiliar company culture, unfamiliar business function, managing a new or unknown project team, etc.
A real skill of project management is identifying, estimating impact and making provision for these important, but difficult to quantify variables. Like all senior project managers, I depend on my expert knowledge, experience and intuition when making assessments.
A few months ago, I attended a fascinating lecture at the LSE entitled “Super Crunchers” given by Yale Professor, Ian Ayres. Professor Ayres in his lecture and in his book of the same title, gave a number of excellent examples of how using even crude mathematical models and working with available data sets that computer generated predictions are often far superior to those of acknowledged experts.
He gave widely diverse examples of the success of mathematics over established experts - from predicting US Supreme Court rulings using a relatively crude mathematical model; to predicting the quality of a Burgundy or Bordeaux vintage using a small number of available environmental criteria; to how much a new movie will take at the box office and what changes have to be made to increase box office earnings.
In the world of IT, it seems that almost everyone believes they are experts and have excellent intuitive and judgement skills. The over optimism of my IT colleagues and the resulting consequences are something that I have to deal with professionally on a daily basis. The reality is, most people are poor at assessing risk and are even worse at estimating how long a piece of work will take, particularly if they are involved in this work.
Most of us conveniently forget our failures very quickly or just cover them up. Perhaps it’s the best approach and a very good survival technique both psychologically and professionally. However, as an IT Project/Programme Manager, the over optimism of others is something I have to successfully assess and manage as it can have serious economic consequences for both project and client.
Professor Ayers’ lecture caused me to revisit a number of techniques learned in the past including Statistical Analysis and Judgement and Decision Analysis techniques. The problems with these techniques are: firstly, they were time consuming; secondly, there is a tendency to get bogged down in the minutiae and finally, often the end results are not defendable against common sense analysis and project experience.
A simple approach was required that could be used and easily refined after the Critical Path of the project had been specified. I selected a small number of key variables and applied a simple weighting to each. The criteria selected included:
- Business Function – Marketing, HR, Finance, Customer Services, etc.
- Technical Component – Web Dev, Software Dev, IT Infrastructure, Product Installation, etc.
- Key Skills – Graphical Design, Software Engineering, Data Integration, etc.
- Average Years of Experience of Key Participants Experience – Designers, Web Developers, Software Engineers
The simple formula was then applied with the weighted tasks on the Critical Path. I have applied the technique to a number of past projects and after some adjustments achieved good results. The weakness being that hindsight is always perfect.
I am now applying the weightings and formula to new projects. I expect there will several iterations before I am happy with the results, but I am hoping that we will be able to use the approach to increase the accuracy of estimating project duration for the benefit of clients.
When I am happy with the approach and the results obtained, I will publish further details.
No comments:
Post a Comment